UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Juvenile-mature correlations in selected Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] provenances.. Musoke, Rachel 1981

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

Item Metadata

Download

Media
UBC_1981_A6 M09.pdf [ 3.59MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0095201.json
JSON-LD: 1.0095201+ld.json
RDF/XML (Pretty): 1.0095201.xml
RDF/JSON: 1.0095201+rdf.json
Turtle: 1.0095201+rdf-turtle.txt
N-Triples: 1.0095201+rdf-ntriples.txt
Original Record: 1.0095201 +original-record.json
Full Text
1.0095201.txt
Citation
1.0095201.ris

Full Text

Juvenile-Mature Correlations i n Selected Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] Provenances and Progenies  by Rachel Musoke B.Sc.F., Makerere University, 1977  A Thesis Submitted i n P a r t i a l F u l f i l l m e n t of the Requirements for the Degree of Master of Forestry in The Faculty' of Graduate 'Studies Faculty of Forestry  We accept this thesis as conforming to the required  standard  The University of B r i t i s h Columbia July, 1981  (c)  Rachel Musoke, 1981  In presenting  t h i s thesis i n p a r t i a l fulfilment of the requirements  for an advanced degree of the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available f o r reference and study.  I further agree that permission for extensive copying of this  thesis f o r scholarly purposes may be granted by the Head of my Department or by h i s representatives.  I t i s understood that copying or  publication of t h i s thesis f o r f i n a n c i a l gain s h a l l not be allowed without my written permission.  Department of  The University of B r i t i s h Columbia 2075 Wesbrook Mall Vancouver, B.C. V6T 1W5 Canada  i  ABSTRACT  Growth and branch characteristics of thirteen year old Douglasf i r trees were analysed with the objectives  of p a r t i t i o n i n g the  variance into additive and non-additive, estimating and  estimating  juvenile-mature genetic  correlations.  heritabilities, High correla-  tions could be used i n early selection to reduce the progeny testing periods with possible advantage of increasing and hence genetic  selection d i f f e r e n t i a l  gains.  Most of the t r a i t s rendered non-significant consequently non-significant  heritabilities.  additive variance,  Among the juvenile  t r a i t s , embryo class and dormancy period revealed  s i g n i f i c a n t genetic  correlations with the thirteen year old root c o l l a r diameter (0.73 and 0.32 r e s p e c t i v e l y ) .  This highlights the possible p r e d i c t a b i l i t y of  root c o l l a r diameter correlated response as a result of early selection based on embryo class or dormancy time.  ii  TABLE OF CONTENTS Page ABSTRACT  i  TABLE OF CONTENTS LIST OF TABLES  i i i i i  LIST OF FIGURES  iv  ACKNOWLEDGEMENTS  v  1.  INTRODUCTION  1  2.  LITERATURE REVIEW  5  2.1  Variation  5  2.1.1  Quantitative Genetic Methods  7  2.1.2  Qualitative Genetic Methods  9  3.  4.  2.2  Heritability  11  2.3  Juvenile-Mature Correlations  15  MATERIALS AND METHODS  25  3.1  Variation  30  3.2  Heritability  37  3.3  Juvenile-Mature Correlations  38  RESULTS AND DISCUSSION  42  4.1  Variation and H e r i t a b i l i t y  42  4.2  Juvenile-Mature Correlations  49  5.  CONCLUSION AND RECOMMENDATIONS  54  6.  REFERENCES  56  7.  APPENDIX  65  iii  LIST OF TABLES  Page 1.  Locations of Douglas-fir cone c o l l e c t i o n areas.  27  2.  Traits recorded on the 13-year-old Douglas-fir trees.  29  3.  Analysis of covariance and expected mean squares f o r height, dbh, RCD, volume, taper and crown width.  33  Analysis of variance and expected mean squares on a number of branches i n a whorl and an interwhorl, yearly growth and diameter of yearly growth.  34  Analysis of variance and expected mean squares on whorl and interwhorl branch lengths, diameters and angles.  36  Early, juvenile and 13-year-old tree t r a i t s used i n the juvenile-mature genetic correlations.  40  H e r i t a b i l i t y estimates and F ratios (for a l l sources of v a r i a t i o n ) from the analyses of covariance and variance for the 15 measured t r a i t s of the 13-year-old Douglas-fir trees.  43  8.  H e r i t a b i l i t y estimates and percentages of various variance components making up the t o t a l phenotypic v a r i a t i o n .  44  9.  H e r i t a b i l i t y estimates and F ratios (for each source of variation) from the analyses of variance of the 18 early and juvenile t r a i t s of the sample used i n this study.  50  4.  5.  6.  7.  IV  LIST OF FIGURES  Page  Figure 1.  D i s t r i b u t i o n of Douglas-fir cone c o l l e c t i o n areas 1962.  26  V  ACKNOWLEDGEMENTS I wish to extend my sincere appreciation to Dr. 0. S z i k l a i , my supervisor for his i n i t i a l d i r e c t i o n and assistance i n providing with invaluable information for this study.  me  I am also grateful to his  continuous encouragement throughout the period of my graduate studies and during the preparation of this thesis.  The funding received in  form of Assistantships and Fellowships, has made this unique experience possible and i s greatly appreciated. Special thanks to the other members of my committee Dr. A. Kozak, Faculty of Forestry and Dr. R.G. Science. ques and  Peterson, Department of Animal  They showed interest, contributed useful s t a t i s t i c a l technisuggestions and constantly reviewed the thesis.  I would l i k e to thank my colleagues Dr. Y.A. Dr. S. Omule and Mr.  El-Kassaby,  S. Chiyenda for the f r u i t f u l discussions we  during the course of writing this thesis.  had  Dr. Y.Z.G. Moyini, you  are  remembered for the i n i t i a l introduction to this University, continuous encouragement and valuable  advice.  F i n a l l y , gratitude i s expressed to my husband George, and  my  son, Stephen whose moral support and immense s a c r i f i c e while I studied in Canada, has made the completion of this thesis possible.  vi  DEDICATION  I dedicate this thesis to my late mother, Ephrance, whose love and dedication to motherhood w i l l always be cherished and remain a source of i n s p i r a t i o n .  1  1.  INTRODUCTION  Forest genetics i s a study of hereditary v a r i a t i o n i n forest trees (Wright, 1976).  These hereditary differences are caused by  genes and/or cytoplasm within the tree.  They are predetermined at the  time the ovule i s f e r t i l i z e d and i n that sense are opposed to d i f f e r ences which are caused by the external environment.  As the world's  population grows, the land available to forestry shrinks because of the a g r i c u l t u r a l demands, expansion of c i t i e s , and road development. Through implementing intensive forest management, forest genetics intends to improve the quality and amount of wood produced per unit area during the shortest possible time most  economically.  Progeny testing i s one of the e a r l i e s t methods developed and i s the main procedure for evaluating the genetic values of an i n d i v i d u a l (Shelbourne, 1969  and S z i k l a i , 1974)  i n forest tree breeding.  testing usually produces r e l i a b l e observations  Progeny  on y i e l d at approxi-  mately half the rotation age, while other c h a r a c t e r i s t i c s such as f r o s t , disease and insect resistance could be evaluated at e a r l i e r years.  For example, i n case of Coastal and I n t e r i o r Douglas-fir  [Pseudotsuga menziesii (Mirb.) Franco] (whose rotation ages are about 80 and 120 years respectively); i t may  take 40-60 years before y i e l d  could be evaluated with a high degree of certainty. period i n that case i s inconvenient, improvement programs.  The long testing  expensive and hampers tree  For this reason attempts are made by s c i e n t i s t s  to obtain v a l i d early testing methods which give a definite advantage in terms of e f f i c i e n c y , easiness and rapidity leading to greatly increased return on c a p i t a l investment (Wyk,  1976b).  2  To increase the rate of genetic gain; early selection i s needed in most tree breeding programs (Nanson, 1967, 1968, 1970).  Selection  therefore i s an important step i n tree improvement programs because this i s where the best provenances, families or individuals are chosen to establish base population for future work.  Measurements made i n  the nursery phases of provenance or progeny tests provide the f i r s t opportunity to estimate performance of the same provenances and progenies grown i n plantations  to various ages.  Thus, the concept of  juvenile-mature correlation becomes useful. The  juvenile-mature c o r r e l a t i o n concept was i n i t i a t e d by Schmidt  (1964) to select provenances and progenies mainly i n Scots pine (Pinus s i l v e s t r i s L.). correlations  He suggested that future research on juvenile-mature  should include  investigations  seed and physiology of the seedlings. v a r i a t i o n , mode of inheritance  related to chemistry of the  Prior knowledge of genetic  and h e r i t a b i l i t i e s of different t r a i t s  is recommended ( S z i k l a i , 1974) before their future performance can be assessed with a high degree of certainty.  Therefore, juvenile-mature  correlation expresses the relationship between q u a l i t a t i v e and quantitative  data collected at different intervals during the l i f e  cycle and depends on the strength of the genetic control. family correlations families, within  The between  indicate the r e l i a b i l i t y of early selection on  family correlations relate to early selection of  individuals within families, while t o t a l / o v e r a l l correlations  indicate  the r e l i a b i l i t y of early mass selection (Squillace and Gansel, 1974).  3  Besides height, diameter at breast height (dbh), volume, stem form, and root c o l l a r diameter,  crown width i s an important  i s t i c for selection i n progeny testing. and Wyk  character-  Denison (1967), Dyson (1969),  (1977) pointed out that trees with smaller crowns are  preferred i n selection as more trees of high wood production and higher timber quality would f i t on a unit area.  The crown i s made of  branches which for selection purposes must satisfy certain conditions. Large numbers and sizes of the branches are not desired as they leave a large proportion of knotwood i n logs which i n turn degrades the quality of sawn timber and pulpwood.  Branch c h a r a c t e r i s t i c s respond  less to genetic manipulation than stem form.  In spite of t h i s , there  i s a tendency to get generally f i n e r more horizontal branches from the improved trees as compared to the plantation stock. In l i g h t of the foregoing statements,  this study was  initiated  to attempt to meet the following objectives: 1.  To investigate v a r i a t i o n i n Douglas-fir populations and  parti-  tion i t into additive versus a l l the rest (the rest being non-additive genetic and environmental) using the following traits: (a)  height  (b)  diameter at breast height  (c)  root c o l l a r diameter  (d)  volume  (e)  taper  (f)  crown width  4  (g)  growing space  (h)  yearly growth  (i)  diameter of yearly growth  ( j ) . number of branches i n a whorl (k)  number of interwhorl branches  (1)  length of branches i n a whorl  (m)  diameter of branches i n a whorl  (n)  angle of branches i n a whorl  (o)  length of interwhorl branches, and  (p)  diameter of interwhorl branches  2.  to estimate h e r i t a b i l i t i e s of the above-mentioned  t r a i t s , and  3.  to study juvenile-mature genetic correlations i n r e l a t i o n to their possible use to reduce the progeny testing period by early selection with possible advantage of increasing selection d i f f e r e n t i a l and hence genetic gains.  5 2. 2.1  LITERATURE REVIEW  Variation Prior to selecting individuals for a tree breeding program, the  genetic v a r i a t i o n within the base population should be known. The existence of a wide range of v a r i a t i o n among the breeding individuals provides a basis for genetic manipulation. Amount of v a r i a t i o n i n a t r a i t i s measured and expressed as the variance.  2 The t o t a l v a r i a t i o n known as the phenotypic v a r i a t i o n (a.p.)  2 2 i s composed of genotypic (O"Q) and environmental variances (og) (Falconer, 1960).  The genotypic and environmental components  generally cannot be estimated d i r e c t l y from observations on the population, though Sakai and Hatakeyama (1963) estimated them.  They  obtained the values from the l i n e of best f i t between observed and expected plot means using least squares.  The most frequent estimates  of these components are obtained from experimental populations. The r e l a t i v e magnitude of these components determines  the genetic  properties of the population, i n particular the degree of resemblance between r e l a t i v e s . The genetic variance i s subdivided into additive, dominance and interaction variances (Falconer, 1960): 2 G (genetic)  2 A (additive)  2 D (dominance)  2 I (interaction)  The additive variance i s the variance of breeding values, the most important component since i t i s the chief cause of resemblance between r e l a t i v e s and therefore the chief determinant  of the observable  genetic properties of the population and the response of the  6  population to selection (Falconer, 1960).  I t i s the only component  that can be readily estimated from observations made on the population.  Therefore the important  p a r t i t i o n i s into additive genetic  variance versus a l l the rest (the rest being non-additive genetic and environmental variance). The interaction variance i s also subdivided according to whether i t involves breeding values or dominance deviations (Falconer, 1960). Considering two l o c i , there would be three sorts of two-factor actions.  inter-  Interaction between two breeding values gives r i s e to  o additive x additive variance (G^A); interaction between breeding value of one locus and dominance deviation of the other gives rise to . 2 additive x dominance variance (a^j)); and interaction between two dominance deviations gives r i s e to dominance x dominance variance 2 (OJJQ)  2  o-j- =  . Therefore the interaction variance i s expressed as 2 2 + O ^2J J + OJJJJ + etc, where the terms designated etc. are  similar to components a r i s i n g from interactions between more than two loci.  The amount of variance contributed by interactions i s usually  very small, especially involving large numbers of l o c i so that they are ignored i n most cases without 2 2 1960).  leading to serious errors (Falconer,  However, ajj and ax are c o l l e c t i v e l y known as non-additive  variances. Using observations of resemblance between r e l a t i v e s , the additive variance i s estimated, enabling the p a r t i t i o n of 2 °A  :  2 (D a  2 +  a  I  + a  2 E)«  If  t n e  inbred lines are available, the environ2 2  ment component can be estimated, providing a p a r t i t i o n of  OQIO^.  combination of the two partitions provides estimates of the three  A  7 d i f f e r e n t parts of the phenotypic  4 (phenotypic)  :  variance  (°f3 +  thus:  4)  (additive) (non-additive)  2 (environmental)  Environmental variance i s composed of a l l non-additive variance, and much of this i s beyond experimental fore, i t can mainly be estimated  control (Falconer, 1960).  There-  from highly inbred lines which  possess no genetic variance. Natural populations of trees contain large amounts of v a r i a b i l i t y f o r both quantitative and qualitative t r a i t s .  Generally,  genetic and environmental factors affect the quantitative t r a i t s , whereas, the qualitative ones are almost exclusively determined by genetic factors (El-Kassaby,  1980).  The extent of genetic v a r i a b i l i t y  in natural populations has been extensively studied with respect to quantitative t r a i t s using mainly quantitative genetic methods.  Some  of the work that has been done on qualitative and quantitative t r a i t s i s presented  2.1.1  i n the following sections.  Quantitative Genetic Methods Douglas-fir cones were analysed by W i l l e t t (1963) who found  their lengths ranging from 5.1 - 7.7 cm with an average of 6.0 cm. The widths ranged from 1.8 - 2.4 cm with an average of 2.1 cm.  Out of  the t o t a l v a r i a t i o n observed within and between provenances for cone length only 9.8% was attributed to longitude, l a t i t u d e , height, diameter at breast height (dbh), crown width, and age of parent trees. The same t r a i t s explained 13.2% of the t o t a l v a r i a t i o n i n cone width. These results postulated that other variables, genetical and  8  environmental; might account for the remaining part of v a r i a t i o n . Douglas-fir seed t r a i t s were studied and analysed (1963) and Dunlap (1964).  by Robinson  Their results i n conjunction with W i l l e t t ' s  (1963), confirmed Allen's (1960) findings that seed shape, colour  and  markings could d i f f e r e n t i a t e with certainty Coast and I n t e r i o r provenances. Kiss (1971) described Douglas-fir seeds and their germinants recording these t r a i t s , endosperm and embryo class; germination, dormancy and growing periods; and height, root c o l l a r diameter and number of branches of the 1+0  seedlings.  The analyses  s i g n i f i c a n t v a r i a t i o n within and between provenances.  (RCD)  revealed  The embryo  class alone accounted f o r 97% v a r i a t i o n i n the endosperm-embryo class, indicating that the embryo can be used alone i n seed c l a s s i f i c a t i o n . Latitude and elevation accounted for the largest part of v a r i a t i o n i n height, while longitude, l a t i t u d e and germination period explained large part of v a r i a t i o n i n RCD. number of branches was period.  a  The largest amount of v a r i a t i o n i n  attributed to longitude, latitude and dormancy  It i s worth mentioning at this point that the  foregoing  studies of Robinson (1963), W i l l e t t (1963), Dunlap (1964) and Kiss (1971) a l l dealt with e a r l i e r materials which gave r i s e to trees that provided  data for this study.  The acids (DNA)  study of i n t r a s p e c i f i c v a r i a t i o n i n deoxyribose nucleic contents of Douglas-fir by El-Lakany and S z i k l a i (1973)  revealed intraphase nuclear volume (INV) and r e l a t i v e amounts of to be correlated with l a t i t u d e of seed source. i n INV and DNA  The  DNA  trend of v a r i a t i o n  content appeared to be c l i n a l with an increase from  9  South to North along the range of species. were found to have higher amounts of DNA Whiteside fir.  The coastal provenances  than the i n t e r i o r ones.  et. ail. (1977) measured timber  s t i f f n e s s i n Douglas-  They found 80% of the v a r i a t i o n to be due to branch size and  wood density. Most of the genetic variation i n height growth of juvenile Douglas-fir trees i n one of the International Union of Forest Research Organization (IUFRO) provenance-progeny tests was  reported by Fashler  (1979) to be attributed to within provenance effects for two seed zones.  However, the trend i n the other two seed cones was  the opposite d i r e c t i o n .  The apparent contradiction may  observed i n  be explained  by different adaptation responses of the different provenances to the progeny test s i t e .  She also obtained s i g n i f i c a n t (0.01  probability  level) additive variances for a l l the years from age two to eight. There was a s l i g h t drop i n the proportion contributed by the additive variance towards the phenotypic  variance as the seedlings advanced i n  age.  2.1.2  Q u a l i t a t i v e Genetic Methods Qualitative genetic methods are made possible through use of  starch gel electrophoresis by which genetic heterogeneity of proteins and isozymes can e a s i l y be detected.  Qualitative analysis provides  information on the d i s t r i b u t i o n of a l l e l i c v a r i a t i o n i n natural populations since the enzymes are composed of polypeptides synthesized by the action of one or more structural genes.  and The  10  electrophoretic v a r i a t i o n of enzymes can be d i r e c t l y related to changes i n gene structure or codon sequence and always follow Mendelian segregation  i n ideal populations  (Yang et: a l . , 1977).  Genetic v a r i a b i l i t y i n natural populations  of Douglas-fir  studied at the enzymatic l e v e l by Yang e_t a l . (1977).  was  He observed  s i g n i f i c a n t a l l e l i c frequency differences among the provenances examined and the genetic differences i n terms of genetic i d e n t i t y and distance between the provenances was more or less similar to the geographic distance.  Heterogeneity i n general was  with increase i n a l t i t u d e and,  found to decrease  to a lesser extent, l a t i t u d e .  Several species showed close agreement between their electrophoretic and quantitative data i n p a r t i t i o n i n g the t o t a l v a r i a t i o n between and within population  level.  Among those i s Douglas-fir,  Yeh and O'Malley (1980) found 3% of genie v a r i a t i o n i n Coastal populations  to be due to between population gene differences.  They  also applied the analysis of gene d i v e r s i t y to the work of Yang eit a l . (1977) and  the results were s t r i k i n g l y similar.  El-Kassaby (1980)  investigated genetic v a r i a t i o n at 27 allozyme l o c i and seven d i f f e r e n t seedling t r a i t s i n a Coastal stand of Douglas-fir with respect to different elevational classes. and needles, length of hypocotyl and root dry weight.  The  t r a i t s were number of  cotyledons  and e p i c o t y l , total height and,  shoot  He found 7% of t o t a l genetic v a r i a t i o n  attributed to differences between while 93% was within elevational classes.  due  to differences  11  In the International Union of Forest Organization (IUFRO) provenances of Sitka spruce  [Picea sitchensis (Bong.) Carr],  Illingworth (1978) reported that 11% of t o t a l v a r i a t i o n for 3-year height growth of seedlings, was tions.  due to differences between the  popula-  Yeh and El-Kassaby (1980) used the same provenances and found  that only 8% of t o t a l genetic v a r i a t i o n was  also explained by d i f f e r -  ences between the populations. In lodgepole pine (Pinus contorta spp l a t i f o l i a ) Yeh and Layton (1979) pointed out that 4% and 96% of the t o t a l genetic v a r i a t i o n  was  due to interpopulation and intra-population gene differences respectively.  The high l e v e l of within population v a r i a t i o n was consistent  with the observations on general physiological functions (Perry and Lotan, 1977).  O'Malley et a l . (1979) estimated 12% of the detected  genetic v a r i a t i o n i n ponderosa pine (Pinus ponderosa) to be attributed to differences between the  stands.  The above studies indicate possible use of electrophoresis techniques  to supplement t r a d i t i o n a l studies i n assessing the amount  and extent of genetic v a r i a b i l i t y i n forest tree species.  2.2  Heritability Individual genes cannot o r d i n a r i l y be i d e n t i f i e d in quantitative  inheritance and therefore studies of quantitative t r a i t s focus on the study of phenotypic  2 variance (°"p).  2 A proportion of the observed °p, 2  which i s due to the genetic variance (°Q) i s known as h e r i t a b i l i t y (h ). 2  H e r i t a b i l i t y then determines the degree of resemblance  j  12  between r e l a t i v e s .  Estimates of h2  s  a r e  necessary to express the  r e l i a b i l i t y with which the phenotypic t r a i t s might be expected to appear i n the next generation. Warner (1952), Lerner (1958), Falconer (1960) and Hattemer (1963) l i s t various ways of estimating h e r i t a b i l i t i e s .  F i r s t l y , by  the use of offspring-parent regression.  Secondly, by the use of  correlation between f u l l and h a l f - s i b s .  Thirdly, by using approxima-  tion of non-heritable variance from genetically uniform populations. Fourthly, by comparing phenotypic t r a i t s displayed by monozygotic as against dizygotic twins. which estimates h  2  And l a s t l y by the use of clonal analysis  i n the broad  sense.  Besides the above methods of estimating h e r i t a b i l i t y ,  Sakai and  Hatakeyama (1963) estimated h e r i t a b i l i t i e s i n Populus euramericana Abies sachalinensis without raising progenies.  and  They used  Shirikhande's (1957) method which i s based on the assumption  that  v a r i a t i o n between plot means consists of one N-th of the genetic variance and one N^-th of the environmental variance.  The number of  individuals i n each plot i s represented by N while b i s a constant depending on the v a r i a t i o n pattern of environmental conditions. expression then i s :  where  a„  = plot variance  ?  Og = genotypic variance a  2  = environmental variance  b = constant whose value i s between zero and  one.  The  13  From these components of variance values of a number of plots, they 2 obtained genotypic (^Go^ 2 and O g  Q  referred  * environmental  2  variances.  The  to the values of best f i t between observed  and  anc  (cr  E o  )  expected plot means using least squares. Using this formula, h 2 G° , estimates of h s were then obtained. The method i s  =  2  Q  2  Go Eo recommended for a population with trees of about the same age,  a  + 0  fairly  uniform spacing and without any damage or heavy thinning. There are two types of h e r i t a b i l i t i e s , the narrow sense and the broad sense (h^b) (Lush, 1949; Legates, 1979).  The h n 2  (h n) 2  Toda, 1958; Warwick and  i s the proportion of the  observed  phenotypic variance which i s additively genetic or which i s associated with differences  i n average breeding values.  H e r i t a b i l i t y i n the  broad sense i s considered as the sum of the additively genetic, the dominance and the interaction (known as epistasis) variances expressed as a proportion of the phenotypic variance. types of h2  s  Expressions of the two  are as f o i l ows: 2 °"A 2 °P  ,2 h n =  hb 2  = !| ? °P  Q  =  2 OA 2^2 A E +  °A 2 °A  CT  +  g  D  °\  +  2 +  °D  +  2  2 °I  +  "E  The higher the h e r i t a b i l i t y value, the stronger the t r a i t expression is controlled by the genetic make up of the i n d i v i d u a l . contrary, a small h contributing  2  On the  value would mean that the environment i s  more to the v a r i a t i o n r e l a t i v e to the genetic factors.  Hence depending on the h e r i t a b i l i t y value, a breeder should be able to t e l l the major source of v a r i a t i o n for the t r a i t i n question.  14  The most important use of h" i s i n predicting the amount of genetic improvement (gains) that might be attained under various breeding Schemes (Squillace e^t a i . , 1966).  Prior knowledge of  h e r i t a b i l i t y a s s i s t s i n selecting the best breeding approach, suggests the amount of money that can j u s t i f i a b l y be spent, indicates the amount of e f f o r t to put on a t r a i t which i s to be improved, indicates the number of trees to be selected and progeny tested, and also indicates the i n t e n s i t y of each phase of the breeding program. genetic gain (AG)  i s given by the following formula AG = i h  is the selection d i f f e r e n t i a l .  2  The  where i  This would be the difference between  the population mean before selection and that of the selected population  (Falconer, 1960;  Wright, 1976).  Some of the work that has been done on h trees i s presented by Birot (1976). h a l f - s i b s from which the following h for cotyledon number, 0.32 0.60  2  estimates i n forest  He worked on IUFRO Douglas-fir 2  estimates were reported;  for growth cessation, 0.84  for height i n year one and 0.46  0.52  for flushing,  for height i n the second year.  The decrease i n height h e r i t a b i l i t y was  attributed to the possible  disappearance of maternal effects or may  be related to increase of  competition e f f e c t s . Fashler (1979) estimated height h e r i t a b i l i t i e s of juvenile Douglas-fir trees i n one of IUFRO provenance-progeny tests. ranged from 0.28 h  2  - 0.52  with an average of 0.38.  The  The  progenies.  2  r e l a t i v e l y high  values indicated opportunities for s i g n i f i c a n t improvement by  selection i n Douglas-fir provenances and  hs  15  The trees that provided data for this study were worked on at the age of one year by Kiss (1971). 0.13  for endosperm embryo class, 0.08  the growing season, 0.14 and 0.15  He estimated h e r i t a b i l i t i e s of for dormancy period, 0.11 f o r  for height, 0.17  for the number of  for root c o l l a r  diameter,  branches.  H e r i t a b i l i t i e s , however, have their own l i m i t a t i o n s , i n that they vary within a species and between locations. uniform the environment, the higher the h whenever a h  2  values.  2  In general the more Therefore,  value i s given, i t must refer to a particular  population under particular conditions (Falconer, 1960; Warwick and Legates, 1979).  In addition h  2  Zobel,  1961;  i s a population concept  which measures the genetic v a r i a t i o n within a population, but not the contributions of the genotype and the environment to the phenotype of the individual (Suzuki and G r i f f i t h s , 1976).  H e r i t a b i l i t i e s are  subject to large standard errors as a result of very large standard errors of the variances (Falkenhagen,  1972).  In an attempt to improve  the accuracy of the variances, the number of observations can easily become impractical.  2.3  Juvenile-Mature Correlations The conventional methods of progeny testing i n tree breeding  l a s t s for long periods of time.  For example, Coastal and Interior  Douglas-fir have rotation ages of 80 and 120 years respectively and their progeny test periods usually l a s t for 40 and 60 years respectively.  Since man  i s impatient, seeks to accomplish i n a few years  16 what nature may be content to wrestle with for centuries, Schmidt (1964) i n i t i a t e d the juvenile-mature correlation concept.  This i s a  measure of the association between the juvenile and mature t r a i t s .  If  they are closely related, the performance of the mature t r a i t can be accurately predicted from the juvenile t r a i t .  However, i t i s  important to r e s i s t the temptation of overestimating the value of early tests. There are three types of correlations, environmental ( r ^ ) , genetic (r^) and phenotypic ( r ) . p  The r  £  i s the one between the  environmental deviations of the two t r a i t s while the r ^ i s the one between the genotypic values.  In p r i n c i p l e i t i s not easy to p a r t i -  tion variance into non-additive components and therefore s t r i c t l y speaking the r  £  consists of correlation of environmental deviations  together with non-additive genetic deviations (Falconer, 1960). r ^ i s the correlation of additive genetic deviations.  A combination  of r ^ and r^ gives r i s e to the observable phenotypic correlation (r ). p where  The relationship i s expressed a s r = h h r , + e e r p x y A x y E x = the juvenile t r a i t y = the mature t r a i t h = the square root of the h e r i t a b i l i t y  If both t r a i t s have low h2  s  then the phenotypic correlation i s  determined mainly by the r ^ . In case of high h s the r ^ i s the z  most important.  Hence  17  Genetic and environmental  source of variation affect the t r a i t  through different physiological mechanisms thus causing the two correlations to be different i n magnitude and sometimes d i f f e r e n t i n sign.  Hence the magnitude and sign of the r ^ cannot be determined  from the phenotypic correlation alone.  Genetic correlations can be  estimated i n 3 different ways (Falconer, 1960).  F i r s t l y , by use of  resemblance between relatives thus: Cov 5Z 2 y  r. =  where  Cov = covariance component 2 a  = variance component  Secondly, by the use of offspring-parent relationship thus:  =  r  A  Cov Cov Cov xx xy yy  V  where  Cov xy Cov  xx  and Cov  yy  = Cross-covariance component r  = the offspring-parent covariances of each p o r t r a i t separately.  And t h i r d l y by using response  V  CR CR r.A = \ R x where  CR  to selection thus:  CR R  = correlated response (one i n which selection i s primarily for one t r a i t , but due to a strong genetic correlation a change occurs i n a second  R  trait).  = response of a t r a i t when selected d i r e c t l y .  18  The basic problem i n tree breeding i s to determine  the extent to  which early selection i s e f f e c t i v e at the u t i l i z a t i o n stage and when i t must be carried out to achieve s u f f i c i e n t correlated gains (Nanson, 1976).  The existing experimental plots established i n the past which  are almost a unique source of information on juvenile-mature r e l a t i o n ship have been concerned primarily with provenances and have lacked proper r e p l i c a t i o n and randomization.  Therefore, correlation  estimates were based on means of individuals of one plot resulting i n high environmental effects and overestimation of genetic gains. However, experimental plots established more recently have the advantage (over old ones) that they incorporate s u f f i c i e n t  replication  and randomization, resulting i n less environmental effects and better estimate genetic gains. Most of the juvenile-mature correlation work i n tree breeding has been primarily concerned with phenotypic correlations.  The work  that has been done i n Douglas-fir (species) includes that of S z i k l a i (1964).  Heights of 132-day-old  f u l l - s i b progenies were correlated  with their heights at 4 years of age obtaining a very high correlation of 0.89.  Results indicated that performance of these trees at age 4  could have been predicted at the age of 132  days.  Kiss (1971) correlated a number of variables with one year old seedling t r a i t s .  The variables were those related to location of  parent trees, parent tree, cone, seed and germinant's  traits.  Seedling t r a i t s included height, root c o l l a r diameter and number of branches.. Strong correlations were observed between seed weight,  19  endosperm-embryo class and seedling height.  Heavier seeds have been  observed to give r i s e to faster growing seedlings which do not necessarily become potential winners (Sluder, 1979).  Therefore,  selection i s not recommended to be based on seed weight as i t might result i n seedlings with a survival and growth disadvantage. Height correlations for the ages 5, 10, 12, 15, 18, 23, 28, 40 and 53 years were estimated Douglas-fir plantations.  (Namkoong et a l . , 1972)  33,  i n one of the  The correlations declined for any given  when correlated with successively older ages.  age  The results pointed to  the p o s s i b i l i t y of selecting at age 28 for height performance at age 53 (r = 0.82).  Since age 53 i s i n the range of half the rotation age  (40-60 years), then selection based on height performance at age  28  might be a good predictor of height performance at the end of the rotation  age.  Haddock et a l . (1976) obtained very high correlations (close to one) between average t o t a l heights of 2-year-old heights at 5, 6, 7, 8 and 11 years of age.  seedlings and  their  Hence height performance  at a l l these ages could have been predicted at age two.  However,  caution should be taken since i t i s well known that during early stages, the genetic t r a i t s change quite often (Namkoong e_t a l . , 1972; Squillace and Gansel, 1974;  S z i k l a i , 1974).  Autocorrelations of heights at the ages of 4, 5, 6 and 7 were estimated  i n one of IUFRO's Douglas-fir provenance-progeny test by  Fashler (1979).  The trend of the results was  Namkoong e_t a l . (1972).  similar to that of  Continuation of the research was  recommended  20  to,determine whether these high correlations would persist i n later ages.  Persistance of the high correlations would indicate preference  of early selection as a predictor of late performance with minimal r i s k s of losing potential  winners.  Similar type of work (juvenile-mature phenotypic correlations) has been done i n Pinaceae  (Wakeley, 1971).  His study involved slash  pine (Pinus e l l i o t t i i , Engelm), l o b l o l l y pine (Pinus taeda L.), longleaf pine (Pinus echinata M i l l ) .  The correlations were of t o t a l  heights (at 3, 4, 5, 8, 10, 15 and 20 years) and diameter at breast height (at 10, 15 and 20 years) with corresponding measurements at age 30.  The correlation c o e f f i c i e n t s obtained were r e l a t i v e l y higher for  diameters compared to heights.  They were increasing with age,  similar  to Namkoong et a l ' s (1972) and Fashler's (1979) r e s u l t s . Squillace and Gansel (1974) correlated growth t r a i t s (height, diameter at breast height, and volume) and oleoresin y i e l d recorded measurements of slash pine at age 25 with those of e a r l i e r years 3, 8, 14 and 18.  Growth t r a i t correlations increased with age while those  of oleoresin y i e l d were moderate (r = 0.61).  Depending on the age of  selection, the highest genetic height gains were realised around 9 and .10 years which led to the conclusion that less than mature trees can be selected for mature performance.  Therefore, a breeder i s advised  to consider the additional cost of short generation intervals when making a f i n a l decision. However, the extra developmental be outweighed by the increased genetic gains per year.  costs could  21  Juvenile-mature  correlation c o e f f i c i e n t s were presented  relating  heights and diameters of ponderosa pine (Pinus ponderosa Laos.) and western white pine (Pinus monticola Dougl.) of early ages to heights and diameters at ages ranging up 50 years (Steinhoff, 1974).  It i s  recommended that i n both species evaluation of provenance t r i a l s or progeny tests would not be very r e l i a b l e at ages less than 15-20  years  though limited selection for c u l l i n g the poorest provenances or families could begin at about 10 years of age. Mean heights of slash pine and l o b l o l l y pine progenies i n the nursery were correlated with mean heights of the same families i n 5 year old plantations (La Farge, 1975).  Only 12% of the correlations  were s i g n i f i c a n t indicating that future performance should not be predicted at the nursery stage.  Therefore, f i e l d plantations should  be established regardless of nursery performance. Nanson (1976), working with Scots pine (Pinus s i l v e s t r i s ) , Norway spruce (Picea abies) and Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco], found that heights i n the nursery, but more importantly i n the f i e l d between ages 5 and 10 were good indicators of wood production at the end of the rotation age. growth was  then suggested  Selection for height  to be suitable between ages 5 and 10 while  10 and 20 could be optimal for form and branching c h a r a c t e r i s t i c s , and perhaps s t i l l l a t e r for disease peculiar to the mature ages. Meier and Goggans (1977a,b) observed low s i g n i f i c a n t correlations between the c o r t i c a l monoterpenes of the eight year old V i r g i n i a pine (Pinus virginiana M i l l . ) and commercially  important  characteristics  22  namely height and diameter.  Therefore, monoterpenes would probably  not be valuable as an indirect selection t o o l . Besides juvenile-mature correlation work i n Douglas-fir and Pinaceae, other species have also been dealt with.  Wyk  (1976a,b)  examined a complete d i a l l e d cross of Eucalyptus grandis.  He corre-  lated data r e l a t i n g to t o t a l height, root c o l l a r diameter, dbh, volume seedling dry weight, rate of growth, and number, length and of branches.  diameter  The results revealed strong relationships between these  t r a i t s , but greenhouse results showed a generally poor relationship with nursery r e s u l t s .  However, the data revealed maternal effects to  be s i g n i f i c a n t at two months and seemed to disappear i n older trees indicating possible effects of embryo-endosperm v i a b i l i t y . effects are most often considered of l i t t l e importance improvement programs.  Maternal  i n tree  Barnes (1973) has confirmed this by showing  that these effects are negligible i n most t r a i t s studied i n Pinus patula especially i n long term studies. Height, diameter and volume of the 8-year-old clones of eastern cottonwood (Populus deltoides Bartr.) were correlated with those of the e a r l i e r ages (Randall, 1977).  Phenotypic and genotypic correla-  tions obtained increased with increase i n age. Nepevue et a l . (1978) correlated wood density of one year old stems of Populus nigra and Populus euramericana  clones.  The  correla-  tions were high indicating that early selection of clones with high wood density i s possible without endangering  yield.  However, no  s i g n i f i c a n t correlation was observed between density and g i r t h  23  increment.  Similar results had been reported by Reck and S z i k l a i  (1970) from their studies of wood quality (ring width, wood density and dry matter content) i n Douglas-fir. Results implied that future wood quality can be assessed using wood samples from increment with the exception of the latest five rings.  cores  Therefore, progeny tests  for estimating these genetic parameters should commence with trees old enough (at least 8 years) to provide wood samples with more than five annual rings. Ring width and wood density i n a 30-year-old stand of Norway spruce was correlated with the juvenile measurements.  The correla-  tions revealed the possible p r e d i c t a b i l i t y of mature wood density from early juvenile stages. Work supporting the above-mentioned studies of the f e a s i b i l i t y of early selection of superior progenies i s presented by Ying and Morgenstern  (1979).  They obtained high height correlations at ages 8  and 22 in white spruce [Picea glauca (Moench) Voss]. Lambeth (1980) reviewed and analysed juvenile-mature phenotypic correlations from the l i t e r a t u r e and found i t more predictable than what could have appeared  to be the case i n the f i r s t place.  Age -  age correlations were estimated with reasonable accuracy, with the exception of very young ages (1-3 years), by a single regression equation which applied to several species and studies.  Optimum  selection ages recommended were 5, 6, 6, 7, 7, 8 and 8 years for 20, 25, 30, 35, 40, 45 and 50 years of economic rotations in that order. Besides work done on juvenile-mature correlations i n trees, similar studies have been attempted  i n other f i e l d s .  Only to mention  24  a few, poultry breeders  found a correlation between the weight of the  egg and that of the chick at l a t e r ages.  Funk et_ al^. (1930) indicated  that chicks were very highly variable individuals while Jaap and Morris (1937) pointed out that growth to 8 weeks of age appeared to be separate and not necessarily related to adult weight. Human g e n e t i c i s t s have also found that trends evident i n early stages may not be correlated to those of other ages.  Bock et a l .  (1973) developed a model to describe human growth that includes two l o g i s t i c functions, the f i r s t accounting tal  for a component for prepuber-  growth, which continues i n reduced degree u n t i l maturity,  the second accounts for the contribution of the adolescent  while  spurt.  Considering Schmidt's (1927, 1930, 1935, 1936 and 1964) suggestion cited e a r l i e r , and S z i k l a i ' s (1974) observation concerning  prior  knowledge of the pattern of genetic v a r i a t i o n , mode of inheritance, and h e r i t a b i l i t i e s of the different t r a i t s , future performance can be assessed with a high degree of certainty.  Repetitive experiments  using various species taking into account a number of economically important  t r a i t s w i l l make i t possible to draw v a l i d conclusions f o r  each species.  Use of juvenile-mature  correlations w i l l lead to short  generation i n t e r v a l permitting the testing of more families and more individuals which would i n turn permit greater selection intensity and r e a l i z a t i o n of greater genetic gains (Nanson, 1970).  25  3.  MATERIALS AND METHODS  The trees that provided data for this study represent open pollinated progenies of 7 Douglas-fir provenances from B r i t i s h Columbia.  Details of location of these provenances are presented i n  Figure 1 and Table 1.  H i s t o r i c a l information on these trees i s given  by Kiss (1971). O r i g i n a l l y there were 29 Douglas-fir provenances from B r i t i s h Columbia and one i n Alberta from which dominant trees were selected. The cones and seeds from these trees were studied and the information is available from Robinson (1963), W i l l e t t (1963) and Dunlap (1964). In 1966, seedlings from 21 provenances consisting of five dominant trees each, were raised in the nursery at the University of B r i t i s h Columbia.  A l l the information concerning provenance location, plus  tree, cone, seed, germinant and seedling t r a i t s was compiled by Kiss (1971). Seedlings were then transplanted into large containers of about 210 cirp i n volume.  The containers are known as "French J i f f y Pots"  made of compressed peat.  In 1967, they were planted out i n three  d i f f e r e n t permanent locations, as 1+0 seedlings with the containers i n situ.  The three locations are Lens Creek on Vancouver Island, the  University Research Forest i n Haney, and at the South Campus Research nursery of the University of B r i t i s h Columbia (UBC).  The layout of  provenances planted on South Campus at UBC i s shown i n Appendix 1, and samples for this study were obtained from this location.  The  DISTRIOUTION OF DOUGLAS -FIR  COME C0L1ECTI0N  A R E A S , 1962 1 -  ;  A2  • 16 Delia Coola  Nimpkish Lake  17- Honey (a)  Gold River  10 Honey (a,)  3  Elk River  4  Robertson Creek  56  Gordon River Cowlchan Lake  20 Honey (b ) 21 Honey (cl  78  Ash River  22 Honey (c ) 23-Borden Creek  9  Courlenay (high elev)  A10  il-  ia  A 1 9 - Honey (b) (  (  Gorrel Lake  24- Quesnel  Courtenay (low elev)  A 2 5 Hopo (high «lav )  Qulnsam Soanlchlon  26 Hope (low elev) A 27- Clinton 28 Keremeos  13- Ml- Prevosl • 14 HIM 6 0  29 Lardoau  15- Errlnglon 20  0  Scale of Miles 60 20 40  80  100  • Provenances used i n this study  FIGURE 1.  So  ro ON  27  TABLE 1. Locations of Douglas - f i r  Code  Provenance area  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 24b 25 26 27 28 29 30  Nimpkish Lake Gold River* Elk River Robertson Gordon R i v e r Cowichan Lake Ash River Garret Lake Courtney (high elev.) Courtney (low elev.)* Quinsam Saanichton Mt. Prevost Hill-60* Errington Bella Coola* U.B.C. Forest A° U.B.C. Forest Check A° U.B.C. Forest B* U.B.C. Forest Check BA U.B.C. Forest C U.B.C. Forest Check C° Borden Creek Quesnel Area A Quesnel Area 2° Hope (high elev.)* Hope (low elev.)A Clinton* Keremeos Lardeau Kananaskis 0  0  0  Longitude degrees  127.0 126.1 125.9 124.2 124.5 124.5 125.1 125.6 125.2 125.3 125.3 123.4 123.7 123.4 124.5 127.0 122.6 122.6 122.6 122.6 122.6 122.6 121.7 122.7 122.7 121.4 121.4 121.3 120.3 117.0 115.0  cone c o l l e c t i o n areas  Latitude degrees  Elevation degrees  50.4 49.9 49.9 48.7 48.9 48.8 49.5 50.1 49.6 50.0 50.0 48.5 48.9 49.4 49.4 52.2 49.3 49.3 49.3 49.3 49.3 49.3 49.0 53.3 53.3 49.4 49.4 51.2 49.4 50.2 51.0  180 900 1000  1750 1150 1000 2050 400 300 250 1400 700 450 20  1600 1300 650  800 2000  750 250 3200  1800 5050  Collection agency  Can.For. Pro. Tahsis Co. B.C.F.S. B.C.F.S. B.C.F.S. B.C.F.P. MB & PR Co. B.C.F.S. Crown Z e l l . Crown Z e l l . B.C.F.S. B.C.F.S. B.C.F.S. B.C.F.S. U.B.C. B.C.F.S. U.B.C. U.B.C. U.B.C. U.B.C. U.B.C. U.B.C. B.C.F.S. U.B.C. U.B.C. U.B.C. U.B.C. U.B.C. U.B.C. Koot. F.P. Kan. F.E.S.  * Provenances selected for this study. A Provenances that did not provide enough seed to r e p l i c a t e the experiment. ° Provenances that did not provide enough seed to warrant their incorporation i n the experiment.  28  plantation s i t e at South Campus of UBC i s of about s i t e index one, with 1550 mm average r a i n f a l l , minimum and maximum daily temperatures of -12° and 38°C respectively. The experiment at South Campus Research Nursery of UBC was l a i d down i n a randomized complete block design with five treatments and eight r e p l i c a t i o n s for each treatment.  The blocks refer to  provenances, while the treatments refer to dominant trees from each provenance.  Replications represent  seedlings from each dominant tree.  In this study seedlings from a dominant tree are called a family and the i n d i v i d u a l seedlings are referred to as trees. At the time of data c o l l e c t i o n for this study (May 1980), trees were 14-year-old with a survival percentage of 41.25.  From the  remaining provenances, seven (2, 10, 14, 16, 19, 25 and 27) could provide at least four families each and three trees from each family and therefore, were selected.  The families and the trees within each  family were randomly selected after eliminating trees that were dying, stunted, crooked and forked. of 84 trees.  Hence observations  A summary of the t r a i t s recorded  presented i n Table 2.  were taken on a total  on each tree i s  It i s worth noting that, a l l measurements did  not include the growth of 1980 (the 14th year).  Therefore  throughout  the discussion of this study, trees w i l l be referred to as "13-year-old trees". Using Huber's formula (Husch et^ a l . , 1972), i n d i v i d u a l measurements of yearly growth and diameter of yearly growth, the volume was computed section by section and then summed to obtain volume for a whole tree.  Huber's formula states that V = hA where m  29  TABLE 2.  No.  Traits recorded on the 13-year-old Douglas-fir trees  Trait  Unit  Remarks  1  Total height  m  up to 1979's growth  2  DBH  cm  at 1.3 m high  3  Root c o l l a r diameter  cm  at ground level  4  Volume  m^  -  5  Taper  -  -  6  Crown width  m  7  Growing space  rn  -  8  Yearly growth  m  -  9  Yearly growth diameter  cm  -  10  No. of branches i n a whorl  -  -  11  Length of branches i n a whorl  m  of the three  12  Diameter of branches i n a whorl  cm  longest branches  13  Angle of branches i n a whorl  14  No. of interwhorl branches  -  15  Length of interwhorl branches  m  16  Diameter of interwhorl branches  cm  2  degrees  i n each whorl  1  of the three longest branches in each interwhorl  30  V  = volume  h  = height or length of the section  ^  = middle cross-sectional area of the section.  The taper was calculated as an index using the following formula (Kozak,  1980) _ dbh - l a s t diameter Taper = = : v  length  where last diameter refers to mid diameter of the youngest whorl and length being the distance between the two diameters. Crown width was measured as the sum of the lengths of the two longest branches on opposite sides (Denison, 1967; Dyson, 1969; 1977).  Wyk,  In addition, growing space available for each tree was com-  puted because of the non-uniform tree spacing (Appendix 1). The branch lengths, diameters and angles measured were to the three longest branches i n each whorl and interwhorl. diameters were taken very close to the main stem.  restricted The  The angles were  recorded only on the whorl branches since the interwhorls do not l a s t long before they are shed. 3.1  Variation The data was analysed using the ANOVAR, a computer  package  available from the University of B r i t i s h Columbia Computing Center. ANOVAR performs analysis of variance and covariance for a wide variety of problems both with equal and unequal number of observations.  In  this study i t was used to perform both analysis of variance (ANOVA) and covariance (ANACOVA) for a 2-way nested design.  31  To begin with, the provenances were subdivided into four groups according to their locations. Provenance 2 and 16 represented the Wet Coastal Region, while 10 and 14 represented and 25 representing the Wet Mainland Interior Mainland.  the Dry Coastal Region; 19  and f i n a l l y , 27 representing the  Analysis of variance carried out on the f i r s t  three locations, (since they were balanced) revealed non-significant differences among locations. Therefore, they were a l l pooled  together  and analysed as the seven provenances. Due to the large number of t r a i t s (Table 2) involved i n this study and different sources of v a r i a t i o n , the analyses were done i n three subsets.  The f i r s t subset involved ANACOVA on height, diameter  at breast height (dbh), root c o l l a r diameter (RCD), volume, taper and crown width using growing space as a covariate.  The ANACOVA table and  the expected mean squares (EMS) based on the following random effects l i n e a r model: Y. .. = u + x. + B ijk  i  j(i)  +  e  k(xj)  where Y. ., = measurement of the k the i  t  n  t n  tree i n the j  t  n  family of  provenance  y = overall mean = effect due to provenances B.,.. = e f f e c t due to families nested within provenances £ , k(ij)  = residual v a r i a t i o n (0, a„) t  i = number of provenances i = 1  ,p  j = number of families nested within provenances j = l ,  ,f  32 k = number of trees nested within families and provenances k.= 1, are  ,t  presented i n Table 3. The second subset involved ANOVA on number of branches i n a whorl  and an interwhorl, yearly growth and diameter of yearly growth.  The  ANOVA table and EMS based on the following random effects l i n e a r model: Y.  = y + T.+  ijkl  i  M 1  B .,.  s  + y, ,,,, r  k(ij)  +  e ..,  l(ijk) w  s  where Y. ., .  = measurement of the 1th whorl on the k tree of the j family i n the i provenance = effect due to provenances = effect due to families nested within provenance t n  t  T f3j(i)  n  t  n  = effect due to trees nested within family within provenance  Yjx^.s  2 £  l(ijk)  =  residual v a r i a t i o n (0, a^)  i = number of provenances i = 1,  ,p  j = number of familes nested within provenances j = 1,  ,f  k = number of trees nested within families and provenances k = 1, t 1 = number of whorls nested within trees, families and provenances 1 = 1,.... ,w are  presented i n Table 4.  The four uppermost whorls (except for that  of 1979) uniform for a l l trees, were selected for this study.  A l l of  these whorls were s t i l l young, had a l l the interwhorl branches i n t a c t , and they were counted despite the fact that some would drop off soon afterwards.  33  TABLE 3.  Analysis of covariance and expected mean squares (EMS) for height, dbh, RCD, volume, taper and crown width  df  Source of v a r i a t i o n  EMS  MS  Provenance  p l  MS,  a , + to , + ftoE F p  Families within provenance  p(f-l)  MS  °E  Residual  pf(t-l)  MS  -  Total  where:  II  2  2  +  to  l  III  pft-1  Op  = variance among provenances MS - MS I];  ft  a  2 F  = variance among families nested within provenances M S  2  II "  M S  III  = residual component of variation "  M S  III  2  34  TABLE 4.  Analysis of variance and expected mean square (EMS) on a number of branches i n a whorl and an interwhorl, yearly growth and diameter of yearly growth  Source of v a r i a t i o n  df  MS  EMS  Provenance  p-1  Families within provenance  p(f-l)  MS  Trees with family, within provenance  pf(t-l)  M  Residual  pft(w-l)  MS  Total  pftw-1  0 £ + wo\£ + twcr£ + f twcK L a p  S  n  m  a2 +  °E  +  w a  w a  2  + twa2  T  IV  2 where:  a  = variance among provenances MSj. - MS-J--J. ftw  2 a-p = variance among families nested within provenances II - III tw M S  M S  =  2 o  T  = variance among trees nested within families within provenances M  S  =  i n  -  M S  iv  w  a  = residual component of v a r i a t i o n  =  M S  iv  35  The third subset of analyses consisted of ANOVA on whorl and interwhorl branch lengths, diameters and angles.  The ANOVA table and  EMS based on the following mixed effects linear model: ^ijklm  y  +  T  i  +  B  j(i)  +  Y  k(ij)  +  a  l(ijk)  +  E  m(ijkl)  where Y. ., •,  = measurement of the m branch of the l whorl on the k tree of the j family i n the i provenance. tn  t  t n  t  t  n  n  n  T.  = effect due to provenance  3^  = effect due to families nested within provenances  k(ij)  =  e  ff due trees nested within families within provenances  l(ijk)  =  e  ff due whorls nested within trees, families and provenances  m(ijkl)  =  Y  a  e  c  t  e c t  t o  t 0  2 residual v a r i a t i o n (0, o^,) s  £  -  i = number of provenances i = 1,  ,p  j = number of families nested within provenances 3 = 1 f k = number of trees nested within families and provenances k = l , . . . . , t 1 = number of whorls nested within trees, families and provenances 1 = 1, ,w m = number of branches nested within whorls, trees, families and provenances m = 1, ,b are given i n Table 5.  36  TABLE 5.  Analysis of variance and expected mean squares (EMS) on whorl and interwhorl branch lengths, diameters and angles  Source of v a r i a t i o n  df  MS  Provenance  p-1  MS  Families within provenance  p(f-l)  MS  Trees within families, within provenances  pf(t-l)  MS  Whorls within trees, families and provenances  pft(w-l)  MS  Residual  pftw(b-l)  MS  Total  pftwb-1  where:  a p  EMS  a  II III  IV  2  o  E 2  a  Ii 2  iL  + bwa + bwta + bwtfa i r p 2  2  + bwa + bwta 2  1  + bwa  2  2  r  2  W  V  = variance among provenances MS. - MS ]  bwtf 2 a = variance among families nested within provenances _ II - III bwt M S  M S  2 a = variance among trees nested within families within provenances MS - MS III IV bw variance among whorls within trees within families within provenances MS - MS  w  IV  y  6  = variance of a fixed term (Anderson and Bancroft 1952)  a  = residual component of v a r i a t i o n -  M S  V  37  3.2  Heritability H e r i t a b i l i t y calculations were based on the assumption that open  pollinated progenies within  families were h a l f - s i b s .  Therefore a | i s  an estimate of a quarter of additive genetic v a r i a t i o n (^^) a  2 A  2 = 4 a (Falconer, 1960). r  s o  that  The assumption regarding the h a l f - s i b i s  p a r t i a l l y incorrect as i t i s d i f f i c u l t to admit that there are so many pollen sources contributing within  to a progeny as the number of individuals  the progeny test (Birot, 1976).  However, the probability i s  high for an individual to be pollinated by immediate neighbours. Therefore, a certain rate of f u l l - s i b s exists within a provenance and this may overestimate the h e r i t a b i l i t i e s .  P o l l i n a t i o n between related  trees may occur but since the rate of self p o l l i n a t i o n i s low (Sorenson 1971,  1973 and El-Kassaby, 1980), this factor can be ignored.  the above discussion,  h e r i t a b i l i t i e s estimated i n this study are the  maximum possible values. dbh,  Based on  The narrow sence h e r i t a b i l i t i e s for height,  root c o l l a r diameter, volume, taper and crown width were given by  the following  formula  2 , 2 2 °P °F °E +  (from Table 3). and  +  Those of whorl and interwhorl branches, yearly growths  their diameters were given by  38  (from Table 4).  H e r i t a b i l i t i e s for whorl and interwhorl  branches,  lengths, diameters and angles were obtained using the following formula  2 P  2 F  2 T  2 E  (from Table 5).  3.3  Juvenile-mature Correlations A tree breeder's major concern i n juvenile-mature c o r r e l a t i o n  work, i s the response of a t r a i t at l a t e r ages or at maturity as a result of selection on the juvenile t r a i t . lated response (CR) to s e l e c t i o n . character  This i s known as a corre-  The response of the correlated  can be predicted from the following formula CR = i h h r.a x y A py  i f the genetic c o r r e l a t i o n and the h e r i t a b i l i t i e s of the two t r a i t s are known (Falconer, Where  1960).  i  = selection intensity of the juvenile t r a i t  h  = square root of the h e r i t a b i l i t y  x  = juvenile t r a i t  y  = t r a i t at l a t e r ages or at maturity  r^  = genetic correlations between t r a i t s x and y  opy = phenotypic standard deviation of t r a i t y. A more detailed discussion of genetic and 17 of the l i t e r a t u r e review. estimating  correlations i s given on pages 16  Therefore, this study aimed at  genetic correlations between a l l the e a r l i e r and later  recorded t r a i t s .  The e a r l i e r ones consisted of c h a r a c t e r i s t i c s of  39  cones, seeds, germinants and from Kiss' (1971) work.  1+0  seedlings and  these were available  T r a i t s concerning parent tree locations  and  their c h a r a c t e r i s t i c s were l e f t out because they do not bear any genetic relationship for additive variance  estimation.  The  later  t r a i t s are those recorded from the 13-year-old trees by the author i n May  1980.  A l i s t i n g of the early, juvenile and 13-year-old tree  t r a i t s , i s presented in Table 6. Analysis of variance  using the ANOVAR package program was  out on a l l the e a r l i e r t r a i t s besides the l a t e r ones.  It was  carried  only  that  t r a i t whose family source of v a r i a t i o n showed s i g n i f i c a n t differences, that was  considered for genetic correlations (Peterson, 1981).  significance of the additive variance  The  depended on the significance of  the F value of the corresponding family component.  The argument  was  based on the nature of derivation of the correlated response (page 38). Non-significant  additive variance  give r i s e to non-significant  genetic  c o r r e l a t i o n which i n turn renders the correlated response meaningless. In addition non-siginificant additive variance  give r i s e to non-  s i g n i f i c a n t h e r i t a b i l i t i e s which also renders the correlated response meaningless.  With the assumption of h a l f - s i b s , the s i g n i f i c a n t e s t i 9  mated family component (o-p) was  multiplied by four (since i t estimates  a quarter of the additive variance) to obtain the estimated additive variance  2 (°^)«  Individual observations of the pair to be correlated  were then added and analysed to estimate the variance 2 (o" y). x+  This variance  covariance of the two 2 a , x+y  was  traits  used in the following formula to obtain a  traits (Cov )  = a  of the two  xv  2 2 + a + 2Cov x y xy  40 TABLE 6:  No.  Early, juvenile and 13-year-old tree t r a i t s used in the juvenile-mature genetic correlations  Traits  Units  Cone and seed measurements 1 2 3 4 5 6 7 8 9 10 11 12  Length of cone Width of cone Length of seed Width of seed Length of seed wing Width of seed wing A thousand seed weight No. seeds per cone No. f i l l e d seed per cone Endosperm class Embryo class Endosperm-embryo  mm mm mm mm mm mm mg  -  Germinants 13 14  Germination period Dormancy time  weeks weeks  15  Growing season  weeks  1+0 seedling measurements 16 17  Height Root c o l l a r diameter  18  No branches  m cm  Measurements at age 13 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33  Height DBH Root c o l l a r diameter Volume Taper Crown width Yearly growth Yearly growth diameter No. of branches i n a whorl No. of interwhorl branches Length of branches i n a whorl Diameter of branches i n a whorl Angle of branches i n a whorl Length of interwhorl branches Diameter of interwhorl branches  m cm cm m3 m m cm m cm degrees m cm  41  2 2 where o and o are the individual estimated additive variances. x x above estimated variances and covariance were then used i n the following formula to obtain the genetic correlations ( ^)« r  Cov  The  42  4.  4.1  RESULTS AND DISCUSSION  Variation and H e r i t a b i l i t y Analysis of covariance on height, diameter at breast height  (dbh), root c o l l a r diameter  (RCD), volume, taper and crown width  resulted i n s i g n i f i c a n t v a r i a t i o n only among provenances except for RCD and crown width (Table 7).  Considering the t o t a l phenotypic v a r i a t i o n  in height, 7.0% was attributed to differences among families within 2 provenances (ap) which i s an estimate of a quarter of the additive genetic variance (Table 8).  Approximately 39.0% of the v a r i a t i o n was  due to differences among provenances,  while the rest of i t 54.0% was  due to residual component (Table 8).  Besides the variation contributed  by the family component, the rest i s considered to be due to the remaining three quarters of the additive e f f e c t s , non-additive effects and environmental e f f e c t s , where non-additive effects consists of dominance and e p i s t a t i c (interaction between n o n - a l l e l i c genes) effects.  However, there was an attempt  of minimizing the environmental  component i n the experimental design by randomizing provenances and families within provenances on the experimental  site.  Analysis of heights of the same trees at age one (Kiss, 1971) revealed s i g n i f i c a n t v a r i a t i o n among both provenances and families. comparison with other studies also conducted provenances,  In  i n other Douglas-fir  Birot (1976), Christophe and Birot (1979) and Fashler  (1979) observed s i g n i f i c a n t v a r i a t i o n among provenances and families. El-Kassaby (1980) determined  the significance mainly among families.  TABLE 7.  H e r i t a b i l i t y estimates and F ratios (for a l l sources of variation) from the analyses of covariance ( t r a i t s 1-6) and variance ( t r a i t s 7-15) for the 15 measured t r a i t s of the 13-year-old Douglas-fir trees  Source of v a r i a t i o n Traits  Heritability  Provenance a  2  P  Families a  2  F  Trees a  2  T  Whorls 2 a  W _  1.  Height  0.28  7.29 **  1.39 ns  2. 3.  Diameter at breast height Root c o l l a r diameter  0.41 0.72  3.02 * 4.61 **  1.43 ns 2.06 *  4. 5.  Volume Taper  0.52  3.47 *  0.36  3.56 *  1.60 ns 1.39 ns  6. 7.  Crown width Yearly growth (YRG)  0.66 0.01  2.47 ns 7.00 A *  1.70 ns 1.35 ns  8. 9.  YRG diameter No. of branches i n a whorl  0.03 0.08  3.19 * 0.98 ns  1.43 ns 1.15 ns  0.93 ns  0.17  1.17 ns  1.30 ns  2.51 *  11.  No. of interwhorl branches Length of branches i n a whorl  0.02  3.04 *  1.15 ns  18.63 **  44.27 **  12.  Diameter of branches i n a whorl  0.00  6.29 *  0.49 ns  8.59 *  5.14 *  13.  Angle of branches i n a whorl  0.30  1.86 ns  4.60 **  2.42 **  14.  Length of interwhorl branches  0.00  2.34 ns  2.15 * 0.82 ns  11.36 **  15.68 **  1.39 ns  11.32 **  9.59 **  10.  15.  Diameter of interwhorl branches  * Significant at 0.05 probability l e v e l . ** Significant at 0.01 probability l e v e l , ns Not s i g n i f i c a n t .  0.10  2.40 ns  0.26 ns 2.13 *  -  TABLE 8.  H e r i t a b i l i t y estimates and percentages of various variance components making up the total phenotypic v a r i a t i o n , a = a family component 2  9  estimating 1/4 of additive variance,  9  op, a£>  9 a n d  Og = provenance,  tree and residual components respectively, estimating the rest of the genetic variance together with the environmental variance.  % contribution of each component towards t o t a l phenotypic v a r i a t i o n Traits  Heritability 2  2  S  a  p  2 o  F  2 a  2 a  T  E  Root c o l l a r diameter  0.72  31.4  17.9  4.  Volume  0.52  21.5  13.1  5.  Taper  0.36  20.8  9.1  6. 7.  Crown width Yearly growth (YRG)  0.66  14.5  16.6  -  0.01  1.3  0.2  0.0  98.5  8.  YRG diameter  0.03  1.8  0.8  0.0  97.4  9.  No. of branches i n a whorl  0.08  0.0  2.0  21.5  76.5  10.  No. of interwhorl branches  0.17  0.8  4.3  26.1  11.  Length of branches i n a whorl  0.02  10.6  2.6  51.6  68.8 35.2  12.  Diameter of branches i n a whorl  8.7  0.0  35.4  56.9  13.  Angle of branches i n a whorl  0.00 0.30  3.9  9.7  19.9  92.1 46.9  1. 2. 3.  14. 15.  Height Diameter at breast height  0.28  39.0  7.0  0.41  17.4  10.4  Length of interwhorl branches  0.00  7.9  0.0  0.0  Diameter of interwhorl branches  0.10  7.2  5.7  40.2  54.0 72.2 50.7 65.4 70.1 68.9  .  66.5  45  There i s a f a i r l y weak genetic control on height because of the r e l a t i v e l y low h e r i t a b i l i t y value of 0.28 (Table 7).  The same h e r i t a -  b i l i t y estimated by Kiss (1971) at the age of one year, was only 0.14. It i s rather inappropriate to compare the two values since Kiss' sample was f a r larger than that of the present study.  However, results from  the l i t e r a t u r e suggest higher h e r i t a b i l i t y values during the nursery phases and/or the early years of outplanting.  This i s due to the  influence of maternal effects and uniformly favourable environment (Barnes, 1973).  Birot (1976) together with Christophe and Birot  (1979), f o r instance reported h e r i t a b i l i t y values of 0.60, 0.46, 0.30 and 0.26 for heights of Douglas-fir trees at ages 1, 2, 3 and 4 i n that order.  The decrease i n h e r i t a b i l i t i e s was attributed to the disappear-  ance of maternal  effects and possibly onset of competition.  Further-  more, i n the same species, Fashler (1979) reported average h e r i t a b i l i t y values of 0.42 and 0.34 for ages 1 and 7 respectively. The family component contributed 10.4% towards the total phenotypic v a r i a t i o n i n dbh (Table 8).  Differences among provenances  accounted f o r 17.4% of the v a r i a t i o n while residual component accounted for 72.2%.  Selection of best individuals within the best provenances  for dbh i s l i k e l y to y i e l d reasonable genetic gains because of the strong genetic contro 1 ( h = 0.41). 2  The root c o l l a r diameter showed s i g n i f i c a n t v a r i a t i o n i n both among provenances and families (Table 7).  Out of the t o t a l  phenotypic  v a r i a t i o n , 17.9% i s attributed to the family component, while 31.4% i s due to differences among provenances and 50.7% due to residual component  46  (Table 8).  A l o t more genetic gains would be expected from RCD selec-  tion of the best individuals within the best families, because of the high h e r i t a b i l i t y value (0.72) as compared to dbh (0.41).  A similar  analysis (by Kiss, 1971) on RCD of the same trees at age one revealed the same v a r i a t i o n as obtained i n this study with a h e r i t a b i l i t y of 0.17 except that his sample was much larger than the one used i n this study. Both volume and taper showed s i g n i f i c a n t v a r i a t i o n only among provenances (Table 7).  The volume showed 21.5% provenance contribution  towards the t o t a l phenotypic  v a r i a t i o n (Table 8).  contributed 13.1% while 65.4% was due to residual. estimate  The family component The h e r i t a b i l i t y  (0.52) showed a f a i r l y strong genetic control.  The taper  showed 20.8% of the phenotypic v a r i a t i o n to come from the differences among provenances.  The family component contributed 9.1% while 70.1%  was due to residual.  The h e r i t a b i l i t y estimated was 0.36. However,  when making a decision concerning volume and taper, caution should be taken i n view of the probable traits.  influence of other factors on these  For example, the two t r a i t s are functions of height and dia-  meter, influenced by spacing i n the f i e l d , site quality geographical location and climate. Crown width showed no s i g n i f i c a n t v a r i a t i o n among provenances and families whose contributions were 14.5% and 16.6% respectively towards the t o t a l phenotypic residual.  variation.  The remaining  68.9% was due to the  There i s a strong genetic control ( h = 0.66) on crown 2  width despite the fact that no s i g n i f i c a n t v a r i a t i o n exists among the  47  partitioned sources.  According  to the results selection concerning  crown width should be based on individual trees within families within provenances. For yearly growth and diameter of yearly growth analysis of variance revealed s i g n i f i c a n t v a r i a t i o n only among provenances (Table 7).  Contributions towards total phenotypic  family components are very small 0.2%  and 0.8%  v a r i a t i o n by the  for yearly growth and  diameter of yearly growth respectively (Table 8).  Low  additive genetic  variances resulted i n low estimates of h e r i t a b i l i t i e s of 0.01  and  0.03  in the same order, suggesting the p o s s i b i l i t y of eliminating these t r a i t s from consideration for selection as breeding  traits.  Significant v a r i a t i o n for number of branches i n a whorl and number of interwhorl branches was  observed only among trees (Table 7).  There was a very small contribution towards the t o t a l phenotypic  varia-  tion by the family component resulting i n very low h e r i t a b i l i t y e s t i mates.  The h^s were 0.08  and 0.17  for number of branches i n a whorl  and interwhorl branches respectively. Again these two t r a i t s would be precluded from consideration for selection as breeding  traits.  The length and diameter of branches in a whorl showed s i g n i f i c a n t v a r i a t i o n among provenances, trees and whorls (Table 7).  Only 2.6%  the total phenotypic v a r i a t i o n i n length of branches in a whorl was  of due  to differences among families, giving rise to a low h e r i t a b i l i t y estimate of 0.02  (Table 8).  The diameter did not show any additive  effects towards the phenotypic h e r i t a b i l i t y estimate  v a r i a t i o n hence resulting in zero  (Table 8).  Once again i t would be advisable to  48  preclude these two t r a i t s from consideration for selection as breeding traits. The analysis of angles of branches i n a whorl revealed cant v a r i a t i o n among families, trees and whorls (Table 7).  signifiThere was a  f a i r contribution of the family component towards the total phenotypic v a r i a t i o n (9.7%) and a s i g n i f i c a n t h e r i t a b i l i t y estimate of 0.3 (Table 8).  Hence selection for branch angles based on good trees among the  best families i s l i k e l y to give a reasonable response i n terms of genetic gains. Significant v a r i a t i o n for length and diameter of interwhorl branches was observed among trees and whorls only (Table 7).  Similar  to many branch c h a r a c t e r i s t i c s , contributions towards the total phenotypic v a r i a t i o n by the family components are low, 0% for the lengths and 5.7% for the diameters (Table 8).  This resulted i n low  estimates of h e r i t a b i l i t i e s , 0.0 for the lengths and 0.1 for the diameters.  Because of low h e r i t a b i l i t i e s , these t r a i t s are also  recommended to be precluded from consideration for selection as breeding t r a i t s i n these particular provenances. C o l l e c t i v e l y , branch c h a r a c t e r i s t i c s except for the angles, have shown low h e r i t a b i l i t y estimates, which i s i n agreement with Dyson's (1969) findings.  Because of the low h e r i t a b i l i t y estimates, branch  c h a r a c t e r i s t i c s are not normally used as a basis f o r selection as breeding t r a i t s .  However, there i s a tendency for improved trees to  carry fine horizontal branches.  49  4.3  Juvenile-mature  Correlations  P a r t i t i o n i n g of the variance into additive and the rest (the rest being non-additive and environmental) revealed that most of the additive variances were not s i g n i f i c a n t except for the RCD and the angle of whorl branches (Table 7). I t i s possible that most of these  traits  showed non-significant additive variances because of the r e l a t i v e l y small sample size of only 7 provenances, 4 f a m i l i e s per provenance and 3 trees per family.  As recommended by Peterson  (1981), i t was only RCD  and branch angles that were considered for genetic correlations since they showed s i g n i f i c a n t additive variances. Analyses  of variance of a l l the e a r l i e r t r a i t s revealed only  embryo class and dormancy time to have s i g n i f i c a n t additive variances. Therefore, these were the ones considered for genetic c o r r e l a t i o n s . Results from the analyses of variance and the corresponding l i t i e s are presented  i n Table 9.  heritabi-  Embryos were c l a s s i f i e d into four  categories according to their v i s i b i l i t y on X-ray photographs i n r e l a t i o n to seed sizes ( S z i k l a i , 1964).  Class one consisted of absent  embryos while class two embryo was less than 50% of the seed length. Classes three and four consisted of embryos 50-70% and over 75% of the seed length respectively. Out of the 84 trees used i n this study 10% were i n embryo class two, 19% i n class three and 71% i n class four. The dormancy time was determined by calculating number of days between dates of sowing and seedling entering dormancy. estimates time.  The h e r i t a b i l i t y  for embryo class was 0.5 while 1.21 was for the dormancy  The overestimate  of dormancy time h e r i t a b i l i t y could be due to  50  TABLE 9.  H e r i t a b i l i t y estimates and F ratios (for each source of variation) from the analyses of variance of the 18 early and juvenile t r a i t s of the sample used i n this study  Source of v a r i a t i o n Trait  Heritability  provenance  4  family 2 °F a  1.  Length of cone  2.  Width of cone  3.  Length of seed  Analyses were impossible  4.  Width of seed  because only mean values  5.  Length of seed wing  were given per family  6.  Width of seed wing  7.  A thousand seed weight  8.  No. seeds per cone  9.  No. f i l l e d seeds per cone  10.  Endosperm class  0.0  3.00 *  0.67 ns  11.  Embryo class  0.5  6.22  **  1.75 *  12.  Endosperm-embryo class  0.32  6.74  *  1.44 ns  13.  Germination period  0.35  4.71 *  1.41 ns  14.  Dormancy time  1.21  2.47 ns  2.74 *  15.  Growing season  0.0  14.42 *  0.76 ns  16.  Height  0.0  10.54 **  0.97 ns  17.  Root c o l l a r diameter  0.0  9.11  18.  No. branches  0.31  5.78 *  * Significant at 0.05 probability l e v e l , ** Significant at 0.01 probability l e v e l . ns Not s i g n i f i c a n t .  **  0.84 ns 1.39 ns  51  the assumption of h a l f - s i b s when i t could have been f u l l - s i b s and also could be due to sampling errors.  Allen and Owens (1972) observed  pollen grains i n Douglas-fir, to be dispersed by wind, larger than most other conifers, and lacks wings or bladders. r e l a t i v e l y short dispersal distance.  As a result, they have  A small f r a c t i o n of pollen i s  dispersed a distance further than 5-10 times the tree height. of the above reasons the chances of f u l l - s i b s increase.  Because  The sample  size also could have had an e f f e c t . The estimated genetic correlation c o e f f i c i e n t between the embryo class and the RCD was r e l a t i v e l y high and positive (0.73).  The high  correlation suggests higher chances of accurate prediction of the correlated response of RCD at age 13 from the size of the seed embryo. The bigger the embryo, the bigger the RCD i s l i k e l y to be. These results suggest the p o s s i b i l i t y of carrying out selection at the seed stage.  Obviously this would be a great r e l i e f to tree breeders i n  terms of reducing the progeny testing periods. Since the majority of trees used i n this study (71%) belonged to the highest embryo class (4), i t i s advisable to select i f possible only those seeds with embryo class four.  Such early s e l e c t i o n would mean increasing selection  intensity which i n conjunction with high h e r i t a b i l i t i e s yields high genetic gains.  I t i s well known that RCD i s highly and positively  correlated with dbh and height, so that any selection imposed on RCD, i n d i r e c t l y r e s u l t i n a similar response i n dbh and height.  It i s  worth mentioning at this point that this i s the f i r s t time,embryo  52  class i s being used i n juvenile-mature correlation studies.  The author  consider i t an important finding worth pursuing i n future. The genetic correlation c o e f f i c i e n t estimated between dormancy time and RCD was 0.32, suggesting again that correlated response at age 13 can be predicted using the dormancy time.  of RCD  Results imply that  the longer the growing period, the larger the RCD would be at age 13. Selection at this very early seedling stage would also f a c i l i t a t e the tree breeding program i n the same way as selection at the seed stage. However, selection at the seed stage i s bound to be less costly i n terms of money and time as compared to selection at the seedling stage. According to the results i t i s obvious that the accuracy of prediction from dormancy time w i l l not be as precise as that from the embryo class.  Therefore, i t i s better to use embryo class i f i t could be  obtained.  Similar to embryo class dormancy time also i s being used f o r  the f i r s t time i n connection with juvenile-mature correlation studies. The angles of whorl branches also showed s i g n i f i c a n t additive variance, but they were not correlated with the embryo class, dormancy time and RCD.  This i s because the angles were analysed on a bigger  model to which the others could not be expanded.  However, there i s  almost a perfect relationship (r = 1.06) between the embryo class and dormancy time.  The overestimate could be due to the assumption of  h a l f - s i b s when i t could have been f u l l - s i b s , possibly the small sample size and may be the sampling errors.  Significant and high genetic  correlation c o e f f i c i e n t s obtained from this study suggest that embryo class, dormancy time and RCD share a certain set of genes.  53  Juvenile-mature correlation work has been primarily concerned with the seedling t r a i t s and those of the advanced ages.  Where the  seed weight (weight of seed coat, endosperm and embryo) has been considered, i t was found to be related with only the early ages' performance.  These effects were concluded to be maternal effects which  l a s t for short periods of time, and are of very l i t t l e importance i n long lived organisms l i k e forest trees (Barnes, 1973; Yao,  1971).  However, this study has revealed that pre-seedling t r a i t s could also be used i n predicting a correlated response at advanced  ages.  54  5. Results  CONCLUSION AND  RECOMMENDATIONS  of this study suggest that dbh,  root c o l l a r diameter  (RCD), volume, taper, and angles of whorl branches have reasonable amounts of additive variance  consequently f a i r l y large h e r i t a b i l i t i e s .  In fact i f the h e r i t a b i l i t i e s are real (since most of them are s i g n i f i c a n t ) selection based on these able genetic gains. RCD  t r a i t s would provide  not  consider-  Selection of a l l the above-mentioned t r a i t s except  and angles of whorl branches should be based on best individuals in  the best provenances.  In case of the two exceptions; which showed  s i g n i f i c a n t additive variance  and h e r i t a b i l i t i e s ; selection should be  based on the best individuals within f a m i l i e s .  In the f i n a l analysis  the best families are noted for more seed c o l l e c t i o n and cuttings.  possibly  On the whole, yearly growth, diameter of yearly growth, and  branch c h a r a c t e r i s t i c s showed very low quantities of additive variances consequently low h e r i t a b i l i t i e s . necessary to substantiate Results revealed  from genetic  However, further studies may  these findings. juvenile-mature correlation analyses  the p o s s i b i l i t y of predicting RCD  performance at age 13 from  the embryo class since a strong relationship ( r ^ = 0.73) The  be  same performance can be predicted  was  found.  from the dormancy time of the  germinants but with less precision as a result of the weaker r e l a t i o n ship ( r ^ = 0.32). Past investigations i n forest genetics have largely been concerned with revealing the genetic architecture of species, among them the extent of v a r i a t i o n , mode of inheritance and estimates of  55  h e r i t a b i l i t y to measure potential gains of selection for the different traits.  It is therefore clear from this study that genetic c o r r e l a -  tions between juvenile and premature or mature t r a i t s offer better alternative c r i t e r i a for s e l e c t i o n .  In view of t h i s , i t is strongly  recommended to continue investigations on genetic  juvenile-mature  correlations of certain tree c h a r a c t e r i s t i c s i n order to accumulate data that might lead to establishing r e l i a b l e and quantitatively testable c r i t e r i a for future selection programs i n tree breeding. would be p a r t i c u l a r l y useful to carry out the analyses on data obtained Lens Creek.  It  presented here  from the other two remaining areas, namely Haney and  The incorporation of larger sample sizes i n future  investigations should lead to more accurate estimates of genetic parameters.  The recording of various t r a i t s on a yearly basis should aid  in understanding possible trends of v a r i a t i o n i n these c h a r a c t e r i s t i c s . It i s further recommended that more experiments be established which include information on seed and  seedling c h a r a c t e r i s t i c s as these  appear to be a possible basis for future prediction of growth performance.  56  REFERENCES A l l e n , G.S.  1960.  A method of distinguishing Coast from Interior  Douglas-fir seed.  B r i t i s h Columbia Lumberman.  Vol. 44, August  26-30. A l l e n , S.G.  and J.N. Owens.  1972.  Information Canada Ottawa. Anderson, R.L.  and T.A.  research. 399  The l i f e history of Douglas-fir. Cat. No. F042-4972.  Bancroft.  1952.  S t a t i s t i c a l theory i n  New York, Toronto, London, McGraw-Hill Book Co.  pp.  Barnes, R.D.  1973.  The genetic improvement of Pinus patula (Schiede  and Deppe) i n Rhodesia.  Ph.D.  Thesis, Department of Forestry,  University of London, England. Birot, Y.  1976.  Genetic parameters between some populations of  Douglas-fir.  Consequences on selection.  I_n: Proc. IUFRO Joint  Meeting Advanced Generation Breeding. Bordeau - June 14-18. Bock, R.D.,  H. Wainer, A. Peterson, D. Thisseu, J . Murray and A. Roche.  1973.  A parameterization for individual human growth curves.  Human Biology 45(1):63-80. Christophe, C. and Y. Birot.  1979.  Genetic v a r i a t i o n within and  between populations of Douglas-fir.  Silvae Genetica (28)  5-6,  197-206. Denison, N.P.  1967.  A preliminary investigation of progeny t r i a l s of  the p r i n c i p l e Pinus species i n the Eastern Transvaal. (Unpublished).  57  Dunlap, L.H.  1964.  D i f f e r e n t i a t i o n of coastal and i n t e r i o r  provenances using morphology of seed. B r i t i s h Columbia, Vancouver. Dyson, G.W.  1969.  BSF Thesis, University of  44 pp.  Improvement of stem form and branching characteris-  t i c s i n Kenya Cypresses.  F.A.O. World Consult, of Forest Tree  Breeding F.O.-FTB-69 315 307-315. El-Kassaby, Y.A.  1980.  Isozyme patterns of a selected Pseudotsuga  menziesii (Mirb) Franco population. B r i t i s h Columbia, Vancouver. El-Lakany, M.H.  and 0. S z i k l a i .  Ph.D. Thesis, University of  137 pp.  1973.  Further investigation of  i n t r a s p e c i f i c v a r i a t i o n i n DNA contents of Douglas-fir, Pseudotsuga menziesii (Mirb) Franco.  Egypt. J. Genet. Cytol (2)  p. 345-354. Falconer, D.S.  1960.  Introduction to quantitative genetics.  Ronald  Press Co., New York, p. 365. Falkenhagen, E.R.  1972.  The h e r i t a b i l i t y concept: i t s v a l i d i t y i n  applied breeding. Unpublished term paper for FRST 502 course at the University of B r i t i s h Columbia, Vancouver. Fashler, K.A.M.  1979.  Intraspecific v a r i a t i o n in non-selected natural  populations of Douglas-fir (Pseudotsuga menziesii (Mirb) Franco). M.F. Thesis, University of B r i t i s h Columbia, Vancouver. Funk, E.M.,  H.C. Knandel, and E.W.  Callenbach.  1930.  Statistical  studies of the v a r i a t i o n i n the growth of SWL and their significance.  Poultry Science, 9:157-163.  74 pp.  58  Haddock, P.G., J . Walters, and A. Kozak. i n t e r i o r provenances of Douglas-fir  1976. Growth of coastal and (Pseudotsuga menziesii  (Mirb)  Franco) at Vancouver and Haney i n B r i t i s h Columbia Research Paper No. 19. Hattemer, H.H.  1963. Estimates of h e r i t a b i l i t y published  tree breeding research.  F.A.O./FORGEN 63-20/3:14-28.  Husch, B., C.I. M i l l e r , and T.W. Beers. Second e d i t i o n . Illingworth, K.  i n forest  1972. Forest mensuration.  The Ronald Press Co., New York.  410 pp.  1978. Stika spruce provenance t r a i t s , three years  after planting i n B r i t i s h Columbia.  _In: Proc. IUFRO Joint  Meeting (Vancouver, B.C., Canada) S2.02-12 (Abstract). Jaap, R.G. and Morris.  1937. Genetical differences i n eight-week  weight and fathering. Kiss, K., Gyula.  Poultry S c i . 16:44-48.  1971. Variation i n seed and seedling c h a r a c t e r i s t i c s  of selected provenances of Douglas-fir. B r i t i s h Columbia, Vancouver. Kozak, A.  Report.  (Unpublished).  University of  103 pp.  1980. Personal communication.  La Farge, T.  1975. Correlations between nursery and plantation height  growth i n slash and l o b l o l l y pine. Lambeth, C C .  1980. Juvenile-mature correlations i n Pinaceace and  implications for early selection. Lerner, I.M. 298 pp.  For. S c i . 21(2):197-200.  For. S c i . 26:571-580.  1958. The genetic basis of selection.  Wiley, New York.  59  Lush, J.L.  1949.  animals.  H e r i t a b i l i t y of quantitative characters i n farm  In: Proc. 8th Intern. Congr., Hereditas, Suppl. V o l .  356-375. Meier, R.J. and J.F. Goggans.  1977a.  H e r i t a b i l i t i e s and correlations  of c o r t i c a l monoterpenes of V i r g i n i a pine (Pinus virginiana Mill.).  Silvae Genetica  Meier, R.J. and J.F. Goggans.  (27) 2(1978) 79-84. 1977b.  H e r i t a b i l i t i e s of height,  diameter and s p e c i f i c gravity of young V i r g i n i a pine.  For. S c i .  23(4):450-456. Namkoong, G., R.A. Usanis, and R.R. S i l e n .  1972.  Age-related  v a r i a t i o n i n genetic control of height growth i n Douglas-fir. Theoretical and Applied Genetics Nanson, A.  1967.  42:151-159.  Contribution to the study of early tests.  Interna-  t i o n a l Scots Pine Provenance Test (1906) Trav Stn de Rech Groenenda Ser E (2), 43 pp. Nanson, A.  1968.  La valeur des tests precoces daus l a selection des  arbres f o r e s t i e r s en p a r t i c u l a r on point de vue de l a croissance. Ph.D. Thesis.  Gombloux, Stat. Rech. Eaux et Forets, Groenendaal,  H.S. n 38, 242 pp. Nanson, A.  1970.  Juvenile and correlated t r a i t selection and i t s  effect on selection programs, pp. 17-26.  Ta: Papers presented at  the second meeting of the Working Group on Quantitative Genetics, Section 22 IUFRO.  1969 USDA Forest Service South Forest Exp.  Stn. New Orleans, La.  60  Nanson, A.  1976.  Juvenile-mature relationships mainly i n provenance  and progeny tests.  Proceed, of the IUFRO Joint Meeting of  Working Parties on Advanced Generation Breeding.  Bordeaux-June  14-18, p. 99-119. Nepevue, G., R. K e l l e r , and E.T. Du Gros.  1978.  Juvenile selection  for wood quality i n Populus nigra and Populus euramericana. Annales des Sciences Forestieres 35(1):69-92. O'Malley, D.M.,  F.W.  Allondorf, and G.M.  Blake.  1979.  Inheritance of  isozyme v a r i a t i o n and heterozygosity i n Pinus ponderosa. Biochemical Genetics 17:223-250. Perry, D.A.  and J.E. Lotan.  1977.  Variation i n lodgepole pine (Pinus  contorta var. l a t i f o l i a ) : green house response of wind pollinated families from five populations to day length and temperaturesoil.  Can. J . For. Res. 8:81-89.  Peterson, R.G. Randall, W.K.  1981. 1977.  Personal communication. Short Note: Growth correlations of cottonwood  clones developed from mature wood cuttings.  Silvae Genetica  26(2-3):119-120. Reck, S.G. and 0. S z i k l a i .  1970.  Juvenile-mature correlation i n wood  c h a r a c t e r i s t i c s of Douglas-fir [Pseudotsuga menziesii (Mirb) Franco]. Robinson, B.A.  Unpublished. 1963.  Variation i n seed characteristics of Douglas-fir  in B r i t i s h Columbia. Vancouver.  66 pp.  BSF Thesis, University of B r i t i s h Columbia,  61  Sakai, K.I. and S. Hetakeyama.  1963.  Estimates of genetic parameters  in forest trees without raising progeny.  FAO World Consult For.  Genet., Stockholm No. FAO/FORGEN 63/20/4, p. 5. Schmidt, W.  1927.  pH-Reaktion bei Kiefern-und Fichtenkeimlingen  verschiedener geographischer Herkunft.  Forstarchiv.  Quoted by  Schmidt, 1957. Schmidt, W.  1930.  Berlin. Schmidt, W.  Unsere Kenutnis von Forstsaatgut Deutche Forstwirt,  256 pp. 1935.  Die Rassendiagnose i n der Praxis der  Waldsamenprufungsanstalt.  Jahresbericht Hauptaussch f o r s t l .  Saatgutenerken., 6-15. Schmidt, W.  1936.  Neue Wegeder Rassenforschung und Kiefernanerkennung.  Deutsch. Forstv., 433-453. Schmidt, W.  1964.  Shelbourne, C.J.A.  Early tests, F.A.O./F0RGEN-63 2a/10, 10 pp. 1969.  Breeding for stem straightness i n conifers.  In: Proc. Second World Consul, on Forest Tree Breeding, l:293-FAO. Shirikhande, V.J. '  1957.  on coconut trees.  Singh, S.M.  Some considerations i n designing experiments J . Indian Soc. Agr. S t a t i s t i c s 982-99.  and Zouros, E.  1978.  Genetic v a r i a t i o n associated with  growth rate i n the American oyster Crassostrea v i r g i n i c a . Evolution 32:342-353. Sluder, E.R.  1979.  The effects of seed and seedling size on survival  and growth of l o b l o l l y pine.  Planters Notes (30):25-28.  62  Sorenson, F.C.  1971.  Estimate of s e l f - f e r t i l i t y i n Coastal Douglas-  f i r from inbreeding Sorenson, F.C.  1973.  studies.  Silvae Genetica 20(4):115-20.  Frequency of seedlings  f e r t i l i z a t i o n in Coastal Douglas-fir.  from natural  self-  Silvae Genetica  22(l-2):20-24. Squillace, A.E.,  R.T.  Bingham, G. Namkoong, and H.F.  Robinson.  1966.  H e r i t a b i l i t y of juvenile growth rate and expected gain from selection i n western white pine. Squillace, A.E.  and R.C.  i n slash pine. Steinhoff, R.J.  Gansel.  S z i k l a i , 0.  In: Proc. Joint IUFRO Meeting S.02.04.1-3  p. 243-250.  and A.J.F. G r i f f i t h s .  analysis.  Juvenile-mature correlations  Juvenile-mature correlations in ponderosa and  Western white pines.  Suzuki, D.T.  1974.  For. S c i . 20(3):225-229.  1974.  Stockholm,  Silvae Genetica 16(1):1-6.  W.H.  1964.  1976.  Freeman & Co.,  An introduction to  San Francisco.  menziesii.  Ph.D.  Vancouver.  136  1974.  Thesis, University of B r i t i s h Columbia, pp.  Juvenile-mature c o r r e l a t i o n .  In:  ecological genetics  Proc. of the  1958.  characters  and  breeding theory and progeny testing.  Stockholm S.02.04.1-3 Session IV.  p. 217-235.  Variation and h e r i t a b i l i t y of some quantitative i n Cryptomeria.  and  (Mirb) Franco var.  IUFRO Joint Meeting of Working Parties on population  Toda, R.  pp.  Variation and inheritance of some physiological  morphological t r a i t s i n Pseudotsuga menziesii  S z i k l a i , 0.  468  genetic  Silvae Genetica 7(3):87-93.  63  Wakeley, C P .  1971.  Relation of thirteenth year to e a r l i e r  of southern pines. Warner, J.N.  1952.  dimensions  For. S c i . 17(2):200-209.  A method of estimating h e r i t a b i l i t y .  Agron. J .  44:427-430. Warwick, E.J. and J.E. Legates. farm animals.  7th Ed.  Whiteside, I.D., M.D.  1979.  Breeding and improvements of  McGraw-Hill Book Co.  Wilcox, and J.R. Tustin.  624 pp.  1977.  New  Zealand  Douglas-fir timber quality i n r e l a t i o n to s i l v i c u l t u r e . For.  22(l):24-44.  W i l l e t t , V.H.S.  1963.  Variation i n cone characteristics of Douglas-  f i r i n B r i t i s h Columbia. Columbia, Vancouver. Wright, J.W.  1945.  BSF Thesis, University of B r i t i s h  65 pp.  Influence of size and portion of cone and seed  weight i n Eastern White Pine. Wright, J.W.  1976.  New York. Wyk,  G.  1976a.  J . For. 43:817-819.  Introduction to forest genetics. Academic Press  463  pp.  Early growth results i n a d i a l l e l progeny test of  Eucalyptus grandis ( H i l l ) Maiden. Wyk,  G.  1976b.  G.  1977.  Silvae Genetica, 25:126-132.  Early growth results i n a d i a l l e l progeny test of  Eucalyptus grandis ( H i l l ) Maiden. Wyk,  N.Z. J .  Silvae Genetica 26(l):44-50.  Does tree improvement influence the harvesting and  transport of trees? Yang, J.-Ch., T.M.  South African For. J . No.  Ching, and K.K.  Coastal Douglas-fir. three enzyme systems.  Ching.  1977.  105.  Isozyme variation of  1. A study of geographical v a r i a t i o n i n Silvae Genet. 26(1):10-18.  64  Yao,  C.  1971.  Geographical  variation i n seed weight, some cone scale  measurements and seed germination menziesii (Mirb) Franco). Columbia, Vancouver. Yeh,  F.C. and C. Layton. variability  88  M.F.  of Douglas-fir (Pseudotsuga  thesis, University of B r i t i s h  pp.  1979.  The organization of genetic  i n central and marginal populations of lodgepole  (Pinus contorta spp. l a t i f o l i a ) .  pine  Can. J . Genet. Cytol.  21:487-503. Yeh,  F.C. and Y.A.  El-Kassaby.  1980.  population of Sitka spruce  Enzyme variation i n natural  (Picea sitchensis (Bong) Carr.).  1. Genetic v a r i a t i o n patterns among trees from ten IUFRO Provenances. Yeh, F.C. and D.M.  Can. J. For. Res. 10:415-422. O'Malley.  1980.  Enzyme variation i n natural  populations of Douglas-fir (Pseudotsuga menziesii (Mirb) Franco) from B r i t i s h Columbia. populations. Ying, C C .  and E.K.  1. Genetic variation patterns i n Coastal  Silvae Genet. Morgenstern.  (In press). 1979.  Correlations of height growth  and h e r i t a b i l i t i e s at different ages i n white spruce.  Silvae  Genetica 28(5-6):181-185. Zobel, B.J.  1961.  Inheritance of wood properties i n conifers. Silvae  Genetica 10(3):65-70.  APPENDIX 1  65  Experimental design f o r provenance - progeny test of Douglas-fir at the South Campus Research Nursery (Universitv of B r i t i s h Columbia')  Prov 1  13  x3x,6 xM  Fins , 6 x 3x 5 i » 3 A  x  **x  *  V"x6 X x?  S x ^ x '  3x  'x  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X.  X.  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X 5  X  X.  X  X  X  X  X  X4  10*  X.  X  X  X  X  X  X  X S  x s 4x xS S X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  ..Prov 15  X  X  X  X  X  *3 2x  X  X  X  X  X  X  X  X  X<y  X  X  X  X  X  X  X  X  X  X  X  X  V  X  X  X  X  X  X  X  X  >3x  X  X.  X  X  X  X  X X  X  <X  X  X  X  X  X  X  X  X 1 J x  X  X  X  X  >  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X4  5"x  X  X  X  x3 «x  X  X  X  X  X  X  X  X X  X  X  X  v  X  X  xff 6x  •x  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X » l((X  X  X  X  X  X  X  X  x  X  x  X  X  X  X  X  X  X  X  X  X  X.  X  X •x  X  X  X  X  X  *  X  X  X  X  X  X  X  X  K  X  X  X  X  X  X  2x  X  X  X  X  X  A  X  f x  X  X  X  X  X  X  X  jx X 4 gx x2 I x  X  X  X  -X  X  • X  X  X  X  X  X  X  X  x  X  X  K  X X  X  X  X  X  X  X  X  xis  6  X S  X  X  X  X  X  X  X  XiS  X  X  X  X  X  X  X  X  X  X  K  xq  II  x  X  X  X  X  X  X  X  X  X  X  X  XjT  K»X  X  X  X  X  X  X  X  X  X  X  X  X  X  X3  i6x  X  X X  X  X  X  X  X  X  X  X  X  X  X  X  * I  18*  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X2  SX  X  X  X  X  X  X  X  X  X  X  X  X  X  X3  «X  X  X  X  X  X  X  SX  li|X X  X  X  X  X  X  X X X  X  X  X  X  X  X  X  IW  X  X  X  X  X  X  X  X  X  X  X  X  X  -X  X  X.<<  X  X  X  X  X  X  X  x y itx  X  X  X  X  X  X  X 6  X  X  X  X  X  *.  X  X i  1*  X  X  X  X  X  X  X  X  X  X  X  X  X  X6 2.x  X  A.  X  X  X  X  X  X  X  X  X  X  X  X  X8  3X  X  X  X  X.  X  X  X  X  X  X  X  X  XI  <|X  X,  X  X  X  X  X  XO  X  X  > x  X  x.  X X  X  X  X A  X  X  X  X  X  X  >M  I X  X  X  X  X  X  X  *  X  X  X  x6  f x  X  X  X  X  X  X  X  X  X  X.  x ?  «x  X  X  X  X  X  X  X  X X  X  X  x n  7-X  X  X  X  X  X  X  X  X  X l l ( IPX  X  X  X  X  X  X  X  X  X  X  x-2,  A  X  X  x i  X  X  X  X,  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  11  12  X  X  X  X  i x  X  X  X  X  X  X  3x  X  X  X  X  X  X  X  X  X  X  X  X  X  X7  5 *  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X !  6 x  X  X  X  X  X  X  X  X  X  X  X  X  X  X  x ' o  8*  X  X  X  X  X  X  X  X  X  X  X  X  X 1  IX  X  X  X  X  X  X  X  X  X  X X  X  X  X  X  x s ax  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X 8  X  X  X  X  X  X  X  X  0  7X  X  X  X  X  X  X  X  X  x i © gx.  X  X  X  X  X  X  X  X  X  X  X  X  X  X  A(2. It*  X  X  X.  K  X  X  X  X X  X  X  X  X  X  X X  X  X  X  X  X  X  X  x3  X  X  X  X  X  X  X  X  X  X  X  X  x.  X  »X  X  X  X  X  X  X  x-  <  X X  X  X  X  X  X  X  X  X  X  X  X  X  X  X X  '*  X  X  X5  X  X  Xg  X  X  X  xq 1*  X  <  X  X  X  X  X  X  X  X  Xis  10*  X  X  X  X  X  X  X  X  X  X  X  X  XN  X  X  X  X  X  ax ax  X  X  >; x i  X  X  X  X  X  X  X  X  <  X  X  X  X  X*» 4 X  X  X  X  X  X  X  X  X  X  X  X  xq  X  X  *;  i x  X  X.  X  X  V  X  X  X  X  X  XI© i | X  X  X  X X  X  X  X  X  X  X  KEY  Prov - Provenances Fms - F a m i l i e s Trees are 3 w i t h i n rows X - Individual trees  20  

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 8 17
United States 8 0
Italy 4 0
Canada 1 0
Greece 1 0
Japan 1 0
Unknown 1 0
City Views Downloads
Beijing 7 6
Ashburn 6 0
Maccarese 4 0
Sunnyvale 1 0
Shenzhen 1 11
Mountain View 1 0
Unknown 1 2
Tokyo 1 0
Athens 1 0
Vancouver 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

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.831.1-0095201/manifest

Comment

Related Items