Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Intraspecific variation in non-selected natural populations of Douglas-fir (Pseudotsuga menziesii (MIRB.)… Fashler, Anita Marie Kvestich 1979

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

Item Metadata

Download

Media
831-UBC_1979_A6 F38.pdf [ 3.67MB ]
Metadata
JSON: 831-1.0075131.json
JSON-LD: 831-1.0075131-ld.json
RDF/XML (Pretty): 831-1.0075131-rdf.xml
RDF/JSON: 831-1.0075131-rdf.json
Turtle: 831-1.0075131-turtle.txt
N-Triples: 831-1.0075131-rdf-ntriples.txt
Original Record: 831-1.0075131-source.json
Full Text
831-1.0075131-fulltext.txt
Citation
831-1.0075131.ris

Full Text

c o p . I INTRASPECIFIC VARIATION IN NON-SELECTED NATURAL POPULATIONS OF DOUGLAS-FIR (PSEUDOTSUGA MENSIESII (MIRB.) FRANCO) by ANITA MARIE KVESTICH FASHLER B.S.F., University of B r i t i s h Columbia, 1976 A THESIS SUBMITTED IN PARTIAL FULFILMENT 0F THE REQUIREMENTS FOR THE DEGREE OF MASTER OF FORESTRY in THE FACULTY OF GRADUATE STUDIES THE FACULTY OF FORESTRY We accept th i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA A p r i l , 1979 ©' Anita Marie Kvestich Fashler, 1979 In presenting th i s thesis in pa r t i a l fu l f i lment of the requirements, for an advanced degree of the University of B r i t i s h Columbia, I agree that the Library shal l make i t f ree ly ava i lable for reference and study. I further agree that permission for extensive copying of th i s thesis for scholar ly purposes may be granted by the Head of my Depart-ment or by his representatives. It i s understood that copying or publ icat ion of th i s thesis for f inanc ia l gain shal l not be allowed without my written permission. Department of ho re.shy The Univers ity of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, B. C. V6T 1W5 Canada - i i -ABSTRACT Seedlings from the International Union of Forestry Research Organiza-tions (I.U.F.R.O) co l lect ions in 1966 and 1968 sampling the range of Douglas-f ir from northern Ca l i f o rn ia to B r i t i s h Columbia were used to establ i sh a provenance-progeny test at the Univers ity of B r i t i s h Columbia Research Forest in Haney. The study analyzed height growth in a to ta l of 384 ha l f - s i b famil ies representing 48 provenances. The objectives were 1) to estimate the degree of genetic var iat ion between and within, proven-ances for height growth, 2) to estimate the addit ive genetic variance, 3) to estimate narrow sense h e r i t a b i l i t y , 4) to estimate juveni le x mature co r re l a t i on , and 5) to se lect the best provenances and progenies for the Haney planting s i t e . Results from seed zone analysis showed that the most s i gn i f i can t differences in the genetic var iat ion in height growth of juven i le Douglas-f i r trees was found in the r e l a t i ve s izes of the variance between proven-ances (Vp) and the variance between fami l ies within provenances (Vp^p). Provenances most adapted to Haney conditions exhibited Vp^p and V.^as the dominant contributors to the phenotypic va r i a t i on . For less adapted provenances, Vp was of greater importance. In a l l zones the greatest component of variance was due to the variance of indiv idual trees within famil ies (V^.)- Variance due to block x provenance interact ion (Vp xg) was the next largest variance between blocks (Vg). Estimates for addit ive genetic variance and h e r i t a b i l i t y for seed zones were quite h igh. Low values for the i r respective standard errors indicated high r e l i a b i l i t y in the resu l t s . High values of h e r i t a b i l i t y indicated that there are opportunities for s i gn i f i can t improvement by se lect ion in Douglas-f ir. Results from the juveni le x mature corre lat ion analysis indicated that r e l i ab l e se lect ion of the best and delet ion of the poorest provenances and fami l ies may begin at age f ive.years. I t i s recommended that se lect ion in seed zones 2 and 3 and the best provenances from seed zone 1 would y i e l d the best results for the Haney s i t e . As an example of genetic gain, a se lect ion intens i ty of only one in four (25%) of the top indiv iduals was chosen. Using th i s low se lect ion in tens i ty , f igures obtained for genetic gain at Haney varied from 17.90% for two year to 10.96% for eight year height growth. Selection in the best provenance, of the best seed zone could increase tota l height growth by almost 33%. An addit ional increase of 70% is suggested i f the best indiv idual within the best provenance is chosen. Further gains in height growth are possible i f higher se lect ion i n tens i t i e s are used. - iv -TABLE OF CONTENTS Page INTRODUCTION 1 LITERATURE REVIEW 5 1. Components of Variance 5 2. H e r i t a b i l i t y . 7 3. Juvenile x Mature Correlat ion 10 MATERIALS AND METHODS 14 RESULTS AND DISCUSSION ; 28 1. Genetic Var iat ion in Height Growth 28 2. Estimation of Additive Genetic Variance (V^) 37 3. Estimation of H e r i t a b i l i t y 39 4. Juvenile x Mature Correlat ion 45 5. Selection of the Best Provenances and Progenies for the Test S i te 46 SUMMARY 54 LITERATURE CITED 57 APPENDICES: I. Least-Squares Analysis of Variance (Seed Zones 1, 2 and 3 and Inter ior) 61 II. Components of Variance (Seed Zones 1, 2 and 3 and Inter ior) 63 I I I. Regression Analysis for Maternal Effects (Seed Zones 1 and 2) 67 2 IV. H e r i t a b i l i t y (h ) and Standard Errors (S.E.) for Total Height Based on Individual Provenances ... 69 V. Ranking According to Mean 1975 and 1978 Total Height (Seed Zones 1, 2 and 3 and Inter ior) 70 VI. Mean (m) and Standard Deviation (S.D.) of 1972 to 1978 Total Height for Provenances in Seed Zones 1, 2 and 3 and Inter ior 74 - V -LIST OF TABLES Table Page 1 Elevat ion, Lat itude, Longitude and Thousand Seed Weight (TSW) of the Douglas-fir Provenances ... ... ... : 17 2 Analysis of Variance and Expected Mean Squares for Analysis of Between Provenance Var iat ion 21 3- Analysis of Variance and Expected Mean Squares for Analysis of Within Provenance Var iat ion 23 4 Total 1978 Height Differences. Between Seed Zones 29 5 Maximum, Minimum, Average and Standard Deviation of 1978 Total Height in Seed Zones 1 and 2 30 6 Maximum, Minimum, Average and Standard Deviation of 1978 Total Height in Seed Zone 3 31 7 Maximum, Minimum, Average and Standard Deviation of 1978 Total Height in the Inter ior Provenances 32 8 Addit ive Genetic Variance (VA), Standard Errors (S.E.) (S.E. (V A)/V A)100 and V A as a Percent of Total Phenotypic Variance for Total Height: Seed Zones 1 and 2 40 9 Addit ive Genetic Variance (VA), Standard Errors (S.E.) (S.E. (V A)/V A)100 and V A as a Percent of Total Phenotypic Variance for Total Height: Seed Zone 3 and Inter ior Provenances 41 10 H e r i t a b i l i t y (h 2) and Standard Error (S.E.) for Total Height Based on Seed Zones 43 11 Juvenile x Mature Correlat ion: Seed Zone 1 47 12 Juvenile x Mature Corre lat ion: Seed Zone 2 ... 48 13 Juvenile x Mature Correlat ion: Seed Zone 3 49 14 Juveni le x Mature Correlat ion: Inter ior Provenances 50 15 Estimates of Genetic Gain for Height Growth in Seed Zones 1,2 and 3 and Inter ior Provenances 53 - vi -LIST OF FIGURES Figure Page 1 Location of Douglas-fir Provenances 16 ACKNOWLEDGEMENTS I am very grateful to Dr. 0. S z i k l a i , Faculty of Forestry and cha i r -man of my thesis committee for valuable assistance and continuous encourage-ment in the preparation of th i s thes i s . Gratitude i s also expressed to the other members of my committee - Mr. Bruce Devitt , Pac i f i c Logging Co. L td . ; Dr. C F . Wehrhahn, Inst i tute of Animal Resource Ecology and Dr. D.H.. Wil l iams, Faculty of Forestry - for reviewing the thesis and contr ibuting the i r useful advice. In add i t ion, acknowledgement i s extended to Drs. J . Hodges and R.G. Peterson, Faculty of Agr icu l ture, Department of Animal Science for recommending the least-squares computer program used herein. I am also grateful to Mr. M. Hoque, PhD. candidate and Mr. D. Lee, research ass i s tant , Department of Animal Science for guidance in handling the program. I wish to thank Ms. G. Ho and Ms. S. Phelps for t he i r assistance in computing. I am greatly indebted to the B r i t i s h Columbia Ministry of Education and to Pac i f i c Logging Co. Ltd. for the i r f i nanc ia l support in the form of the Graduate Research Engineering and Technical (G.R.E.A.T.) grant. Thanks are given to Mrs. M.A. DeVescovi fo r her technical assistance and advice and to Ms. J . Kel ly for the typing of the manuscript. And f i n a l l y , I wish to thank my husband, Robert, fo r his support and cooperation. INTRASPECIFIC VARIATION IN NON-SELECTED NATURAL POPULATIONS OF DOUGLAS-FIR (PSEUDOTSUGA MENZIESII (MIRB.) FRANCO) INTRODUCTION The study of natural var iat ion within a tree species of commercial value i s required to derive basic information on population structure. The information of greatest interest includes the various components of variance.contributing to the tota l phenotypic variance and estimates of the h e r i t a b i l i t y of characters of in teres t . The components of variance are useful in assessing the degree of genetic d i f f e ren t i a t i on among the trees. H e r i t a b i l i t y values are useful in ca lculat ions to determine the amount of possible genetic gain. An understanding of both the amount and nature of components of variance and h e r i t a b i l i t y w i l l allow the tree breeder to increase the e f f i c i ency .o f tree improvement programs by evaluating the se lect ion procedures and predict ing what improvements can be ant ic ipated by various breeding methods (Campbell, 1972; Meier and Goggans, 1978; Rink and Thor, 1976; Owino, 1977 and Stonecypher, 1966). For example, based on information on geographic and indiv idual tree var iat ion within the l o b l o l l y pine range Zobel and Dorman (1973) have defined seven d i f fe rent provenances that could be used in d i f fe rent areas as exot ics. Estimates of variance components and h e r i t a b i l i t y have been prev-iously calculated for several tree species (see Hattemer, 1963 for a review), - 2 -although few studies were designed s p e c i f i c a l l y to estimate genetic v a r i -ance, the majority i n i t i a t e d as progeny and provenance tests . Such progeny tests and provenance studies designed to minimize the variance of a mean are not necessari ly good for estimating of genetic and environmental v a r i -ance (Goggans, 1962). Therefore, experiments designed espec ia l l y for estimating variance components w i l l generally y i e l d more r e l i a b l e and useful information (Stonecypher, 1966). However, the high cost of estab-lishment, maintenance and evaluation of long term tree breeding experi=-ments requires a m u l t i p l i c i t y of objectives to j u s t i f y the i r usefulness. Sacr i f i ces in maximum r e l i a b i l i t y in a l l estimates for many tree breeding experiments must, therefore, be made. Recent work on the ca lcu lat ion of genetic parameters for Douglas-f ir (Pseudotsuga menziesii (Mirb.) Franco) has been published. Campbell (1972) studied the genetic var iat ion in seedling height increment of Douglas-f ir and estimated addit ive and dominance genetic variances. Results showed that the dominance portion of genetic variance was quite important account-ing for 12 to 65 percent of the to ta l genetic variance; estimates of addit ive genetic variance were low with h e r i t a b i l i t i e s ranging between 0.08 and 0.22. Later, B i rot (1974 and 1976) studied 26 Douglas-f ir prov-enances from the f i r s t co l l ec t i on made by International Union of Forest Research Organizations (I.U.F.R.O) (Barner, 1973; Fletcher and Barner, 1978). Several character i s t i c s were measured and calculated h e r i t a b i l i -t ie s ranged from 0.21 to 1.0 for cotelydon number, 0.05 to 0.75 for growth cessat ion, 0.15 to 1.51 for f lushing time, 0.19 to 1.0 for tota l f i r s t year height and 0.14 to 0.79 for tota l second year height. In contrast to Campbell, these r e l a t i v e l y high estimates suggested that the - 3 -above character i s t i c s are under strong genetic cont ro l . Working on 10 to 15 year old rac i a l crosses of Douglas-f ir Orr-Ewing and Yeh (1978) also found low estimates of h e r i t a b i l i t y suggesting that both growth and survival were dependent on weak addit ive genetic variance. The apparent contradictory results of these studies indicate that the estimation of genetic parameters of Douglas-f ir populations require greater e f f o r t in the future. The material used in th i s study (Kvestich, 1976) i s unique since i t samples the range of Douglas-f ir from northern Ca l i f o rn ia to B r i t i s h Columbia, and provides the opportunity to assess the genetic character i s t -ics of several d i f fe rent populations of Douglas-fir simultaneously. A combined provenance-progeny tes t of 48 non-selected provenances of Douglas-f i r including.384 open-pollinated fami l ies were used to test var iat ion in height growth. The objectives of the study were: 1. To estimate the degree of genetic var iat ion for height growth among various Douglas-f ir provenances when grown at Haney, B r i t i s h Columbia. Var iat ion in height was studied at two l eve l s : (a) between provenance (b) within provenance 2. To estimate the addit ive genetic component for open-pollinated progeny from d i f fe rent provenances. 3. To estimate h e r i t a b i l i t y . 4. To study the consistency of height growth of fami l ies in each year since 1971 ( i . e . to estimate " juveni le x mature" cor re lat ion of progeny). - 4 -5. To select the best provenances and progenies for the Haney p lant-ing s i t e . - 5 -LITERATURE REVIEW 1. Components of Variance The phenotypic variance of an indiv idual cha rac te r i s t i c may be p a r t i -tioned into two major categories; the genetic variance and the environ-mental variance (Falconer, 1960). It has long been recognized that the tota l genetic variance could be subdivided into three d i f fe rent portions: a f i r s t due to the average effects of genes (V^), a second due to the interactions of a l l e l i c gene effects (dominance) (Vp) and a th i rd due to interact ions of non -a l l e l i c gene effects (epistas i s ) (Vj) (Fisher, 1918). General theoret ica l expectations for further subdivision of the to ta l genetic variance into portions have been carr ied out more recently by several authors (Cockerham, 1954; Henderson, 1954; Kempthorne, 1955; and L i , 1954). Their additions included interact ions among average effects of two or more non -a l l e l i c genes ( V ^ + V ^ A + e t c . ) , interact ions of two or more non -a l l e l i c dominance effects ( V D D + V D D Q + etc.) and interact ions between dominance and average effects of two or more non -a l l e l i c genes ^VDA + VDDA + VDAA + e t c - ) - T n e di vi s i on of the tota l phenotypic variance (Vp) can then be expressed as VP = V A + V D + VI < V A A + V A A A + + VDD + VDDD + + V D A + V D D A + V D A A + - > + V E + V J where Vj equals ef fects due to ep i s ta s i s , V £ equals environmental l i near ef fects and Vj equals environmental non-l inear ef fects (genotype x environ-ment interact ion) (Toda, 1958). - 6 -The components of tota l genetic variance may be derived from analysis of variance of ha l f or fu l T:.sibs. For example, Lacaze and Arbez (1971) studied within provenance var iat ion in Pinus s y l ves t r i s L. and subdivided the variance into between family and within family components. The v a r i -ation between family expresses the addit ive genetic variance of mother trees and the variance within family expresses various genetic (addit ive or not) and environmental e f fec t s . With the assumption of ha l f - s i b s , the genetic effects include three-quarters of the addit ive genetic variance and a l l the dominance va r ia t ion . Eiche and Anderson (1974) provided a good explanation of some of the appl icat ions of the components of variance. If the addit ive genetic variance is high with respect to the tota l phenotypic variance, the breeder should aim for mass se lect ion and cross breeding. High dominance variation.means that family se lect ion and hybr id izat ion of trees with high spec i f i c combining a b i l i t y would y i e l d the best results from breeding. I f both the addit ive and dominance genetic variance ef fects are low and the genotype x environmental interact ion i s high, breeding e f fo r t s should produce separate l i nes for each ecological region. When overdominance i s important, inbreeding with the object of producing hybrids between unrelated l ines i s the recommended program aim. Any model developed for the estimation of genetic variance involves various b io log ica l assumptions. Common genetic assumptions concerning parent populations and progenies include: a) normal d ip lo id and Mendelian inher itance, b) population in linkage equi l ibr ium, c) re la t i ves not inbred, - 7 -d) re la t i ves random members of a non-inbred population (no se lec t i on ) , e) re lat ives are ha l f - s i b s , f ) no maternal or cytoplasmic e f fec t s , g) no mult iple a l l e l e s , h) no ep i s tas i s (Stonecypher, 1966 and Sprague, 1966). Attention should be given to the v a l i d i t y of such assumptions in the test considered and the ef fects of any such i nva l i d assumptions on the estimates of genetic variance should be discussed. 2. H e r i t a b i l i t y H e r i t a b i l i t y (h ) i s the f ract ion of the observed or phenotypic variance which was caused by differences between the genes or the genotypes of the ind iv idua l s , that i s , the re l a t i ve amount of var iat ion passed on to the next generation by each ind iv idual (Lush, 1949). H e r i t a b i l i t y values can be reported in two ways; in the "narrow" sense or in the "broad" sense (Toda, 1958). When material i s propagated by sexual means, the non-addit ive ef fects of the genotypes are not passed on to the progeny. This r a t i o of the r e l a t i ve amount of addit ive genetic variance (V^) to the to ta l phenotypic variance (Vp) i s the narrow sense h e r i t a b i l i t y (h n s = ^ / V p ) / When material i s propagated vegetat ively, the. ef fects of dominance and epis tas i s are passed on since the genotypes are transferred unchanged. In th i s case the r a t i o of the re l a t i ve amount of tota l genetic variance (VQ) to the to ta l 2 phenotypic variance, that i s , the broad sense h e r i t a b i l i t y (h ^ s = V^/Vp)./ 2 2 Narrow sense h estimates are frequently less than h in the broad sense so possible gains from breeding organisms that reproduce sexually are usually smaller than gains from organisms reproducing asexually. - 8 -H e r i t a b i l i t y can be calculated in several ways (Hattemer, 1963): a) ' s ib analysis (estimates h^ n ) 2 b) clonal analysis (estimates h ^ ) c) parent-offspring regression (estimates .h ). 2 However, although h estimates the re l a t i ve importance of the genetic com-ponents of variance, they are most useful in the assessment of genetic gain. 2 Several factors should be kept in mind when dealing with h estimates. 2 F i r s t , h estimates are not constant for a l l conditions (Zobel, 1961). Any-thing that causes an increase in var iat ion (eg. environmental ef fects ) w i l l 2 cause a change in the h estimate. In general, the more.uniform the s i t e 2 the greater the h value. Even the method of ca lcu lat ion can change the 2 2 h estimate. With such l im i t a t i on s , h can almost be considered a param-eter of a f i e l d t r i a l rather than a genetic parameter since i t refers to se lect ion in a de f in i te population under de f i n i te environmental conditions 2 (Hattemer, 1963). Therefore, whenever a value i s stated for h i t must refer to a par t i cu la r population under pa r t i cu la r conditions (Falconer, 1960 and Zobel, 1961). 2 In the second case, h i s a population concept and a property of a character and the environment. It.does not measure the contr ibut ion of the genotype and the environment to the phenotype of the indiv idual (Suzuki and G r i f f i t h s , 1976). The th i rd factor to consider includes several problems associated 2 2 with h (Falkenhagen, 1972). The h model depends upon a s imp l i s t i c s t a t i s t i c a l model which i s based on the addit ive genetic ef fects model. However, there i s no b io log ica l reason to believe that most of the genetic - 9 -ef fects resu l t from the addit ive ef fects of many indiv idual genes. In f a c t , the dominance genetic variance (V^) i s rarely found to be greater than the addit ive genetic variance (V^)(even though i t may be) because of the method used in ca lcu lat ion of the components of variance. That i s , the way that breeding value and dominance deviation are defined. Kimura and Crow (1964) i l l u s t r a t e d the re l a t i ve amount of addit ive genetic variance and dominance genetic variance for one locus in d i f fe rent s i tuat ions and showed that even when most of the variance i s due to dominant gene e f f ec t s , the ca lcu lat ion of the components of variance continues to represent, rather low dominance variance except in a narrow range of gene frequency. Evaluation of 2 2 2 indiv idual h , family h and within family h together with the components of variance w i l l provide a more accurate assessment.of what gene action i s actua l ly occurr ing. For instance, i f var iat ion within fami l ies i s high and var iat ion between fami l ies i s low and i f i t i s proven that most of the phenotypic var iat ion i s due to genetic e f f ec t s , then V Q i s high (Wehrhahn, 1979). Mendelian genes may not form a representative sample of the tota l genetic material although normal Mendelian inheritance i s an important assumption in the ca lcu lat ion of the components of variance. Selection a l ter s gene frequencies which in turn cause changes in genetic parameters so that predict ion of future gains based on h estimates of parent populat-ions may not be appl icable to the o r i g ina l population. F i n a l l y , estimates of variance components often have very large standard errors since tests are usually re s t r i c ted to sampling small numbers of re la t i ves and very few environments. Reduction in standard errors would require large sample s izes. Pirchner (1969) estimated that a sample s ize f i ve times larger was required to decrease the standard error by one-half. Therefore, how-ever,useful h estimates may be in assessing genetic gain, one must be - 10 -2 aware of the assumptions made, and the complexities of h when applying or 2 discussing h values. 3. Juveni le.x Mature Correlation Juvenile x mature corre lat ion refers to the interdependence between qua l i t a t i ve or quant itat ive data co l lected at d i f ferent: interva l s during the l i f e cycle ( S z i k l a i , 1974) and depends upon the strength of genetic control of a character. It i s a useful measurement to predict the perform-ance of the same or other att r ibutes at a more advanced stage. Schmidt (1963) was one of the f i r s t workers to recognize the importance of early tes t ing. Nanson (1965, 1968, and 1970) carr ied on theoret ica l work on the value of early test ing concluding that at equal se lect ion i n t e n s i t i e s , ear ly se lect ion permits higher genetic gain than late se lect ion i f the cor re lat ion between early and late measurements of a t r a i t are greater than that t r a i t ' s 2 • • • broad sense h . An advantage of the early test i s that i t allows the evaluation of a greater number of fami l ies or provenances which would give a higher se lect ion intens i ty and further increase the genetic gain. Because trees are long- l ived organisms, forest genet ic i s t s espec ia l l y hope that juveni le x mature correlat ions would be high to.al low the results of ear ly tests to be used prof i tab ly in se lect ion work. However, i t i s important to r e s i s t the temptation of overestimating the value of early tes t s . The l i t e r a t u r e includes examples of cases both accepting and reject ing the e f f i cacy of early test ing in forest trees. For example, Toda (1964 and 1972) found that the diameter of ages less than s ix years correlated negatively with diameter at 30 years in Lar ix  l epto lep i s . (Sieb. and Zucc.) Gord. Evidence gives varying results for Pinus sy l ves t r i s L. Nanson (1968) found that growth data co l lected at age - 11 -10 correlated well with estimates made l a te r in mature stands and Pat laj (1973) calculated a pos i t ive cor re lat ion coe f f i c i en t of early to l a t e r height growth of 0.75. However, Gierytech (1974) reported low cor re la t ion between various growth characters of early and l a te r years also in P.  s y l ve s t r i s L. Four d i f fe rent pine species in Louisiana were studied over a 30-year period by Wakely (1971). His results suggested that se lect ion should not be done before age 10 and that r e l i a b i l i t y of results remained low at 15 years, not increasing appreciably; un t i l approximately age 20. Working with Pinus ponderosa Laws and P_. monticola Dougl. Steinhoff (1974) also found that evaluation of results from provenance or progeny t r i a l s was not very re.liable^at ages of less than 15 to 20 years. He said that a minimum corre lat ion coe f f i c i en t (R) of 0.70 was required for good r e l i a -b i l i t y of early test results since the coe f f i c i en t of determination (R ) would account for at least half of the var iat ion present although select ion to c u l l the poorest and favour the best fami l ies or ind iv iduals could begin e a r l i e r (approximately age 15). Namkoong and Conkle (1976) showed a decline in the ra t i o of family var iat ion to tota l var iat ion between ages f i ve to eight years in P_. ponderosa. Laws. Such rapid changes in the re l a t i ve s i ze of variance components are consistent with low cor re lat ion between early height and those taken at l a te r ages. Results with P_. v i rg in iana M i l l . have shown strong corre lat ion between heights at very early ages and l a te r years. Genys and Forbes (1973) found that at least 64 percent of the best s t ra ins for height at age 16 could have been predicted at age seven. Meier and Goggans (1977) also found pos i t ive correlat ions between annual growth and tota l height although i t was not s i gn i f i can t un t i l age eight suggesting that select ion should not proceed - 12 -unt i l the material was at least age eight. In Douglas-f ir, Namkoong e t a l _ . . (1972). studied age related var iat ion in genetic control of height growth of a 53 year old p lantat ion. Results indicated that the genetic var iat ion changed during the d i f fe rent stages of the l i f e cycle ( juven i le , early reproductive and l a t e r periods). Additive genetic variance (and therefore h ) was highest from the e a r l i e r l i f e stages of establishment to the early reproductive years. Af ter th i s period the re la t i ve amount of decreased as trees.reached dominance and height growth became more uniform. The rapid change in the r e l a t i ve s ize of variance components was s im i la r to that observed in P_. pohderosa Laws (Namkoong and Conkle, 1976) and suggests low corre lat ion between early and l a te r height measurements. Work in human genetics has also pointed out that trends evident in early l i f e stages may not be correlated to those at other stages. Bock et al_. (1963) 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 accounts for a component for prepuberta l growth which continues in reduced degree un t i l maturity and the second accounts for the contr ibution of the adolescent spurt. From the var iety of evidence presented, i t can be concluded that the r e l i a b i l i t y of ear ly tests d i f f e r s between species as well as within species. Generally, pos i t ive juveni le x mature corre lat ions were observed in older provenance experiments where material was genet ica l ly very diverse. But when material was not very diverse, results were not as dependable (eg. for progeny te s t s ) . Longer test periods seem to be required for progeny tests with s i gn i f i can t family x s i t e interact ion or for tests that do. not contain a wide range of va r i a t ion . Also much experimentation on juveni le - 13 -x mature c o r r e l a t i o n assumes t ha t the r e l a t i v e ra te s o f growth f o r d i f f e r -ent genotypes do not vary g r e a t l y w i th age. Th is assumption has i t s l i m i t -a t i on s s i nce the examples g iven above show tha t pa t te rns o f growth r a t e change as organisms mature. Given t h i s ev idence, i t i s c l e a r t ha t f u r t h e r exper imentat ion i s needed to de f i ne the j u v e n i l e x mature c o r r e l a t i o n of f o r e s t t r e e s . - 14 -MATERIALS AND METHODS Seedlings representing the open-pollinated progeny of 48 provenances from the f i r s t and second I.U.F.R.O. seed co l lect ions in 1966 and 1968 established in a provenance-progeny test at the Univers ity of B r i t i s h Columbia Research Forest in Haney in 1971 provided the material for th i s study. Details on the location of the provenances are given in Figure 1 and Table 1. The test was designed as a randomized complete block repl icated three times, each provenance represented by eight fami l ies including f i ve seedlings per family per block. The mother trees may be considered as representing the natural var iat ion within the stand and as non-related since there was no select ion applied and trees were separated by an interva l that exceeded normal seed- fa l l distance (B i ro t , 1974 and Lines, 1967). Further information on experimental layout and the test s i t e are given by Kvestich (1976). The trees in the plantation were measured to the nearest centimetre in September, 1978 for 1976, 1977 and 1978 height increment as well as 1978 tota l height. Growth data for e a r l i e r years had been previously co l lected in 1973 and 1975 (Kvestich, 1976). Very l i t t l e morta l i ty occurred since the 1975 measurements were taken with the average morta l i ty to May 1978 being 9.8%. Only four provenances exhibited morta l i ty f igures greater than 15%. These were provenance 99 (seed zone 1) at 29.21, provenance 115 (seed zone 2) at 40.8%, provenance 103 (seed zone 3) at 34.2% and provenance 66 (seed zone 7) at 24.2%. Because of the high su r v i va l , competition ef fects within the plantation were minimal for a l l years, each tree having approximately the same avai lable growing space. However, effects of competition are - 15 -becoming more important as the trees mature and plans are being made to th in the test during the summer of 1979. S t a t i s t i c a l analyses were carr ied on for indiv idual provenances and for provenances arranged in groups according to seed zones. Grouping was made according to four categories: seed zones 1, seed zone 2, seed zone 3 and a fourth representing i n t e r i o r provenances (seed zones 4, 5, 7 and 8). A l l expected mean squares were derived using a random effects model. The analysis for var iat ion between provenances was based on the fol lowing l i nea r model: FIGURE 1 TABLE 1 E levat ion, Lat i tude, Longitude, and Thousand Seed Weight (TSW) of the Douglas^fir Provenances Cl imatic Provenance No. Elev. ( f t ) Latitude .Longitude TSW Region deg. min. deg. min. la 23 Cassidy 650 49 03 123 57 9.1725 32 Duncan 200 48 45 123 45 10.2133 laa 55 Gard Station 1500 lb 90 01 a l l a KJ 1100 95 Corva l l i s 250 96 M i l l C i ty 550 99 Roseburg 900 lc 104 Ashland* 4900 111 Hawkinsville 3500 117 Big Bar 4300 124 Lower Lake 3100 2a 12 Jeune Landing 550 2b 51 Humptuli ps 450 52 Matlock 1650 53 Matlock 400 2c 67 Naselle 150 79 Prindle 1500 83 Vernonia 700 2d 87 Waldport 200 89 Coqui l le :.250 48 00 123 05 9.0224 43 05 123 34 14.2130 44 42 123 13 11.5786 44 48 122 24 11.2457 43 19 123 30 11.4927 42 05 122 39 15.8762 41 47 122 40 16.1968 40 47 123 12 16.7025 38 50 122 42 16.0461 50 27 127 27 10.0874 47 19 123 54 8.8665 47 18 123 26 10.0426 47 15 123 25. 9.9037 46 22 123 44 10.5983 45 37 122 08 11.7812 45 46 123 13 11.7173 44 24 123 52 9.5850 43 12 124 10 10.4077 * Provenances excluded from the analys is, (see page 21) TABLE 1 (continued) Cl imatic Provenance No. Elev. ( f t . ) Region 2d 91 Brookings 92 Burnt Woods* 1000 1100, 2e 115 Areata* 2900 3a 29 Caycuse . 700 3b 25 Squamish 27 Chi l l iwack 40 Darrington 42 Perry Creek* 50 3000 500 2000 3c 60 Denny Creek 61 Gold Bar* 73 Cougar 76 Packwood 1800 400 1650 2150 3d 86 Pine Grove* 97 Detroit 101 Oakridge 103 Wolfcreek* 2400 1600 2900 1400 3e 112 Dunsmuir 113 Burney* 3300 3350 4 6 K l i n ak l i n i 10 5b 46 Twi.sp 64 Cle El urn 77 Glenwood 80 Wi l lard 2600 2100 1600 1650 Latitude Longitude TSW deg. min. deg. min. 42 07 124 12 12.0091 44 36 123 42 9.8074 40 54 123 46 13.9353 48 55 124 26 10.1977 49 47 123 09 9.0917 49 06 121 42 7.3223 48 16 121 •-38 10.9612 48 03 121 28 10.9612 47 24 121 32 10.9442 47 51 121 39 9.6996 46 05 122 18 10.9332 46 34 , 121 40 8.4889 45 06 121 23 14.2043 44 44 122 10 11.1186 43 54 122 22 12.4379 42 41 123 23 12.5340 41 12 122 18 15.1707 41 05 121 39 18.0451 51 01 125 36 10.3733 48 23 120 24 13.2947 47 13 121 07 10.3615 46 00 121 10 13.5443 45 48 121 41 12.2576 * Provenances excluded from the analysis, (see page 21) TABLE 1 (continued) Cl imatic Provenance No. Elev. ( f t . ) Lati tilde Longitude TSW Region deg. min. deg. min. 7a 10 Revel stoke 2000 51 00 118 12 8.8204 18 Salmon Arm 1550 50 44 119 13 8.7206 7b 28 Nelson 2700 49 30 117 16 9.2489 16 Spokane 2000 47 47 117 12 13.5056 8 11 Golden 2700 51 23 117 00 9.0777 - 20 -Y i j km = H + P i + B j - (PB)^- + F / P k ( i ) + ( B x F / P ) J k ( i ) + E m(ijk) w h e r e Y i j km the mean measurement of the tree in the k *^1 family in the j t h block mean of a l l trees over a l l f ami l i e s , blocks and provenances provenance e f fect B. = block e f fect ( P B ) 1 d F/P k( i ) = provenance x block interact ion = family within provenance (BxF/P)j k^.^ = block x family within provenance interact ion (sampling error) ^m(ijk) = e x P e r i m e n ' t a l error A s t a t i s t i c a l test for the s ign i f icance of the highest order interact ion (BxF/P) was performed by using ANOVAR, a computer package ava i lab le from the Univers ity of B r i t i s h Columbia Computing Centre, and was found to be i n s i gn i f i can t at the one percent l e v e l . Therefore, a revised model was used for the genetic analysis in the form: Y. .. = M + P. + B. + (PB). . + F/P. / .> + E ,. .. s i j km ** i j i j ' k ( i ) m ( i j k ) where E^-j includes both the sampling and experimental error. This pooled error term increases the precis ion of the F tes t s . The expected mean squares of the analysis of variance are given in Table 2. - 21 -TABLE 2 Analys is of Variance and Expected Mean Squares f o r Analys is of Between Provenance Var ia t ion Source of Var ia t ion df Provenance P - l Block b-1 Provenance x Block Family/Provenance Error (Pooled) (p-D(b-l) P'(f-D P(b-l)(f-l) + bpf (t -1) Expected Mean Squares V e p + b t V p / p + f t V p B + b f tV r ep + f t V p B + bftVg V + btV ep + f tV PB F/P V. ep where p = No. b = No. f, = No. t = No. V P = v a n VPB = van" V B = van' F/P = van' V n = res i ep ance due to block x provenance i n te rac t i on ance between blocks ance between fami l ie s with in provenance dua l variance (pooled) - 22 -The analysis for within provenance var iat ion was based on the fol low-ing 1inear model: Y i j k + B i + F j + ( B F ) i j + E k ( i j ) where Y... = the mean measurement of the tree in the family in the i jk i t h block y = mean of a l l fami l ies over a l l blocks B. = block e f fect Fj = family e f fect (BF)..j = block x family interact ion F-k(ij) = experimental error. Table 3'rishows the appropriate analysis of variance showing the expected mean squares. - 23 -TABLE 3 Analys is of Variance and Expected Mean Squares f o r Analys is of Within Provenance Var i a t ion Source of df Expected Mean Squares Var ia t ion Block b-1 Family f-1 V g + t V g F + b tV p Block x Family ( b - l ) ( f - l ) V g + t V B R Error b f ( t - l ) V g where Vp = variance between fami l i e s V D r - = variance due to block x family i n te rac t i on Dr V = res idua l variance e - 24 -I f the interact ion term was found to be i n s i g n i f i c an t , the model was s imp l i f i ed to include the e f fect of block x family interact ion in the pooled error term. Corresponding s imp l i f i c a t i on of the analysis of variance and expected mean squares was also done. The analysis of variance was performed by least squares and maximum l i ke l ihood procedures to obtain unbiased estimates of the constants and the sums of squares for the tests of s ign i f icance (Harvey, 1975). This was required since the designrwas unbalanced due to missing trees caused by morta l i ty. The procedure u t i l i z e d the Mixed Model Least-Squares and Maxi-mum Likel ihood computer programs LSMLGP and LSML76 of Harvey (1968 and 1976). A l im i ta t i on of these programs was using a maximum of ten provenances for each seed zone grouping..; ^ Provenances were eliminated i f the data included missing subclasses ( for example, missing fami l ies with in blocks). Further omissions were made to obtain a sampling of provenances throughout each seed zone range. Omitted provenances are marked in Table 1. The programs were designed for use in animal breeding research and compute various parameters including least-squares analysis of variance, estimates of variance and 2 covariance components, estimates of h and i t s standard error as well as genetic, phenotypic and environmental corre lat ions . Assumptions made for the analysis of variance included momogeneous variances, independent errors and normal d i s t r i bu t i on of observations. Through p lo t t ing of the data, the homogeneity of variances and normality of observations appeared to be v a l i d . The assumption of independence of errors was assumed correct. Assumptions associated with the genetic interpretat ion of the compon-ents of variance included those mentioned in the l i t e r a t u r e review. The - 25 -assumptions that re lat ives are not inbred and are random members of a non-inbred population are v a l i d . The maternal ef fects were tested by estimat-ing the corre lat ion coe f f i c i en t between thousand seed weight (Yao, 1971) and tota l height in the f i r s t and second years. Deviation from these assumptions i s most l i k e l y with respect to the re lat ionsh ip of indiv iduals within a family. I f the open-pollinated seeds had only a few pollen parents then they would be more c lose ly related than ha l f - s ibs and the estimates of 2 h would be biased upward. The genetic components of variance are estimated from the analysis of variance as fo l lows: a) variance among outcrossed ha l f - s i b family means (Vp and Vp/p) i s the covariance of ha l f - s ibs and estimates 1/4 V^.* b) variance within outcrossed half-s ib..famil ies (V .)estimates cp 3/4 V A + V,. The analysis of open-pollinated progenies can therefore only y i e l d information about addit ive genetic variance since the dominant ef fects are confounded in \ T . ep There are some expected problems from the analys is . The number of fami l ies per provenance i s rather small (n = 8) leading to possible sampling errors for males and females selected from the population (B i rot , 1974). Also the number of open-pollinated progeny for each family is low (n ^ 5). Such small sample sizes may lead to large sampling errors (Owino, 1977). F i n a l l y , genetic variance estimates based on data from a s ingle * The standard error of V f l was calculated from the formula derived by Anderson and Bancroft (1952). - 26 -environment may be biased upward because of confounding of addit ive genetic effects with the interact ion of addit ive genetic ef fects and the par t i cu la r environment in which the experiment i s conducted (Gardner, 1963). H e r i t a b i l i t y estimates based on indiv idual se lect ion derived from the analysis of variance were calculated from the formulas: a) for seed zones: variance among fami l ies = covariance ha l f - s ibs cov HS = h V A VA = 4 VF/P then h = — Vep + VF/P where i s the expected mean square for the pooled B x F/P and within plot va r ia t ion , b) for ind iv idual provenances assuming s i g n i f i c an t block x family interact ion : V A - 4 V F 2 4 VF then h = and V e + VBxF + V F for ind iv idual provenances assuming non-s ign i f icant block x family i n t e r ac t i on : '* V + V ep F - 27 -where V i s the expected mean square for the pooled BxF and within p lot va r ia t ion . The estimation of juveni le x mature correlat ions was done by simple correlat ions between progeny annual growth and to ta l height and simple correlat ions between tota l heights at various ages. Progeny annual growth was computed from the average of 1975, 1976, 1977 and 1978 height increments. The l a s t four years were chosen since a comparison of the standard deviations of height increments for each of the years using the coe f f i c i en t of var iat ion (Steel and Tor r ie , 1960) indicated that the v a r i a b i l i t y s t ab i l i z ed by 1975. - 28 -RESULTS AND DISCUSSION I n i t i a l analysis of the var iat ion in height growth for a l l seed zones i s summarized in Table 4 giving the overal l average eight year height growth for the provenance-progeny t r i a l . Large differences for the 1978 tota l height are shown between seed zones. The results for a l l provenances within seed zones are presented in Tables 5, 6 and 7. S i gn i f i cant var iat ion between provenances within seed zones is also found. The large range in provenance means indicates considerable genetic va r ia t ion . This high v a r i -a b i l i t y in height growth suggests that substantial gains can be made through select ing the most desirable provenances. These f indings are consistent with those of Kvestich (1976) that indicated s i gn i f i c an t between and within provenance var iat ion for 1972 to 1975 height increment and for 1975 to ta l height. Least-squares analysis of variance for 1972 to 1978 to ta l height in seed zones 1, 2 and 3 and the i n t e r i o r provenances reveal s i g n i f i c an t d i f f e r -ences between blocks, provenances and fami l ies within provenances as well as s i gn i f i c an t block x provenance interact ion for a l l seed zones (Appendix I). S i gn i f i cant family within provenance var iat ion gives the opportunity for se lect ion of the best fami l ies in the best provenances. A l l tests are s t a t i s t i c a l l y s i gn i f i c an t at the 99% l e v e l . The only exception i s the var iat ion between blocks in the seed zone 3 provenances which is s i gn i f i c an t at the 95% level only for 1972, 1975 and 1976 and not s i gn i f i c an t for 1977 and 1978 to ta l height. 1. Genetic Variat ion in Height Growth Studies of the genetic var iat ion in the Douglas-fir population were - 29 -TABLE 4 Total 1978 Height Differences Between Seed Zones Seed Zone Average 1978 Standard Total Height Deviation (cm.) (cm.) 1 284.2 97.7 2 347.8 101.4 3 313.4 96.2 4 273.9 80.1 5 253.3 86.8 7 , 200.4 68.4 8 164.7 76.0 Average 297.2 104.9 - 30 -Seed Zone TABLE 5 Maximum, Minimum, Average and Standard Deviation of 1978 Total Height in Seed Zones'1 and 2 (cm.) Prov. Max. Min. Average Standard Deviation 23 472.0 100.0 288.6 93.1 32 480.0 107.0 345.3 73.4 55 493.0 70.0 319.5 77.7 90 492.0 31.0 297.8 102.5 95 500.0 151.0 327.7 76.3 96 570.0 . 175.0 372.1 72.7 99 470.0 92.0 281.8 88.7 104 405.0 52.0 211.3 . 82.9 111 557.0 66.0 227.2 77.6 117 366.0 71.0 211.8 64.7 124 411.0 68.0 238.1 88.4 12 530.0 74.0 313.0 92.2 51 564.0 140.0 335.7 87.7 52 590.0 134.0 380.4 100.3 53 580.0 115.0 387.7 101.5 67 570.0 115.0 377.6 104.8 79 503.0 37.0 323.1 80.9 83 542.0 • 158.0 358.2 84.5 87 533.0 133.0 363.7 82.0 89 620.0 68.0 383.4 109.1 91 520.0 165.0 370.8 83.0 92 585.0 91.0 315.6 106.9 115 401.0 84.0 223.7 77.3 - 31 -TABLE 6 Maximum, Minimum, Average an of 1978 Total Height (cm.) Prov. Max. Min. 29 458.0 72.0 25 605.0 133.0 27 520.0 125.0 40 475.0 100.0 42 575.0 174.0 60 500.0 145.0 61 500.0 149.0 73 510.0 66.0 76 475.0 149.0 86 410.0 57.0 97 470.0 153.0 103 378.0 61.0 101 485.0 126.0 112 425.0 84.0 113 405.0 48.0 Standard Deviation n Seed Zone 3 Average Standard Deviation 324.3 68.5 381.3 86.1 369.3 89.4 307.9 79.7 384.3 94.1 299.2 92.5 341.9 75.9 330.0 87.9 315.5 73.2 237.4 86.2 331.9. 67.7 228.5 86.0 314.9 82.0 254.2 100.5 237.9 85.0 - 32 -TABLE 7 Maximum, Minimum, Average and Standard Deviation of 1978 Total Height in the Inter ior Provenances (cm.) Zone Prov. Max. Min. Average Standard De\ 4 6 434.0 78.0 273.9 80.1 5 46 368.0 60.0 192.9 63.7 64 420.0 58.0 263.0 88.5 77 490.0 • , 45.0 269.1 95.9 > 80 442.0 115.0 287.2 64.2 7 10 354.0 87.0 192.0 63.4 18 330.0 63.0 183.6 59.1 28 384.0 78.0 237.7 72.0 66 330.0 49.0 185.3 64.0 8 11 331.0 37.0 164.7 76.0 - 33 -performed to evaluate the o r i g i n of the large amount of observed phenotypic va r i a t i on in height growth. Var ia t ion was studied at two l e v e l s ; with in provenance analyzing provenances i n d i v i d u a l l y and between provenance ana lyz -ing provenances grouped according to seed zone groupings. The ana lys i s f o r va r i a t i on with in provenances i s not included because the r e l i a b i l i t y of the resu l t s fo r several sample provenances was found to be low due to large sampling e r r o r s . Greater e f f o r t was, there fore , placed in the ana lys i s of pooled information (that i s , between provenance v a r i a t i o n ) . The large increase in sample s i ze s u b s t a n t i a l l y improved a l l tes t est imates. Appendix II gives the components of variance fo r between provenance va r i a t i on fo r a l l seed zone groupings. In seed zone 1 the component f o r provenance var iance (Vp) i s approximately f i v e times l a rger than the com-ponent fo r fami ly with in provenance (Vp^ p ) , accounting fo r an average of 26% of the to ta l phenotypic va r i a t i on from 1972 to 1978. A gradual increas ing trend f o r V p from 14.43% in 1972 -to 32.67% in 1978 i s shown. A trend in the opposite d i r e c t i o n i s noted f o r Vp^p which dec l ined from 5.91% to 4.54% although the components of var iance were s l i g h t l y higher in 1975 and 1976. The smal lest amount of phenotypic va r i a t i on i s due to the component of variance f o r blocks (V g ) which accounted fo r only about 1%. This component appears to be more important as the te s t ages. The amount of va r i a t i on accounted f o r by block x provenance i n te rac t i on (Vp xg) shows a s t ead i l y decreasing trend from 18.78% in 1972 to 7.21% in 1978. The greatest contr ibut ion to phenotypic va r i a t i on i s made by the pooled e r ro r variance (V e p) (50% to 60%). Again, there appears to be a decreasing trend in V from 1972 to 1978. The components of variance f o r seed zone 2 suggest a d i f f e r e n t pattern of v a r i a t i o n . The increas ing trend in V p and decreasing trend in - 34 -Vp^p i s s im i l a r to that for the provenances in seed zone 1. However, the variance component Vp i s less than Vp^ p for a l l years. In seed zone 2, Vp i s about one-half Vp^p on average (3.88% compared to 8.07%) although in 1978 Vp and Vp^ p are almost equal. The Vp^p i s s l i g h t l y higher in seed zone 2 in re la t ion to seed zone 1, showing a decreasing trend. S im i l a r l y , the components Vp xg and V are also higher for seed zone 2; the average amount of var iat ion in V p x B i s 8.07% and 70.98% for the pooled error var ia t ion . Variance due to pooled error remains constant throughout the test period. Observations of in teres t concerning the variance components of seed zone 3 include that Vp and Vp^p are nearly equal in 1972 with Vp becoming about twice Vp^p by 1978. On average, V p i s greater than Vp^p but not by the f i v e - f o l d degree observed in seed zone 1. The components of variance for blocks are negative (represented by zeros) for a l l years. Variance due to block x provenance interact ion accounts for greater var iat ion than in e i ther seed zone 1 or seed zone 2, averaging about 25%. The amount of var iat ion in V i s comparable t o seed zone 1, remaining constant at an average of 56%. The i n t e r i o r provenances show variance patterns s im i l a r to seed zone 1 for a l l components except Vp xg which accounts for about hal f as much var iat ion as seed zone 1. However, Vp xg remains approximately constant in the i n t e r i o r provenances while in seed zone 1 i t exhib its a decl in ing yearly trend. Again Vp is about f i ve times Vp / p . The variance component Vg accounts for the least and the variance component V gp accounts for the greatest amount of tota l va r i a t ion . No trends for increasing or decreasing component s i ze are shown. The preceding discussion has shown that the patterns of var iat ion d i f f e r appreciably between the Douglas-f ir populations studied, showing differences in the r e l a t i ve s i z e s ' o f some of the components of variance. In a l l seed zones, over half of the tota l phenotypic var iat ion (53.68% to 62.32%) i s concentrated in the t rees-with in-p lots term, V ^ . This pooled error expresses various e f f ec t s ; genetic (includes 3/4 + Vp with the assumption of ha l f - s ibs ) and environmental (measuring e r ror , e t c . ) . Because th i s value does not d i f f e r substant ia l ly between seed zones, i t may be assumed that ef fects such as measuring errors and V^ also remain constant between seed zones. The within provenance variance, which i s in fact the difference among fami l ies within provenance, also does not vary s i g n i f i -cantly between the seed zones accounting for 5.12% to 8.07% of the to ta l va r ia t ion . However, the re lat ionship of Vp^p with respect to Vp varies among the four groups. For seed zone 1 and the i n t e r i o r provenances, Vp i s greater than Vpyp by a magnitude of approximately f i ve with V p accounting for from 13.15% to 25.85% of the va r ia t ion . This magnitude i s reduced to approx-imately a two-fold difference for seed zone 3. The re lat ionsh ip is reversed for seed zone 2 where the part of the var iat ion due to provenance ef fect i s less than the percent due to family within provenance va r i a t i on . I f se lect ion i s carr ied out at Haney in the future, the approach w i l l be governed by the differences in the re l a t i ve sizes of Vp and Vp^p. In cases where Vp i s greater than Vpyp, provenance select ion for increased tota l height i s indicated; in cases where V p i s less than Vp^ p , family se lect ion would be a more e f fect i ve strategy. The above differences in variance patterns are caused by differences in adaptation of the provenance mater ia l . Haney.represents a low elevation s i t e (about 400 feet) within seed zone 3. - 36 -The provenances from seed zone 1 and the i n t e r i o r zone represent trees adapted to very d i f f e r e n t condit ions than the Haney area leading to the greatest d i f ferences in performance to be evident at the provenance l e v e l . These provenances a l so exh ib i ted the greatest d i f fe rences between Vp and Vpyp. There s t i l l may be s i g n i f i c a n t family with in provenance va r i a t i on in these provenances but i t i s not revealed at the Haney s i t e . The provenances in seed zone 3 expressed a reduced d i f fe rence between the magnitude of Vp and Vp^p. These provenances occur in the same seed zone as Haney but represent high e levat ion locat ions (average e levat ion 1895 feet ) not well adapted to the te s t s i t e . The provenances of seed zone 2 represent low e levat ions most s i m i l a r to Haney (average e levat ion 685 feet ) and hence are most adapted to Haney cond i t ions . The above resu l t s ind ica te that in the more adapted provenances, Vpyp and V g p are the dominant contr ibutors to the phenotypic v a r i a t i o n . For less adapted provenances, Vp i s of greater importance and becomes greater than V p ^ . The overa l l r e su l t s of the genetic va r i a t i on between the various Doug las - f i r populations suggest that a combination of provenance, family with in provenance and w i th in -p l o t s e l ec t i on would y i e l d the greatest increase in height growth. The amount of improvement poss ib le f o r a l l s e l ec t i on schemes i s subject to the leve l of h . D i f ferences in the va r i a t i on with in populations w i l l a lso in f luence the e f f e c t of se l ec t i on in fo res t tree improvement. F luctuat ing estimates of V p x B (7.42% for i n t e r i o r to. 25.33% for seed zone 3 provenances) averaging a rather high 15.67% for the four areas combined ind icates s i g n i f i c a n t genotype x environment i n t e r a c t i o n . I dea l l y , the s i t e chosen f o r a provenance of progeny tes t should be r e l a t i v e l y homo-geneous leading to l i t t l e or no block x provenance i n t e r a c t i o n . The Haney - 37 -s i t e was selected for i t s r e l a t i ve homogeniety. However, other factors such as an invasion of grass from' a nearby experiment has increased the differences between blocks. S i gn i f i cant genotype x environment i n te rac t -ion suggests that se lect ion from provenances and fami l ies for increased height growth, for a l l locations may not be possible,, and indicates the need for a greater number of tests in d i f fe rent s i tes to further i n v e s t i -gate genotype x environment e f fec t s . That i s , more than one seed orchard would be required to provide improved seed for alT l o c a l i t i e s . I t should be noted that the re l a t i ve s ize of V p x g i s diminishing for two regions ind icat ing that the genotype x environment effect.may be decreasing in importance. 2. Estimating Addit ive Genetic Variance (V^) <"• The ca lcu lat ion of V^, according to the re lat ionsh ip Vp/p = depends upon several assumptions already mentioned on page s ix of the l i t e r a tu re review. The assumptions that the re la t i ves are not inbred and are random members of a non-inbred population have been shown to be correct. Further invest igat ion on the assumption concerning maternal ef fects was necessary since the reported results re ly on development in the juveni le stage. The existence of maternal effects, can have s i g n i f i -cant consequences on estimated parameters. Maternal ef fects were estimated using the correlat ion, coe f f i c i en t (B i rot , 1976) between thousand seed weight (TSW) and to ta l height in 1972 and 1973. Provenances in ^seed zones 1 and 2 were analyzed. The amount of var iat ion due to maternal effects: in seed zone 1 was 2.29% and 2.61% respectively for 1972 and 1973 growth. The figures for seed zone 2 were - 38 -3.14% for 1972 and 4.16% for 1973. The very low coef f i c ient s of determin-2 ation (R ) for these provenances suggested that analysis did not require to be continued for the remaining provenances. On the basis of the results for seed zones 1 and 2 .the assumption of no maternal ef fects for a l l provenances and progenies was accepted, although the large amount of var iat ion in TSW within provenances reported by Yao (1971) suggests that TSW may s i g n i f i c an t l y influence height growth. The addit ive genetic variance, i t s standard error and expressed as a percent of tota l phenotypic. var iat ion for to ta l height (Appendix II) are given in Tables 8 and 9. Rather a high percentage of tota l height va r i a t -ion i s accounted for by with average estimates ranging from a low of 21.80% for seed z o n e ! to a high of 32.27% for seed,zone 2. The component for remains r e l a t i v e l y constant between the years accounting for an average of 24.74% of the tota l var iat ion in height for a l l regions combined. These estimates of are lower than (approximately one-half) the estimates obtained by B i rot (1976) who analyzed f i r s t and second year height growth for a sample of the I.U.F.R.0. co l l ec t i on from Washington State grown in France. Several of the provenances tested at Haney were included in his study. The estimates of for f i r s t and second year tota l heights were 54% and 42% respect ively. Differences in the estimates of B i rot and those presented in th i s paper may be due to differences in the model, the test s i t e , and the test mater ia l . B i r o t 1 s analysis did not take account of block ef fects causing.differences due to blocks to be incorporated in.other variance components. The test s i t e he used was very homogeneous, leading to decreased environmental var iat ion and increased estimates of genetic parameters. The test material included d i f fe rent provenances which could exh ib i t s l i g h t l y d i f fe rent patterns of var iat ion and only f i r s t and second - 39 -year height growth were studies. I t w i l l be interest ing to continue compar-isons of the pa ra l l e l provenance-progeny tests between the two locations in the future. Examination of the r a t i o of the standard error of over mu l t i -p l ied by 100 (S.E. (V A)/V A ) 100) given in Tables 8 and 9 reveal differences between seed zones. The average rat ios for seed zone 1 and the i n t e r i o r provenances are very s im i la r at 14.40% and 15.21% respect ive ly. Results for seed zones 2 and 3 are also comparable to each other averaging lower at values of 6.86% and 7.21%. This s i m i l a r i t y of the rat ios (S.E. (V A )/ V A) 100) for pairs of provenance groupings i s s i gn i f i c an t since i t supports the explanation proposed e a r l i e r that the more adapted provenances in seed zones 2 and 3 exh ib i t lower var iat ion in results while those less adapted provenances from seed zone 1 and the i n t e r i o r express higher v a r i a b i l i t y . 2 3. Estimation of H e r i t a b i l i t y (h ) 2 The importance of h in the estimation of genetic gain has been pre-2 viously out l ined. The h estimates presented in th is thesis were ca lcu-lated according to the relat ionships given on page 23 of the Material and Methods section and were based on ident ica l assumptions to those given for V A > The assumption of open-pollinated progeny representing ha l f - s ibs i s considered 2 accurate since the h estimates are a l l less than one. 2 Estimates of h according to seed zones for to ta l height growth given in Table 10 indicate that there is moderate control of the var iat ion found in th is cha rac te r i s t i c . H e r i t a b i l i t y estimates range from 0.28 for 1978 tota l height in the i n t e r i o r provenances to 0.52 for 1972 tota l height in the seed zone 2 provenances and average 0.38 over a l l years and regions. - 40 -TABLE 8 Add i t i ve Genetic Variance ( V A ) , Standard Er as a Percent of Tota l Phenotypic Va r i a t i on Location Year Seed Zone 1 S.E.(VA) VA + S.E. 100(—' — — ) Pheno. VA Var. % 1972 .41.121 + 6.01 14.61 23.64 1973 135.26 + 19.67 14.54 20.96 1974 288.35 + 42.05 14.58 20.80 1975 594.67 + 77.06 12.96 25.44 1976 945.28 + 130.88 13.85 24.08 1977 1299.47 + 194.86 15.00 19.52 1978 1911.97 + 291.29 15.24 18.16 Average 14.40 21.80 (S .E . ) , ( S .E . (V A ) /V A )100 and V A Total Height: Seed Zones 1 and 2 Seed Zone 2 S.E. (V A ) VA + S.E. 100(- -) Pheno. v A Var. % 94.86 + 5.82 6.14 43.40 275.29 + 17.76 6.45 36.48 485.14 + 33.12 6.83 32.28 694.04 + 50.24 7.24 28.52 1209.20 + 86.94. 7.19 27.72 2040.77 + 142.35 6.98 29.08 2854.15 + 203.66 7.14 28.40 6.86 32.27 Es t imate s of V A are a l l s i gn i f i can t at the 0.01 l e v e l . - 41 -TABLE 9 Addit ive Genetic Variance (V A ) , Standard Errors (S.E.), (S.E.(V A )/V A)100 and V A as a Percent of Total Phenotypic Var iat ion for Total Height: Seed Zone 3 and Inter ior Provenances Year Location Seed Zone 3 Inter ior Provenances v A + S . E . • 100( — ) Pheno. VA + S . E . 100(- -) Pheno. v A Var. % . Var. % 1972 52.30 1+ 3.49 6.67 26.76 23.31 + 3.27 14.01 25.24 1973 148.16 + 10.52 7.10 20.76 102.18 + 13.64 13.35 25.72 1974 302.67 + 21.43 7.08 21.32 232.42 + 30.88 13.29 26.96 1975 470.87 + 34.97 7.43 20.92 392.19 + 55.64 14.19 24.32 1976 711.38 + 55.79 7.84 18.56 ' 580.12 + 89.14 15.37 20.68 1977 1302.22 + 96.06 7.38 21.00 940.67 + 112.72 12.40 19.12 1978 2209.26 + 153.08 6.93 25.12 1308.30 + 217.02 16.59 17.52 Average 7.21 22.07 15.21 22.80 •'•Estimates of V A are a l l s i gn i f i c an t at the 0.01 l e v e l . o - 42 -2 For seed zone 1, h remains r e l a t i v e l y constant varying only s l i g h t l y from 2 1972 to 1978. There i s a decreasing trend for h for seed zone 2 from 1972 to 1975 which s t ab i l i ze s to change very l i t t l e in the f i n a l three years tested. Provenances in seed zone 3 also exh ib i t a steady decrease 2 in h continuing to 1976. Estimates for 1977.and 1978 are increasing in magnitude. H e r i t a b i l i t y estimates for 1972, 1973 and ,1974 are increasing for the i n t e r i o r provenances, although values are decreasing in the l a t e r years. H e r i t a b i l i t y declines in the early years according to the re l a t i ve 2 increase of Vp versus Vp^p. One explanation of the decline in h i s a disappearance of maternal ef fects not indicated by the .regression analysis for maternal inf luence. The r e l a t i v e l y high amount of addit ive genetic control indicated by 2 the h values suggest that there are opportunities for s i gn i f i c an t improve-ment by se lect ion in provenance and progeny tests for Douglas-f ir. Never-theless, the differences in genetic variance between regions for Vp and Vp^p and the resu l t ing differences in genetic parameters may cause varying responses to se lect ion between regions. This p o s s i b i l i t y should be further investigated by studying the within provenance variance i nd i v i dua l l y . B i rot (1976) found s i gn i f i can t differences in genetic variance between trees within provenance as well as between provenances. H e r i t a b i l i t y estimates for indiv idual provenances at the Haney s i t e had very large standard errors due to small sample s izes. Appendix IV gives estimates of 2 h and standard errors for several indiv idual provenances. Although these values have low r e l i a b i l i t y , many are s im i la r in magnitude to estimates derived from seed zone analys is . Therefore, emphasis was again placed on pooled data for'provenances within seed zones. - 43 -TABLE 10 H e r i t a b i l i t y (h 2 ) and Standard Error (S.E.) for Total Height Based on Seed Zones Location Year Seed Zone 1 Seed Zone 2 Seed Zone 3 In ter ior h 2 ± S.E. h 2 ± S.E. h 2 ± S.E. h 2 ± S.E. 1972 0.36 ± 0.10 0.52 ± 0 . 1 1 0.42 ±0 .10 0.39 ±0 .10 1973. 0.36 ± 0.10 0.46 ± 0.11 0.37 ± 0.10 0.44 ± 0.10 1974 0.36 ± 0.10 0.40 ± 0.10 0.37 ±0 .10 0.45 ± 0 . 1 1 -1975 0.42 ± 0.10 0.35 ± 0.10 0.33 ± 0.09 0.38 ± 0.10 1976 0.40 ± 0.10 0.36 ± 0.10 0.30 ± 0.09 0.33 ± 0.09 1977 0.34 ± 0.10 0.38 ± 0.10 0.34 0.09 0.31 0.09 1978 0.33 ± 0.10 0.37 ± 0.10 0.39 0.10 0.28 0.09 - 44 -2 The magnitude of h values calculated for d i f fe rent regions are in agreement with B i rot (1976) but are greater than those of Campbell (1972) 2 and Orr-Ewing and Yeh (1978). Campbell estimated h of juveni le Douglas-f i r trees between 0.08 and 0.22 for f i r s t and second year height. These low estimates were mainly due to high family x locat ion in te rac t ion . The potential for genetic improvement in seedling height growth nonetheless appeared good because of considerable genetic variance (high dominance genetic variance). Orr-Ewing and Yeh also suggested that growth was under 2 weak addit ive genetic control obtaining low estimates for h . The h e r i t -a b i l i t i e s they presented had low r e l i a b i l i t y because of the unbalanced 2 design and small s ize of the experiments. Standard errors for h based on seed zones at the Haney s i t e are consistent at approximately 0.10 for a l l years and seed zones giving a high degree of r e l i a b i l i t y to these estimates The merits of each of the se lect ion systems and combinations recommend ed according to the re la t i ve s ize of the components of variance can be com-2 2 pared on the basis of expected genetic gains using h estimates. The h values calculated here are based on family se lect ion. Another concept to be discussed here i s that i t may be more advantag-eous to calculate genetic gain on a per unit time basis.' This was suggested by Namkoong et al_. (1966) since the time to obtain gains may be a c r i t i c a l factor in - t ree improvement programs. Not only i s the interval between the i n i t i a l se lect ion and the time when the next generation i s avai lable for select ion important, but the appl icat ion of se lect ion w i l l act to change gene frequencies that in turn change estimates of genetic parameters u t i l -ized in the gain pred ict ion, making gain estimates on the basis of i n i t i a l measurements inaccurate. Although select ion has not been applied yet in - 45 -the I.U.F.R.O. experiment, Namkoong's idea should be considered in the future. 4. Juvenile x Mature Correlat ion Increases in simple cor re lat ion coef f i c ient s between tota l heights at various ages are inversely associated with the number of years between measurements (Tables 11, 12, 13 and 14). The R values are s im i la r for a l l relat ionships in a l l regions and are a l l s i gn i f i can t at the 99% l e v e l . The tota l height in 1972 and 1973 accounts for almost half of the var iat ion in 1977 and 1978 tota l height. Between 58% and 63% of the var iat ion in 1978 tota l height can be explained by var iat ion in 1974 tota l height. Var iat ion in 1975 tota l height explains approximately 75% of the var iat ion in 1978 tota l height. The corre lat ion for 1976 and 1977 versus 1978 tota l height is over 0.80. In general, results from height measurements at ear ly ages projected forward four to f i ve years can predict over 50% of height growth var iat ion in l a te r years. The strong cor re lat ion between heights at very early ages and l a te r years indicates that se lect ion can be made at ear ly ages to predict l a te r performance with minimal r i s k of losing good ind iv idua l s . Measurements should be continued to determine i f the high re lat ionsh ip of the early juveni le to l a te r juveni le growth and growth in l a te r age classes i s perpetuated. The simple correlat ions between progeny annual growth (P.A.G.) and tota l height are s i gn i f i c an t at a l l ages with the corre lat ion coe f f i c i en t 2 increasing throughout the test period. However, the R does not account for greater than 50% of the var iat ion un t i l 1975. It i s , therefore, - 46 -suggested to wait unt i l at least age f i ve or l a te r to make se lect ions. Nursery effects are not considered s i gn i f i can t in th i s test since a l l the trees received s im i la r treatment before outplanting. S im i l a r i t y in genetic parameter estimates for a l l years also indicates that early results do not vary s i g n i f i c an t l y from la te r ones. Results from the ranking of provenances according to mean 1975 and 1978 tota l height for a l l provenances are presented in Appendix V. These re i terate the close re lat ionship between early and l a te r height growth mentioned above showing that for a l l seed zones, there i s very l i t t l e change in the ranking of the best 25% of the provenances ind icat ing good r e l i a b i l i t y in select ion of the best provenances at age f i v e . The largest deviation in performance between 1975 and 1978 occurs in seed zone 2, but i t should be noted that a l l of the top seven provenances in seed zone 2 are greater than 350 centimeters i n tota l 1978 height. The performance of the remaining provenances also does not change s i g n i f i c an t l y so the removal of the poorest trees at age f i ve may be feas ib le as w e l l . Early evaluation w i l l allow for test ing of a greater number of fami l ies or provenances which w i l l lead to a higher se lect ion in tens i ty . Selection intens i ty i s another important factor in the determination of genetic gain. 5. Selection of the Best Provenances and Progenies for the Test S ite The f i na l objective of th i s study i s to se lect.the best provenances and progenies on the basis of performance of tota l height growth. From the information presented in the beginning of the Results and Discussion sect ion, provenances from seed zones 2 and 3 (Washington sources Matlock (52 and 53), Naselle (67) and Oregon sources Vernonia (83), Waldport (87), - 47 -TABLE 11 Juvenile x Mature Correlat ion: Seed Zone 1 Variables Dependent Independent R2 72 Tot. Ht. vs 73 Tot. Ht. 0.82 1 vs 74 Tot. Ht. 0.70 vs 75 Tot. Ht. 0.54 vs 76 Tot. Ht. 0.47 vs 77 Tot. Ht. 0.40 vs 78 Tot. Ht. 0.31 vs P.A.G.2 0.16 73 Tot. Ht. vs 74 Tot. Ht. 0.91 vs 75 Tot. Ht. 0.71 vs 76 Tot. Ht. 0.62 vs 77 Tot. Ht. 0.54 vs 78 Tot. Ht. 0.41 vs P.A.G. 0.22 Variables Dependent Independent R2 74 Tot. Ht. vs 75 Tot. Ht. 0.88 vs 76 Tot. Ht. 0.81 vs 77 Tot. Ht. 0.72 vs 78 Tot. Ht. 0.58 vs P.A.G. 0.35 75 Tot. Ht. vs 76 Tot. Ht. 0.95 vs 77 Tot. Ht. 0.87 vs 78 Tot.- Ht. 0.72 vs P.A.G. 0.52 76 Tot. Ht. vs 77 Tot. Ht. 0.95 vs 78 Tot. Ht. 0.83 vs P.A.G. 0.67 77 Tot. Ht. vs 78 Tot. Ht. 0.91 vs P.A.G. 0.81 78 Tot. Ht. vs P.A.G. 0.90 A l l of the simple correlat ions are s i gn i f i c an t at the 0.01 level of tes t ing. P.A.G. = progeny annual growth calculated as (75 height increment + 76 height increment + 77 height increment + 78 height increment)/4 - 48 -TABLE 12 Juvenile x Mature Corre lat ion: Seed Zone 2 Variables Variables 2 ? Dependent Independent R Dependent Independent R 72 Tot. 73 Tot. Ht. vs 73 Tot. Ht. 0.80 1 74 Tot. Ht. vs 75 Tot. Ht. 0.86 vs 74 Tot. Ht. 0.67 vs 76 Tot. Ht. 0.76 vs 75 Tot. Ht. 0.49 vs 77 Tot. Ht. 0.67 vs 76 Tot. Ht. 0.40 vs 78 Tot. Ht. 0.58 vs 77 Tot. Ht. 0.34 vs P.A.G. 0.31 vs 78 Tot. Ht. 0.29 vs P.A.G.2 0.11 75 Tot. Ht. vs 76 Tot. Ht. 0.94 vs 77 Tot. Ht. 0.87 Ht. vs 74 Tot. Ht. 0.90 vs 78 Tot. Ht. 0.77 vs 75 Tot. Ht. 0.68 vs P.A.G. 0.53 vs 76 Tot. Ht. 0.57 vs 77 Tot. Ht. 0.49 76 Tot. Ht. vs 77 Tot. Ht. 0.96 vs 78 Tot. Ht. 0.41 vs 78 Tot. Ht. 0.86 vs P.A.G. 0.18 vs P.A.G. 0.69 77 Tot. Ht. vs 78 Tot. Ht. 0.93 vs P.A.G. 0.82 78 Tot. Ht. vs P.A.G. 0.89 *A11 of the simple correlat ions are s i gn i f i can t at the 0.01 level of tes t ing. 2 P.A.G. •= progeny annual growth. - 49 -TABLE 13 Juvenile x Mature Corre lat ion: Seed Zone 3 Variables Dependent Independent Variables Dependent . Independent 72 Tot. Ht. 73 Tot. Ht. vs 73 Tot. Ht. 0.81 1 74 Tot. Ht. vs .75 Tot. Ht. vs 74 Tot. Ht. 0.72 vs 76 Tot. Ht. vs 75 Tot. Ht. 0.58 vs 77 Tot. Ht. vs 76 Tot. Ht. 0.51 vs 78 Tot. Ht. vs 77 Tot. Ht. 0.45 vs P.A.G. vs 78 Tot. P.A.G.2 Ht. 0.38 vs 0.20 75 Tot. Ht. vs vs 76 Tot. 77 Tot. Ht. Ht. vs 74 Tot. Ht. 0.92 vs 78 Tot. Ht. vs 75 Tot. Ht. 0.74 vs P.A.G. Tot.Ht vs 76 Tot. Ht. 0.66 vs 77 Tot. Ht. 0.58 76 Tot. Ht. vs 77 Tot. Ht. vs 78 Tot. Ht. 0.49 vs 78 Tot. Ht. vs P.A.G. 0.25 vs P.A.G. 0.87 0.79 0.71 0.61 0.36 0.94 0.87 0.75 0.52 0.96 0.85 0.68 77 Tot. Ht. vs 78 Tot. Ht. 0.93 vs P.A.G. 0.82 78 Tot. Ht. vs P.A.G. 0.89 A l l of the simple correlat ions are s i gn i f i can t at the 0.01 level of tes t ing. P.A.G. = progeny annual growth. - 50 -TABLE 14 Juvenile x Mature Corre lat ion: Inter ior Provenances Variables Dependent Independent Variables Dependent Independent 72 Tot. Ht. 73 Tot. Ht. vs 73 Tot. Ht. 0.83 1 74 Tot. Ht. vs 75 Tot. Ht. 0.84 vs 74 Tot. Ht. 0.73 vs 76 Tot. Ht. 0.78 vs 75 Tot. Ht. 0.57 vs 77 Tot. Ht. 0.71 vs 76 Tot. Ht. 0.51 vs 78 Tot. Ht. 0.63 vs 77 Tot. Ht. 0.47 vs P.A.G. 0.40 vs 78 Tot. Ht. 0.42 vs P.A.G.2 0.26 75 Tot. Ht. vs 76 Tot. Ht. 0.96 vs 77 Tot. Ht. 0.88 vs 74 Tot. Ht. 0.92 vs 78 Tot. Ht. 0.75 vs 75 Tot. Ht. 0.71 vs P.A.G. 0.56 vs 76 Tot. Ht. 0.64 vs 77 Tot. Ht. 0.58 76 Tot. Ht. vs 77 Tot. Ht. 0.96 vs 78 Tot. Ht. 0.51 vs 78 Tot. Ht. 0.86 vs P.A.G. 0.30 vs P.A.G. 0.71 77 Tot. Ht. vs 78 Tot. Ht. 0.94 vs P.A.G. 0.85 78 Tot. Ht. vs P.A.G. 0.92 A l l of the simple correlat ions are s i gn i f i can t at the 0.01 level of tes t ing . P.A.G. = progeny annual growth. - 51 -Coquil le (89) and Brookings (91) in seed zone 2; B r i t i s h Columbia sources Squamish (25) and Chi l l iwack (27) and Washington sources Perry Creek (42) and Gold Bar (61) in seed zone 3) produce the t a l l e s t trees at the Haney s i t e . Some provenances from seed zone 1 also exh ib i t excel lent growth (Duncan (32), B r i t i s h Columbia and M i l l City (96), Oregon). In addit ion, 2 from the information on V^, h and juveni le x mature co r re la t i on , i t i s evident that indiv idual and family se lect ion in the provenances of the best zones in 1975 w i l l y i e l d s i gn i f i can t improvement in tota l height growth. Applying these results to a Douglas-fir tree improvement program in south coastal B r i t i s h Columbia, the best provenances in seed zones 2 and 3 would be selected because of the i r s i g n i f i c an t l y greater rate of height growth. Spec i f i c indiv idual fami l ies or trees may be selected below the top provenances in the best zones and exceptional provenances in seed zone 1 would be included. Selection of some provenances in other seed zones may also be desireable to increase the genetic base of the resu l t ing breeding program and to provide the material for i n te r r ac i a l crosses to further expand the avai lable genetic va r i a t i on . The number of provenances selected would depend on the spec i f i c program objectives of the pa r t i cu la r agency involved. The response to se lec t ion , that i s , the genetic gain (R) at the Haney 2 s i t e was estimated from the formula R = i V p h (Falconer, 1960) where i i s the in tens i ty of se lect ion and V p i s the phenotypic standard deviation (Appendix VI). A se lect ion intens i ty of only one in four indiv iduals was chosen. Assuming a normal population, the theoret ica l value for i saving 25% of the best indiv iduals i s 1.23 (Falconer, 1960). Estimates of genetic gains for a tree improvement program of phenotypic se lect ion and es tab l i sh -ment of clonal orchards without progeny test ing (Squi 1 lace et al_., 1966) - 52 -are given in Table 15. Results using the r e l a t i v e l y low se lect ion intens i ty of 25% indicate genetic gains from 17.90% in 1972 to 10.96% in 1978 for a l l regions. The genetic gain estimates suggest that appreciable improvement in growth rate of Douglas-fir i s possible by merely se lect ing from the top indiv iduals from any provenance. However, the results of the seed zone and provenance performance show that even greater gains can be achieved by se lect ing the best indiv iduals from the best provenances in the best seed zones. For example, comparing the mean 1978 tota l height of provenance 53 (Matlock) at 387.7 centimetres to the mean for the whole plantation at 297.2 cen t i -metres (Tables 4 and 5) an increase in tota l height growth of almost 33% is suggested. I f the best indiv idual within provenance 53 is also selected (1978 to ta l height of 580 centimetres), a further gain of 70% above the plantation mean is indicated. TABLE 15 Estimates of Genetic Gain for Height Growth for Seed Zones 1, 2 and 3 and Inter ior Provenances Year Seed Zone 1 Seed Zone 2 Seed Zone 3 Inter ior Average % of Pop. Mean in cm. % of in cm. % of in cm. % of in cm. % of pop. mean pop. mean pop. mean pop. mean 1972 4.74 15.72 8.61 23.10 5.75 16.96 3.69; 15.82 17.90 1973 8.55 15.14 13.83 18.89 9.16 13.81 8.25 18.75 16.65 1974 12.53 14.64 17.08 15.09 13.04 12.99 12.65 18.83 15.39 1975 19.02 15.74 19.06 11.99 15.26 10.80 14.94 15.27 13.45 1976 23.82 14.50 25.69 11.88 18.02 9.40 17.15 12.83 12.15 1977 25.85 11.62 34.09 11.84 26.02 10.15 20.88 11.73 11.34 1978 30.93 10.88 40.24 11.22 36.27 11.31 23.51 10.41 10.96 co - 54 -SUMMARY The genetic var iat ion in height growth of juveni le Douglas-f ir trees in the I.U.F.R.O. provenance-progeny test at Haney varied between seed zones . in the re l a t i ve sizes of V p and Vpy p. For two of the four regions, V p was greater than Vpy p by a magnitude of f i ve (seed zone 1 and i n te r i o r ) and d i f fered by a two-fold margin for seed zone 3. The reverse re lat ionship was shown for seed zone 2 where V p was less than Vpy p for a l l years. The con-clusion was made that for provenances most adapted to the Haney s i t e , Vpy p and V are the dominant contributors to the phenotypic var iat ion and that for less adapted provenances, V p i s of greater importance becoming greater than Vp^ p. By far the greatest component of variance was due to the pooled error term V (which includes the var iat ion of indiv idual trees within fami l ies ) accounting for 53% to 62% of the tota l phenotypic va r i a t i on . Variat ion due to V p and V p x g were the next largest sources. The components for family within provenance cons istent ly accounted for 4% to 10% of the to ta l var iat ion and the smallest component was due to Vg. The overal l resu l ts of the study of genetic var iat ion between the various Douglas-f ir populations suggest that a combination of provenance, family within proven-ance and indiv idual tree select ion would y i e l d the greatest increase in height growth. S ign i f i cant genotype x environmental interact ion suggests that select ion from provenances and fami l ies for increased height growth for a l l locations may not be possible. The re l a t i ve s ize of Vp xg shows a diminishing trend indicat ing that i t s e f fect may be decreasing in importance. Estimates for V^ were quite high ranging from 21.80% to 32.27% and accounted for an average of 24.73% of the var iat ion in to ta l height for a l l - 55 -regions combined. The standard error of i s r e l a t i v e l y low for a l l areas ind icat ing r e l i a b i l i t y in the results and consistent values for the r a t i o (S.E. (V A)/V A) 100) for each seed zone are observed. The r e l a t i v e l y large 2 amount of addit ive genetic variance caused estimates of narrow sense h to 2 be moderately high, ranging between 0.28 and 0.52. High values for h i n -dicate that there are opportunities for s i gn i f i can t improvement by se lect ion in Douglas-f ir. The cor re la t ion analysis between to ta l heights at various ages and progeny annual growth and tota l height for juveni le height growth in Douglas-fir suggest that r e l i ab l e se lect ion of the best and delet ion of the poorest provenances and fami l ies may begin at age f i v e . The results from the ranking of provenances according to mean 1975 and 1978 tota l heights sub-s tant iate th i s conclusion. Studies should be continued to determine i f the re lat ionsh ip of ear ly juveni le to l a te r juveni le height growth can be extra-polated to l a te r age classes. Early test ing i s advantageous since i t increases the se lect ion i n tens i t y , which in turn increases the possible genetic gains through se lect ion and breeding. Reccommendations a r i s ing from the analysis presented in th i s thesis relevant to a tree improvement program for Douglas-f ir in south coastal B r i t i s h Columbia would include se lect ion of the top provenances in seed zones 2 and 3 and the best provenances from seed zone 1. These would include B r i t i s h Columbia, Washington and Oregon sources. Additional selections from other provenances within the selected seed zones and other seed zones are suggested to increase the genetic base and provide the potential for i n t e r r a c i a l crosses. Using as an example a selected proportion of 25% of the best indiv iduals at Haney, the figures obtained for genetic gain show that an increase - 56 -in mean two to eight year to ta l height from 17.90% to 10.96% i s poss ib le. I f se lect ion was made in the best provenance in the best seed zone, an increase in tota l height growth of almost 33% is suggested; se lect ing the best indiv idual within that provenance could give:.an addit ional increase of 70%. Further gains can be derived i f higher se lect ion i n tens i t i e s are used. - 57 -LITERATURE CITED Anderson, R.L. and Bancroft, T.A. 1952. S t a t i s t i c a l theory in research. McGraw-Hill Book Co., New York. 399 pp. Barner, H. 1973. Procurement of Douglas-fir seed for provenance:-research. IUFRO Douglas-f ir Working Party, Gottingen, September: 17-24. B i ro t , Y. 1974. Reduction of v a r i a b i l i t y in f lushing time in re l a t i on with the natural se lect ive pressures within some.populations of Douglas-f i r from the state of Washington: consequences for the breeding program. P r o c , Jo int IUFRO Meeting s. 02.04.1 3, Stockholm, Session V: 339-350. B i ro t , Y. 1976. Genetic parameters var iat ions between some populations of Douglas-f ir: consequences on se lect ion. P r o c , IUFRO Jo int Meeting on Advanced Generation Breeding, Bordeaux, August: 19-38. Bock, R.D., Wainer, H., Petersen, A., Thissen, D., Murray, J . and Roche, A. 1973. A parameterization for indiv idual human growth curves. Human Biology 45(1): 63-80. Campbell, R.K. 1972. Genetic v a r i a b i l i t y in juveni le height-growth of Douglas-f ir. S i lvae Genetica 21(3-4): 126-129. Cockerham, C.C. 1954. An extension of the concept of pa r t i t i on ing hered-i t a r y variance fo r ' ana l y s i s of covariances among r e l a t i v e s when epistas i s i s present. Genetics 39: 859-882. Crow, J.F. and Kimura, M. 1970. An introduction to population genetics theory. Harper and Row, New York. 591 pp. Eiche, V. and Anderson, E. 1974. Survival and growth rates of Scots pine. Theor. and Appl. Genetics 44(2): 49-57. Falconer, D.S. 1960. Introduction to quant itat ive genetics. Ronald Press Co., New York. 365 pp. 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 in applied breeding. Essay for Forestry 302. U.B.C. Faculty of Forestry 16 pp. (mimeo). Fisher, R.A. 1918. The corre lat ion between re lat ives on the supposition of Mendelian inheritance. Trans. Roy. Soc. Edinb. 52: 399-433. Fletcher, A.M. and Barner, H. 1978. The procurement of seed for prov-enance research with pa r t i cu la r reference to co l lect ions in NW America. Proc. IUFRO Congr., Vancouver, 1978, Vol . I. 14 pp. ( in press). - 58 -Gardner, CO. 1963. Estimates of genetic parameters in c r o s s - f e r t i l i z -ing plants and the i r implications in plant breeding. From a sym-posium and workshop sponsored by the Committee on Plant Breeding and Genetics of the Agric. Board at NC State College, National Academy of Sciences - NRC, Pub. 982: 225-248. Genys, J.B. and Forbes, D.C. 1973. V i r g in i a pines, Pinus v i rg in iana, from Chesapeake Bay region rank high in growth rate in Tennessee. Chesapeake Sc i . 14: 131-134. Giertych, M. 1974. Inadequacy of early tests for growth characters as evidenced by a 59 year old experiment. P r o c , Jo int IUFRO Meeting, S. 02.04. 1-3, Stockholm, Session IV: 237-242. Goggans, J.F. 1962. The co r re l a t i on , va r i a t i on , and inheritance of wood properties in Lob lo l ly pine. Tech. Rep. No. 14, School of For., N.C Agric. Exp. Sta. , Raleigh, North Carol ina. 17 pp. Harvey, W.R. 1968. Least-squares and maximum l i ke l i hood general purpose program. Ohio State Univ. 35 pp. Harvey, W.R. 1975. Least-squares analysis of data with.unequal subclass numbers. Agr ic. Res. Service H-4, U.S. Dept. of A g r i c , Data Systems Appl icat ion D iv i s ion. 157 pp. Harvey, W.R. 1976. Mixed model least-squares and maximum l i ke l i hood computer program. Ohio State Univ. 76 pp. Hattemer, H.H. 1963. Estimates of h e r i t a b i l i t y published in forest tree breeding research. F.A.0./F0RGEN 63-20/3: 14-28. Henderson, CR. 1954. The e f fect of inbreeding on genetic components of general and spec i f i c combining a b i l i t y . Jour. An. Sc i . 13: 959-963. Kempthorne, 0. 1955. The theoret ica l values of corre lat ions between re lat ives in random mating populations. Genetics 40: 153-167. Kvestich, A.M. 1976. Studies in height growth of selected Douglas-f ir (Pseudotsuga mensiesii (Mirb.) Franco) provenances and progenies. B.S.F. Thesis. Univ. of B.C. 55 pp. Lacaze, J.F. and M. Arbez. .1971. V a r i a b i l i t y intraspecfique de l ' ep icea (Picea abies Karst. ) . H e r i t a b i l i t e et correlat ions genetiques de quelques caracteres au stade juven i le . Ann. Sc i . Forest. 28(2): 141-183. L i , C C 1954. Population genetics. Univ. of Chicago Press. 368 pp. Lines, R. 1967. Standardization of methods for provenance research and test ing (Report of WG meeting at Pont-a-Mousson, 1965). IUFRO Congress, Sec. 22 - AG 22/24, Munchen, I I I : 672-718. Lush, J.L. 1949. H e r i t a b i l i t y of quantitative characters in farm animals. Proc. 8th Intern. Congr., Hereditas, Suppl. Vo l . : 356-375. - 59 -Meier, R.J. and Goggans, J.F. 1977. H e r i t a b i l i t i e s of height, diameter and spec i f i c gravity of younq V i r g in i a pine. Forest Science 23(4): 450-456. Meier, R.J. and Goggans, J.F. 1978. H e r i t a b i l i t i e s and correlat ions of the co r t i ca l monoterpenes of V i r g in i a pine (Pinus v i rg in iana M i l l . ) . S i lvae Genetica 27 (2): 79-84. Namkoong, G. and Conkle, M.T. 1976. Time trends in genetic control of height growth in Ponderosa pine. Forest S c i . 22: 2-13. Namkoong, G., Snyder, E.B. and Stonecypher, R.W. 1966. H e r i t a b i l i t y and gain concepts for evaluating breeding systems such as seedling;-orchards. S i lvae Genetica 15(3): 76-84. Namkoong, G., Usanis, R.A. and Si len,. R.R. 1972. Age-related var iat ion in genetic control of height growth in Douglas-f ir. Theor. and Appl. Genet. 42: 151-159.' Nanson, A. 1965. Contribution a I 'etude de la valeur des tests precoces. I. Experience i n t e r n a t i o n a l sur I ' o r i g ine des graines d 'epicea (.1938). Trav. Stat. Rech. Groenendaal, Ser. E., No. 1. 60 pp. Nanson, A. 1968. La valeur des tests precoces dans la se lect ion des arbres fo res t ie r s en p a r t i c u l i e r au point de vue de l a croissance. Dissertat ion. Station de Recherches des Eaux et Forets, Groenendaal-Hoei laart. 242 pp. Nanson, A. 1970. Juvenile and correlated t r a i t se lect ion and i t s e f fect on se lect ion programs. In papers presented at the second meeting of the Working Group on Quantitative Genetics, Section 22, IUFRO, Aug. 18-19, 1969, Raleigh, N.C.: 17-26. Orr-Ewing, A.L. and Yeh, F.C.H. 1978. Survival and growth t r a i t s of rac ia l crosses with Douglas-f ir. Res. Note No. 85, Province of B.C., Min istry of Forests. 44 pp. Owino, F. 1977. Genetic divergence in selected populations of Lob lo l l y pine. S i lvae Genetica 26 (2-3): 64-66. P a t l a j , I.N. 2973. Geograficeskie Kultury sosny obyknoviennoj v Ukrainskoj SSR. Proc. IUFRO Symp. Genetics of Scots pine. Warsaw-Kornik: .1-26. Pirchner, F. 1969. Population genetics in animal breeding. Freeman W.H. and Co. 274 pp. Rink, G. and Thor, E. 1976. Variance components and gains in volume growth in V i r g in i a pine (Pinus v i rg in iana M i l l . ) . S i lvae Genetica ... . ;25(1): .17-22. Schmidt, W. 1963. Early tests . F.A.O./FORGEN 63 2a/10, In World Consult-ation on Forest Genetics and Tree Improvement P r o c , Vol. 1, Rome: 10-20. - 60 -Sprague, G.F. 1966. Quant i tat ive genetics in plant improvement. From a symposium held at Iowa State Un iver s i t y on Plant Breeding. Edited by K.J. Frey. Iowa State Univ. Press: 315-354. S q u i l l a c e , A . E . , Bingham, R.T., Namkoong, G. and Robinson, H.F. 1966. H e r i t a b i l i t y o f j uven i l e growth rate and expected gain from s e l e c t i o n in Western White pine. S i l vae Genetica 16(1): 1-6. S t e e l , R.G.D. and T o r r i e , J .H . 1960. P r i nc i p l e s and procedures of s t a t i s t i c s . McGraw-Hill Book Co. Inc. New York, Toronto, London. 481 pp. S te inhof f , R.J. 1974. Juveni le-mature c o r r e l a t i o n in Ponderosa pine and Western White pine. P r o c , Jo i n t IUFRO Meeting, s. 02.04. 1-3, Stockholm, Session IV: 243-250. Stonecypher, R.W. 1966. The Lob l o l l y pine h e r i t a b i l i t y study. PhD. Thes i s . Southlands Experiment Forest , Tech. Bui . No. 5, Internat ional Paper Co., Bainbr idge, Ga. 128 pp. Suzuki, D.T. and G r i f f i t h s , A . J . F . 1976. An int roduct ion to genetic ana ly s i s . W.H. Freeman and Co. San Franc i sco. 468 pp. S z i k l a i , 0. 1974. Juveni le-mature c o r r e l a t i o n P r o c , Jo i n t IUFRO Meeting S. 02.04. 1-3, Stockholm, Session IV: 217-235. Toda, R. 1958. Var i a t ion and h e r i t a b i l i t y of some quant i t a t i ve charac-ters in Cryptomeria. S i l vae Genetica 7(3): 87-93. Toda, R. 1964. A b r i e f review and ..conclusions of the d i scuss ion on seed orchards. S i l vae Genetica 13 (1-2): 1-4. Toda, R. 1972. H e r i t a b i l i t y problems in fo res t genet ics . IUFRO - SABRA0 Jo in t Symposia, Tokyo. A-3, I: 1-9. Wakely, P.C. 1971. Relat ion of t h i r t i e t h - year to e a r l i e r dimensions of southern pines. For. S c i . 17: 200-209. Wehrnhahn, C F . 1979.' Personal communication. 8 February. Yao, C. 1971. Geographic va r i a t i on in seed weight, some cone sca le measurements and seed germination of Doug las - f i r (Pseudotsuga  menzies i i (Mirb.) Franco). M.F. Thes i s . Un iver s i t y of B.C. 8.8 pp. Zobel , B. 1961. Inheritance of wood propert ies in con i f e r s . S i l vae Genetica 10(3): 65-70. Zobel , B.J. and Dorman, K.W. 1973. L o b l o l l y pine as an exo t i c . Forest Genetics Resources Information, F.A.0. Rome: 3-15. - .61 -APPENDIX I Least-Squares Analysis of Variance Seed Zone 1 F Values Source of Var iat ion df 1972 1973 1974 1975 1976 1977 1978 Block 2 18.30*- 15.26* 12.38*15.82* 15.90*24.85*;41.26* Provenance 9 26.75* ' 43.48*' 53.10* 54.92*-59.28* 66.16* 70.53* Block x Prov. 18 12.25* 15.63*12.62* 8.83* 7.71* 6.65* 6.11* Family/Prov. 70 2.35* 2.36* 2.35* 2.49* 2.53* 2.27* 2.23* Least-Squares Analysis of Variance Seed Zone 2 F Values Source of Var iat ion df 1972 1973 1974 1975 1976 1977 1978 Block 2 5. .96* 17. .40* 18. ,42* 13. ,37* 12. .42* 11. .90* 7. .10* Provenance 9 4. ,07* 6. .0* 4. .79* 5. ,49* 8. .65* 10. .58* 10. .42* Block x Prov. 18 8. ,81* 9, .75* 9. .10* 8. ,96* 10, .37* 10. .90* 9. .59* Family/Prov. 70 3. ,13* 2, .84* 2. .58* 2. .37* 2, .39* 2, .50* 2, .42* * P ^ 0.01 APPENDIX I (continued) Least-Squares Analys i s of Variance Seed Zone 3 Source of Var i a t ion df 1972 1973 F Values 1974 1975 1976 1977 1978 Block 2 3.58**15.99* ' 14.05* 4.49** 4.39** 1.79NS 0.48NS Provenance 9 20.64* 28.18* 28.30* 25.49* 28.96* 30.63* 31.45* Block x Prov. 18 18.75* 23.71* 22.25* 17.26* 16.48* 15.99*' 13.47* Family/Prov. 70 2.68* 2.44* 2.45* 2.30* 2.15*' 2.31* 2.53* Least-Squares Analys i s of Variance In ter io r Provenances F Values Source of Var i a t ion df 1972 1973 1974 1975 1976 1977 1978 Block 2 23.16* 27.34* . 42.81* 48.30* 46.18* 45.53* 50.67* Provenance 9 51.81* 62.64* 52.69* 45.60* 43.91*- 50.59* 54.14* Block x Prov. 18 4.68* 7.33* 6.47* 4.77*. 5.06* 5.59* 6.63* Family/Prov. 70 2.49* 2.68* 2.70* 2.44* 2.21* 3.45* 2.03* * P ^ 0 . 0 1 **P < 0.05 NS not s i g n i f i c a n t APPENDIX II Components of Variance: Seed Zone 1 Total Year Pheno. VP. V VPxB VF/P Variance 1972 24.92 (14.43) 1 1.76 (1.01) 32.66 (18.78) 10.28 (5.91) 104.27 (59.97) 173.88 1973 133.82 (20.74) 0.0 (0.0)2 138.21 (21.42) 33.81 (5.24) 339.39 (52.60) 645.23 1974 352.02 (25.37) 0.0 (0.0) 235.47 (16.97) 72.09 (5.20) 728.07 (52.46) 1387.75 1975 630.75 (26.97) 24.54 (1.05) 274.60 (11.74) 148.67 (6.36) 1206.48 (53.89) 2339.05 1976 1139.94 (29.04) 48.09 (0.34) 393.58 (10.03) 236.32 (6.02) 2107.43 (53.69) 3925.36 1977 2113.86 (31.73) 177.09 (2.66) 550.16 ( 8.06) 324.87 (4.88) 3495.31 (52.47) 6661.29 1978 3437.20 (32.67) 521.28 (4.95) 758.31 (7.21) 477.99 (4.54) 5326.65 (50.63) 10521.44 Percent of estimated phenotypic variance Negative component estimate APPENDIX II (continued) Components of Variance: Seed Zone 2 Year V P 4.32 (1.98) 1 VB VPxB VF/P Total Pheno. Variance 1972 Q . 0 (o.or 32 97 ( 15.01) 23.72 (10.85) 157.50 (72.08) 218.51 1973 23.62 (3.16) 10.84 (1.43) 124 11 ( 16.42) 68.82 (9.12) 528/67 (69.92) 756.06 1974 36.74 (2.44) 27.06 (1.78) 235 31 ( 15.65) 121.29 (8.07) 1083.25 (72.04) 1503.64 1975 71.65 (2.94) 21.13 (0.87) 381 65 ( 15.67) 173.51 (7.13) 1787.36 (73.39) 2435.30 1976 209.57 (4.80) 16.85 (0.39) 769 75 ( 17.64) 302.30 (6.93) 3064.51 (70.24) 4362.98 1977 411.97 (5.87) 12.99 (0.19) 1276 36 ( 18.18) 510.19 (7.27) 4809.12 (68.50) 7020.63 1978 598.10 (5.95) 0.00 (0.0) 1636 80 ( 16.28) 713.54 (7.10) 7105.36 (70.67) 10053.81 Percent of estimated phenotypic variance Negative component estimate APPENDIX II (continued) Components of Variance: Seed. Zone 3.Provenances Year V P VB VPxB VF/P V ep Total Rheno. Variance 1972 19.20 (9.83) 1 0.0 (0.0) 2 52. .05 (26.64) 13.08 (6.69) 111.03 (56.84) 195.35 1973 88.02 (12.34) 0.0 (0.0) 220. .60 (30.92) 37.04,(5.19) 367.74 (51.55) 763.40 1974 179.17 (12.63) 0.0 (0.0) 418, .47 (29.50) 75.67 (5.33) 745.45 (52.54) 1418.77 1975 279.47 (12.42) 0.0 (0.0) 556, .84 (24.75) 117.72 (5.23) 1296.27 (57.60) 2250.30 1976 543.20 (14.19) 0.0 (0.0) 902, .20 (23.56) 177.85 (4.64) 2206.19 (57.61) 3829.44 1977 925.61 (14.92) 0.0 (0.0) 1404, .69 (22.64) 325.56 (5.25) 3547.45 (57.19) 6203.30 1978 1384.07 (15.73) 0.0 (0.0) 1700 .75 (19.33) 552.32 (6.28) 5162.56 (58.67) 8799.69 Percent of estimated phenotypic variance Negative component estimate APPENDIX II (continued) Components of Variance: . Interior,Provenances Year V P VB VPxB VF/P V e P Total Pheno. Variance 1972 25.05 (27.12) 1 2.73 (2.96) 5.44 (5.89) 5.83 (6.31) 53.30 (57.72) 92.23 1973 117.75 (29.62) 11.47 (2.88) 36.27 (9.12) 25.55 (6.43) 206.57 (51.95) 397.60 1974 222.12 (25.76) 46.84 (5.43) 70.56 (8.18) 58.11 (6.74) 464.63 (53.89) 862.15 1975 381.25 (23.65) 111.62. (6.93) 96.78 (6.00) 98.05 (6.08) 924.26 (57.34) . 1611.96 1976 650.33 (23.18) 186.96 (6.66) 184.66 (6.58) 145.03 (5.17) 1638.84 (58.41) 2805.82 1977 1267.18 (25.74) 306.17 (6.22) 352.16 (7.15) 235.17 (4.78) 2763.10 (56.12) 4923.78 1978 2129.03 (28.52) 0.00 (0.0) 2 676.50 (9.06) 327.08 (4.38) 4332.47 (58.04) 7465.06 Percent of estimated phenotypic variance Negative component estimate - 67 -APPENDIX III Regression Analys i s fo r Maternal E f fec t s in Seed Zone 1 Provenance 1972 R Percent of 1973 R Percent of No. Var i a t ion Var i a t ion Explained Explained 23 0.0112 i:12 0.0175 1.75 32 0.0209 2.09 0.0263, 2.63 55 0.0744 7.44 0.0611 6.11 90 0.0272 2.72 0.0341 3.41 95 0.0188 1.88 0.0081 0.81 99 0.0655 6.55 0.0631 6.31 96 0.0027 0.27 0.0025 0.25 117' 0.0002 0.02 0.0000 0.00 111 0.0003 0.03 0.0001 0.01 124 0.0194 1.94 0.0589 5.89 104 0.0114 1.14 0.0154 1.54 Average 2.29 2.61 - 68 -APPENDIX III (continued) Regression Analysis for Maternal Effects in Seed Zone 2 2 2 Provenance 1972 R Percent of 1973 R Percent of No. Var iat ion Var iat ion Explained Explained 12 0.0121 1.21 0.0027 0.27 51 0.0516 5.16 0.0385 3.85 52 0.0419 4.19 0.0265 2.65 53 0.0241 2.41 0.0139 1.39 67 0.0124 1.24 0.0268 2.68 79 0.0021 0.21 0.0210 2.10 83 0.0000 0.00 0.0010 0.10 87 0.0543 5.43 0.0308 3.08 89 0.0075 0.75 0.0000 0.00 91 0.0179 1.79 0.0008 0.08 92 0.0091 0.91 0.0280 2.80 115 0.1461 14,61 0.1207 12.07 Average 3.14 4.16 - 6 9 -APPENDIX IV 0 , H e r i t a b i l i t y ( l / ) 1 and Standard Errors (S.E.) for Total Height Based on Individual Provenances Provenance No. Year 42 52 55 66 1972 0.00 + 0. 00 0.57 + 0. ,43 0.41 + 0.37 0.76 + 0.51 1973 0.00 + 0. 00 0.38 + 0. ,36 0.37 + 0.36 0.63 + 0.47 1974 0.10 + 0. 21 0.27 + 0. ,31 0.39 + 0.36 0.19 + 0.30 1975 0.26 + 0. 29 0.21 + 0. ,28 0.41 + 0.37 0.06 + 0.24 1976 0.39 + 0. 34 0.20 + 0. 28 0.49 + 0.40 0.00 + 0.00 1977 0.43 + 0. 36 0.12 + 0. ,24 0.55 + 0.43 0.00 + 0.00 1978 0.39 + 0. 34 0.15 + 0. 25 0.40 + 0.37 0.00 + 0.00 H e r i t a b i l i t i e s were ca l cu la ted according to the fo l lowing formula: V F + W + Vep 1978 lCVT 117 111 1975 104 111 117 •''Provenances ranked in increasing 2 Approximately 25% of provenances APPENDIX V Ranking According to Mean 1975 and 1978 Total Height: Seed Zone 1 Provenances 124 99 23 90 124 99 90 23 height from l e f t to r ight , represented to the r ight of the l i ne APPENDIX V (continued) Ranking According to Mean 1975 and 1978 Total Height: Seed Zone 2 Provenances 1978 115 1 12 92 79 51 83 87 91 67! 1975 115 12 92 79 51 87 89 53 83 Provenances rank in increasing height from l e f t to r ight . Approximately 25% of provenances represented to the r ight of the l i ne 1978 103 1 86 113 112 1975 103 86 113 112 Provenances ranked in increasing Approximately 25% of provenances APPENDIX V (continued) Ranking According to Mean 1975 and 1978 Total Height: Seed Zone 3 Provenances 60 40 76 101 29 73 60 101 76 97 73 29 eight from l e f t to r ight . presented to the r ight of the l i n e . APPENDIX V (continued) Ranking According to Mean 1975 and 1978 Total Height: Inter ior Provenances 1978 l l 1 18 66 10 46 28 64 1975 11 66 18 46 10 77 64 Provenances ranked in increasing height from l e f t to r ight . Approximately 25% of provenances represented to the r ight of the l i ne . 77 2 6 80 28 80 CO - 74 -APPENDIX VI Mean (iii) and Standard Deviation (S.D.) of 1972 to 1978 Total Height for Provenances in Seed Zone 1, 2 and 3 and Inter ior (cm.) Location Year Seed Zone 1 Seed Zone 2 Seed Zone 3 Inter ior iii ± S.D. iii ± S.D. iii ± S.D. iii ± S.D. 1972 30.15 ± 10.70 37.27 ± 13.46 33.91 ± 11.14 23.33 ± 7.69 1973 56.49 ± 19.32 73.22 ± 24.44 66.32 ± 20.12 44.01 ± 15.24 1974 85.57 ± 28.29 113.18 ± 34.71 100.37 ± 28.66 67.19 ± 22.86 1975 120.87 ± 36.81 158.91 ± 44.28 141.36 ± 37.60 97.81 ± 31.97 1976 164.31 ± 48.41 216.17 ± 58.02 191.64 ± 4 8 . 8 3 133.62 ± 42.24 1977 222.47 ± 61.81 287.88 ± 72.93 256.26 ± 62.23 177.96 ± 54.76 1978 284.37 ± 76.19 358.72 ± 88.42 320.79 ± 75.60 225.77 ± 68.26 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0075131/manifest

Comment

Related Items