Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Early results of the Douglas-fir cooperative progency test Bartram, Victor Cameron 1978

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

Item Metadata

Download

Media
831-UBC_1978_A6_7 B37.pdf [ 5.08MB ]
Metadata
JSON: 831-1.0075299.json
JSON-LD: 831-1.0075299-ld.json
RDF/XML (Pretty): 831-1.0075299-rdf.xml
RDF/JSON: 831-1.0075299-rdf.json
Turtle: 831-1.0075299-turtle.txt
N-Triples: 831-1.0075299-rdf-ntriples.txt
Original Record: 831-1.0075299-source.json
Full Text
831-1.0075299-fulltext.txt
Citation
831-1.0075299.ris

Full Text

EARLY RESULTS: OF THE DOUGLAS'-FIR COOPERATIVE.PROGENY TEST b y VICTOR- CAMERON BART RAM B.S.P., U n i v e r s i t y , o f B r i t i s h C o l u m b i a , 1973 A THESIS: SUBMITTED, I N P A R T I A L FULFILLMENT OF THE REQUIREMENTS FOR THE: DEGREE. OF MASTER OF. SCIENCE i n THE FACULTY:,: OF GRADUATE. STUDIES ( F o r e s t r y ) We a c c e p t t h i s t h e s i s ass c o n f o r m i n g t o t h e r e q u i r e d s t a n d a r d THE: UNIVERSITY; -OF BRITISH'COLUMBIA O c t o b e r , 1977 * * • © V i c t o r C a m e r o n B a r t r a m , 19 77 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of Brit ish Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of FORESTRY The University of Brit ish Columbia 2075 Wesbrook P l a c e Vancouver, Canada V6T 1W5 OCTOBER M, 1977 ABSTRACT In 1969, a cooperative progeny te s t of Douglas-fir, Pseudotsuqa menzies-ii (Mirb.) Franco, was i n i t i a t e d to evaluate the growth performance of progeny from selected plus trees. Using height data measured i n 19 75, i t i s shown that the mean height of plus tree progeny i s s i g n i f i c a n t l y greater than the mean height of the con t r o l progeny. Much of t h i s gain may, however, be due to the hetero t i c e f f e c t of crossing parents from a l l o p a t r i c populations. The breed-ing value of i n d i v i d u a l plus trees showed a wide range of v a r i a t i o n . This range was markedly reduced when the few extreme plus tree parents were excluded. An i n v e s t i g a t i o n of possible geographic trends showed that i n only one instance did the progeny of parents of s i m i l a r o r i g i n perform comparably. I t i s therefore con-cluded that, over the range of plus tree s e l e c t i o n , geographic o r i g i n i s ; of l i t t l e , importance i n determining breeding value. I n i t i a l juvenile-mature data demonstrated that nursery height i n 1969 and 19 70 was s i g n i f i c a n t l y c o r r e l a t e d with plantation height i n 19 75. As expected, 19 73 and 19 74 i i i p l antation heights were highly c o r r e l a t e d with 19 75 height. An attempt to predict genotype x environment i n t e r -actions from the l a t i t u d i n a l and l o n g i t u d i n a l displacement of the progeny from t h e i r plus tree parents proved unsuc-c e s s f u l . Other variables must therefore be considered be-fore: progeny performance at a s p e c i f i c l o c a t i o n can be su c c e s s f u l l y predicted. i v TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES v i LIST OF FIGURES v i i i ACKNOWLEDGEMENT i x INTRODUCTION 1 LITERATURE REVIEW . . . . . 4 A. Quantitative Genetics, 4 B. Progeny Testing. 7 C. Least Squares Analysis 9 (1) Q u a l i t a t i v e independent variables (Analysis of Variance) 10 (2) Quantitative independent variables (Multiple Regression) . . . . 12 (3) Q u a l i t a t i v e and quantitative independent variables 12 (4) Assumptions allowing v a l i d s t a t i s t i c a l inference 13 MATERIALS 14 METHODS 20 A. Data C o l l e c t i o n and Processing 20 B. Least Squares Analyses 21 (1) Analysis of tree height and condition of family types and plantations . . . . 21 (2) Analysis of parent performance i n 19 75, 22 (i ) h a l f - s i b parents 22 ( i i ) f u l l - s i b parents 23 (3) Juvenile-"mature" c o r r e l a t i o n s 25 (i ) nursery performance 25 ( i i ) plantation performance,- 26 (4) Interpretation of genotype x environment i n t e r a c t i o n s . . . . . . . 26 V Page C Computer Programs 28 RESULTS AND DISCUSSION . 29 A. Tree Height and Condition of Family Types and Plantations - 29 (1) Tree height i n 1973, 1974 and 1975 . . . . 29 (2) Tree condition i n 19 75 33 B. Parent Performance 35 (1) Height of progeny from h a l f - s i b parents i n 19 75, . . 35 (2) Height of progeny from f u l l - s i b parents i n 1975 42 (i ) maternal GCA i n 19 75 43 ( i i ) paternal GCA. i n 19 75 .45 (3) Maternal e f f e c t s i n 19 75 4 7 C. Problems i n the Family Type, Plantation and Parent Analyses 49 D. Juvenile-"Mature" Co r r e l a t i o n s 50 (1) Nursery measurements 50 (2) Ea r l y plantation measurements 52 E. Interpretation of Genotype x Environment Interactions . . . . . 54 F. H e r i t a b i l i t y Estimation 55 CONCLUSIONS AND RECOMMENDATIONS 57 SUMMARY 60 LITERATURE CITED 61 APPENDICES 64 v i LIST OF TABLES Table Page 1. Results of the 1968 p a r t i a l d i a l l e l c rossing program 15 2. The three plantations of the Coop-erativ e Progeny Test 17 3. Analysis of variance of family type and plantation height i n 19,73. 30 4. Analysis of. variance of family type and plantation height i n 19 74 30 5. Analysis of variance of family type and plantation height i n 19 75. . . 30 6. Mean height i n centimetres of family type i n 1973, 19 74 and 19 75 31 7. Mean plantation height i n centimetres i n 1973, 1974 and 1975 34 8. Number of trees included i n the family type and plantation analysis 36 9. Summary of tree condition i n 19 75 per plantation . . . . . . . . 36 10. Analysis of variance of family height i n 19 75 3 7 11. Mean 19 75 height i n centimetres of h a l f -s i b f a m i l i e s from open-pollinated plus trees 38 12. Mean 1975 height i n centimetres of h a l f -s i b f a m i l i e s from open-pollinated clone banks 40 r v i i Table Page 13. Analysis of variance of maternal and paternal GGA i n 19 75 413 14. GCA i n centimetres of maternal parents i n 19 75 44 , 15. GCA i n centimetres of paternal parents i n 19 75 46 16. Maternal eff.ects i n centimetres during 19 75 . . . . . . . . . . . 48 17. Reciprocal cross differences i n c e n t i — metres during.: 19 75 48 18. C o r r e l a t i o n of 19 75 family height with 1969 and 19 70^  family measurements i n the nursery 51 19. C o r r e l a t i o n of 19 75 tree, height with 1974 and 1973 tree heights . . . . 53 20. Summary of genotype., x environment i n t e r a c t i o n analysis 55 v i i i . LIST. OF FIGURES Figure Page 1. Family p o s i t i o n within each plantation — s u b s e c t i o n 1 18 2. Family p o s i t i o n within each plantation — s u b s e c t i o n 2 19 i x ACKNOWLEDGEMENT The; author would f i r s t l i k e to< express h i s sincere; thanks to Dr. 0. S z i k l a i , Professor of Forestry, f o r h i s encouragement and h e l p f u l advice throughout t h i s study. The author i s also indebted to: Dr. A. Kozak, Profess-or of Forestry; Dr. C. Hornby, Associate Professor of H o r t -i c u l t u r e and Dr. R. Peterson, Assistant Professor of Animal Science f o r t h e i r comments; and c r i t i c i s m s . Special thanks are due to: B r i t i s h Columbia Forest Products Ltd., Crown Zellerbach Canada Ltd. and Tahsis Co. Ltd. f o r measuring the. progeny heights. Their assistance i s most appreciated:. Acknowledgement i s also made to the- National Research Council f o r the f i n a n c i a l assistance o f f e r e d during the l a s t year of the author's studies. F i n a l l y , the author would l i k e to thank hi s wife, E l a i n e , f o r her understanding and;great assistance during the completion of t h i s study. 1 EARLY RESULTS OF THE DOUGLAS-FIR COOPERATIVE PROGENY TEST INTRODUCTION The program f o r the genetic improvement of Douglas-f i r , . Psejod^tsu^a men^ie_sii (Mirb.) Franco, began on the coast of B r i t i s h Columbia i n 1956. A shortage of high elevation seed was causing concern within the B r i t i s h Columbia Forest Service (B.C.F.S.). The decision was therefore made to s e l e c t and vegetatively propagate superior phenotypes or plus trees i n c l o n a l seed orchards. These seed orchards would then be used to provide high q u a l i t y seed su i t a b l e f o r future r e f o r -estation (Orr-Ewing, 1958; Heaman, 1967). The s e l e c t i o n work progressed slowly i n 195 7 and 1958 because of l i m i t e d resources (Orr-Ewing and S z i k l a i , 1960). However, the formation of the Plus Tree Board (subsequently the Tree Improvement Board of the Tree Farm Forestry Committee) i n 1959 proved to be of great assistance. This organization was composed of both the Federal and P r o v i n c i a l Forest Services, major coastal companies and the U n i v e r s i t y of B r i t i s h Columbia (U.B.C.). The objectives of t h i s organization were: 1) to coordinate tree improvement a c t i v i t i e s and 2) to stimulate 2 i n t e r e s t i n tree s e l e c t i o n and seed oxchard establishment. Flus Tree Weeks were organized from 1 9 5 9 t o 1 9 6 5 to inform i n d u s t r i a l f o r e s t e r s of se l e c t i o n c r i t e r i a . , thus allowing independent company se l e c t i o n programs. In t h i s cooperative atmosphere the work progressed r a p i d l y . By 1 9 6 7 , a t o t a l of 4 1 4 plus trees: had been chosen and the se l e c t i o n objectives were considered l a r g e l y f u l -f i l l e d at t h i s times. The majority of these trees had also been vegetatively preserved. Interest therefore turned towards progeny t e s t i n g of the selected trees to evaluate t h e i r breeding p o t e n t i a l . In the spring of 1 9 6 8 , when a considerable amount of flowering was; observed on several company clone: banks, a p a r t i a l d i a l l e l crossing program was completed ( S z i k l a i , 1 9 7 1 ) . The Cooperative Progeny Test was i n i t i a t e d when seedlotss c o l l e c t e d from t h i s crossing program and from open-pollinated^ plus tree o r t e t s and clones were sown i n the spring of 1 9 6 9 . The seedlings were raised i n the nursery f o r two years and then outplanted at three t e s t s i t e s on Vancouver Island. In the autumn of 1 9 7 5 each tree was: measured f o r t o t a l 1 9 7 5 height and height growth i n each of 1 9 7 5 and 1 9 7 4 . Using the data c o l l e c t e d from t h i s progeny t e s t , the objectives of t h i s thesis were to: 1 . Evaluate the ea r l y phenotypic performance of the plus tree progenies by comparing t h e i r height to the height of the con t r o l progenies i n 1 9 7 3 , 1 9 7 4 and 1 9 7 5 . , Such an 3 e v a l u a t i o n w o u l d a l l o w an a s s e s s m e n t o f t h e e a r l y g a i n s t h a t c a n b e e x p e c t e d .from p l u s t r e e s e l e c t i o n . 2. E s t i m a t e t h e b r e e d i n g p o t e n t i a l o f s p e c i f i c p l u s t r e e s f r o m t h e h e i g h t o f t h e i r p r o g e n y i n 19 7 5 . 3. E x a m i n e i n i t i a l j u v e n i l e - m a t u r e c o r r e l a t i o n s . 4. A t t e m p t t o i n t e r p r e t g e n o t y p e x e n v i r o n m e n t i n t e r a c t i o n s . 4 LITERATURE:, REVIEW, A.. Quantitative Genetics Only a b r i e f review of quantitative genetics i s given. Those further interested are r e f e r r e d to Lush (1945), Lerner (1958), Falconer (I960);, Stonecypher (1966)., Namkoong et a l . (1966), Sprague (1967), Mather and Jinks (1971), G i l b e r t (1973) and Wright (19 76),. Quantitative genetics r e f e r s to the geneticai analysis of; those t r a i t s which e x h i b i t continuous v a r i a t i o n . The theory states that many genes contribute to the phenotypic expression of a continuous t r a i t . The contribution of each gene i s sub-j e c t to the influence of:: 1) o.ther genes and 2),* the environment. The nature of t h i s gene action makes? the i d e n t i f i c a t i o n of s p e c i f i c genes impossible. Thus genotypes cannot be accurately i d e n t i f i e d ; and:; breeding values must be estimated from pheno-1 t y p i c measurements or i n d i c e s . The theory of quantitative genetics o u t l i n e s several phenotypic indic e s which can be measured to evaluate an i n d i v i d u a l ' s breeding value. The author w i l l introduce two of these indic e s which are commonly used i n f o r e s t tree T h e o r e t i c a l l y , the breeding value of an i n d i v i d u a l i s the sum of the average e f f e c t s of a l l i t s genes which a f f e c t the continuous t r a i t . Breeding values are most accurately estimated from mean progeny performance (Falconer, 1960). 5 improvement. The f i r s t and simplest phenotypic index of an i n d i v i d -ual's breeding value i s i t s own performance. I t i s assumed that those i n d i v i d u a l s which ex h i b i t phenotypic s u p e r i o r i t y i n a population also possess a high frequency of superior genes. I f these superior i n d i v i d u a l s are selected and i n t e r -mated, t h e i r progeny would receive these superior genes. Mean progeny performance should then surpass the mean performance of the population from which the parents were selected. Selection based s o l e l y on the phenotypic performance of i n d i v i d -uals i n the population i s termed i n d i v i d u a l s e l e c t i o n (Ledig, 19 74). I f the selected i n d i v i d u a l s are grouped en masse f o r mating the term, mass s e l e c t i o n i s often used (Falconer, 1960). The second phenotypic index of an i n d i v i d u a l ' s breeding value commonly used i n f o r e s t tree improvement i s the mean performance of that i n d i v i d u a l ' s family. This~index often gives a more accurate estimate of breeding value because e n t i r e f a m i l i e s which express phenotypic s u p e r i o r i t y are more l i k e l y to be g e n e t i c a l l y superior than s i n g l e , phenotypically superior i n d i v i d u a l s . Selection of i n d i v i d u a l s based on t h e i r mean family performance i s termed family s e l e c t i o n (Wright, 1976). The value of family s e l e c t i o n i s dependent on a large family s i z e and l i t t l e environmental v a r i a t i o n among f a m i l i e s (Falconer, 1960). 2 -As noted by Ledig (19 74);, l i t t l e or no family information i s a v a i l a b l e i n natural f o r e s t stands. T h u s f a m i l y s e l e c t i o n i n f o r e s t tree improvement i s dependent on the creation of fa m i l i e s through progeny t e s t i n g . 6 When superior i n d i v i d u a l s are selected according to t h e i r own performance ( i n d i v i d u a l selection) and intermated, Falconer (1960) gives the following formula to prredict the response to se l e c t i o n or genetic gain ( A G ) : The s e l e c t i o n i n t e n s i t y ( I ). i s the mean s u p e r i o r i t y of the selected i n d i v i d u a l s above the population mean. The additive 2 variance ( C T A ) i s the amount of phenotypic v a r i a t i o n i n the population a t t r i b u t a b l e to additive gene act i o n . The f i n a l 2 values i n the formula, (J p and 't are the standard deviation and variance of the phenotyp.es i n the population from which the. parents were selected. The f r a c t i o n off the additive variance over the phenotypic 2 2 variance: ( Cf ^  /(J^ ^ *-ne a b ° v e genetic gain formula i s termed h e r i t a b i l i t y i n the narrow sense (Snyder, 19 72). This value i s most e a s i l y interpreted as the regression c o e f f i c i e n t of the expected genetic gain on the se l e c t i o n attempted (Namkoong et a l . , 1966). Thus, the greater the h e r i t a b i l i t y , the greater the; genetic gain that can be expected at a constant s e l e c t i o n i n t e n s i t y . Prediction of response following family s e l e c t i o n i s also possible. The readerris referred to Falconer (1960) and Pirchner (1969) f o r s p e c i f i c genetic gain formulae. 7 B. Progeny Testing Progeny t e s t i n g i s a form of family s e l e c t i o n i n which the parents are selected according to the performance-of t h e i r o f f s p r i n g (Snyder, 19 72; Falconer, 1960). The prime advantage of progeny t e s t i n g i s i t s power to accurately estimate parental breeding values f o r t r a i t s of low h e r i t a b i l i t y . Lush (1945) a t t r i b u t e s t h i s power to the laws of sampling— each o f f s p r i n g represents an independent estimate of parental breeding value plus the e f f e c t s of gene i n t e r a c t i o n and environmental deviation. As the number of progeny increases, a l l environmental e f f e c t s and most gene i n t e r a c t i o n e f f e c t s cancel, y i e l d i n g a r e l i a b l e estimate of parental breeding value. Peterson (1970) substantiates t h i s concept with a path diagram. According to t h i s model, the greater the number of progeny, the higher the c o r r e l a t i o n between progeny mean and parental breeding value. Sprague (1966) divides prorgeny t e s t i n g i n t o two groups: 1) h a l f - s i b — o n e known parent and 2) f u l l - s i b — t w o known parents. These groups are outlined i n the following paragraphs. The h a l f - s i b progeny te s t generally assumes that the un-known parents contribute equally and therefore do not influence progeny performance. I f t h i s assumption i s true, and the progeny experience s i m i l a r environments, progeny performance i s influenced only by the known parent's breeding value plus some 8 of i t s e p i s t a t i c gene i n t e r a c t i o n (Pirchner, 1969) The f u l l - s i b progeny t e s t allows a more precise i n t e r -pretation of progeny performance than the h a l f - s i b t e s t because the contribution of both parents can be evaluated. Generally, a model i s assumed which contains both a maternal and a paternal e f f e c t plus an i n t e r a c t i o n e f f e c t between the two parents ( G i l b e r t , 1967). In plant breeding the e f f e c t of a p a r t i c u l a r parent plus one-half of the general mean i s termed that parent's General Combining A b i l i t y (GCA). The i n t e r -action e f f e c t of two p a r t i c u l a r parents i s termed t h e i r S p e c i f i c Combining A b i l i t y (SCA), (Falconer, 1960). Lerner (1950) discusses a problem which may bias the r e s u l t s of both the f u l l - and h a l f - s i b progeny t e s t . I f the progeny of each parent experience unequal environments, the evaluation of parental breeding values from the mean of t h e i r progeny i s not s t r i c t l y v a l i d . Mather and Jinks (1971) state that proper experimental design should a l l e v i a t e t h i s problem i n plant breeding. The ad d i t i o n a l problem of a lengthened generation i n t e r v a l , which accompanies progeny t e s t i n g , has been discussed (Falconer, 1960; Pirchner, 1969). Parents cannot be selected u n t i l r e l i a b l e progeny information becomes a v a i l a b l e , thus reducing genetic gain per unit time. Namkoonq et a l . (1966) compare rates of gain between mass selected and progeny tested seed orchards. They concluded that no one breeding system i s 9 superior over d i f f e r e n t combinations of h e r i t a b i l i t y and s e l e c t i o n i n t e n s i t y . The progeny test has become an i n t e g r a l part of f o r e s t tree improvement f o r three main reasons. F i r s t l y , almost a l l f o r e s t tree t r a i t s vary continuously. Thus, some phenotypic index i s required to estimate breeding values. Secondly, the accurate estimates of breeding value provided by the progeny t e s t are often desirable because an i n d i v i d u a l ' s own performance i s not always a good i n d i c a t o r of i t s breeding value. Ledig (1974) substantiates t h i s point when he notes that, based on published information, parent-offspring h e r i t a b i l i t i e s are less than 0.5 f o r most t r a i t s and probably below 0.2 f o r growth t r a i t s . T h i r d l y , those organisms capable of producing large numbers of progeny are best suited f o r progeny t e s t i n g . The p r o l i f i c nature of most tree species undoubtedly influences the popularity of the progeny t e s t i n f o r e s t tree improvement. C. Least Squares Analysis Least squares techniques are used extensively i n the genetical analysis of t r a i t s which vary continuously (Falconer, 1960; Harvey, 1966). The author has therefore made considerable use of l e a s t squares analysis i n t h i s study and s h a l l b r i e f l y review t h i s t o p i c . Searle (1966), using matrix notation, shows that l e a s t squares analysis minimizes the v a r i a t i o n i n the response v a r i a b l e which i s not explained by v a r i a t i o n i n the indepen-dent p r e d i c t i v e v a r i a b l e s . B a s i c a l l y , a p a r t i a l d e r i v a t i v e 10 i s derived f o r each independent v a r i a b l e and equated to zero. -. A\ system of so-called "normal equations" r e s u l t s which, when solved with matrix inv e r s i o n , y i e l d s a set of lea s t squares c o e f f i c i e n t s . Least squares analysis can be divided i n t o three categories according to the nature of the independent v a r i -ables used i n the model. These categories w i l l now be o u t l i n e d . (1) Q u a l i t a t i v e independent variables (Analysis of Variance) Many s t a t i s t i c s texts discuss q u a l i t a t i v e independent va r i a b l e s under the heading "Analysis of Variance" (Dixon and Massey, 1969; Hicks, 1964). Levels of treatments and i n t e r a c t i o n s are ei t h e r present or absent and are therefore coded using dummy va r i a b l e s . Commonly, the r e s t r i c t i o n that the deviations with-i n a f a c t o r or i n t e r a c t i o n sum to zero i s imposed to prevent the inversion of a singular matrix. This r e s t r i c t i o n i s l o g i c a l because Analysis of Variance (ANOVA) models .are generally expressed i n deviations from the general mean (Harvey, 19,66). The complexity of analysis and i n t e r p -r e t a t i o n of models with q u a l i t a t i v e independent variables depends p r i m a r i l y on the experimental design and the number of observations per c e l l . Designs with no crossed f a c t o r s (nested) present no problem regardless of the number of observations per c e l l . However, the complexity of analysis and i n t e r p r e t a t i o n of experimental designs with crossed f a c t o r s ( f a c t o r i a l ) i s 11 dependent on the frequency of observations per c e l l . An equal number of observations per c e l l (balanced design) i n a f a c t o r i a l experiment i s highly desirable ( L i , 1964). The lea s t squares s o l u t i o n , generally r e q u i r i n g matrix in v e r s i o n , reduces to simple sums and sums of squares c a l -c u l a t i o n s . In addition, the contribution of each f a c t o r can be d i r e c t l y assessed because each f a c t o r ' s c o n t r i b u t i o n i s unique to the model. Also, a l l treatment l e v e l s within a f a c t o r receive the same contribution from other f a c t o r s i n the experiment. Thus, mean differences among the l e v e l s of a f a c t o r can be at t r i b u t e d s o l e l y to that f a c t o r . Unequal numbers of observations per c e l l (unbalanced design) i n a f a c t o r i a l experiment cause problems. F i r s t , a computer solut i o n i s required f o r most models. Second, the v a r i a t i o n i n the dependent va r i a b l e that can be at t r i b u t e d to a f a c t o r i s ambiguous because i t depends when the f a c t o r i s entered i n t o the model ( G i l b e r t , 19 73; Kerlinger and Pedhazur, 1973). Overall and Spiegel (1969) suggest three procedures to handle t h i s problem. Their most conservative approach i s to~ note the contribution of a f a c t o r when i t i s entered l a s t into the model. Third, the comparison of means i n an unbalanced design i s less precise. One can never be sure i f two l e v e l s of a f a c t o r d i f f e r because of a r e a l d i f f e r e n c e or because of an unequal input from the other fac t o r s i n the model ( G i l b e r t , 1973). 12 (2) Quantitative independent variables (Multiple Regression) The least squares analysis of models composed s o l e l y of quantitative variables i s often discussed under the heading of "Multiple Regression" (Searle, 1966). In such models the l e v e l of a treatment i s s p e c i f i e d by the measured value of a continuous v a r i a b l e . The main d i f f i c u l t i e s with multiple regression analysis are analogous to the problems of the unbalanced f a c t o r i a l experiment discussed previously. As noted by L i (1964), the v a r i a t i o n a t t r i b u t a b l e to an independent v a r i a b l e i s ambiguous i f i t i s cor r e l a t e d with other independent v a r i a b l e s . I f entered f i r s t i n t o the model i t w i l l explain more v a r i a t i o n than i f entered l a s t . One procedure to t e s t a variable's contribution i n a multiple regression i s s i m i l a r to the most conservative procedure suggested by O v e r a l l and Spiegel (1969) f o r f a c t o r i a l experiments. The v a r i a b l e i s entered l a s t i n t o the model and i t s contribution tested by the so-c a l l e d " P a r t i a l - F " t e s t ( L i , 1964). (3) Q u a l i t a t i v e and quantitative independent va r i a b l e s Least squares analysis of models with both q u a l i t a t i v e and quantitative independent variables combines ANOVA and multiple regression procedures. The importance of any v a r i a b l e , d i s c r e t e or continuous, i s commonly determined by noting i t s contribution when entered l a s t into:the model. The s i g n i f i c a n c e 13 of t h i s contribution is ;; then tested by the same " P a r t i a l - F " t e s t used i n multiple regression ( L i , 1964; Kerlinger and Pedhazur, 19 73). (4) Assumptions allowing v a l i d s t a t i s t i c a l inference The c o e f f i c i e n t s derived from l e a s t squares analysis are s t a t i s t i c s which vary from sample to sample. To v a l i d l y t e s t hypotheses containing these c o e f f i c i e n t s : or to i n f e r population parameters, E i (1964) l i s t s the following f i v e assumptions:. 1. The unexplained v a r i a t i o n about the dependent v a r i a b l e i s normally d i s t r i b u t e d 2. The unexplained v a r i a t i o n about the dependent va r i a b l e has a constant variance 3. The regression of the dependent variable i s l i n e a r on each of the: independent variables} 4. The samples are drawn at random 5. The: independent variables remain constant f o r a l l samples The v a l i d i t y of these assumptions was tested i n the l e a s t squares analyses performed i n t h i s study. S p e c i f i c assumption v i o l a t i o n s are discussed. V 14 . MATERIALS The materials of t h i s study c o n s i s t of data c o l -l ected from the Cooperative Progeny Test. The h i s t o r y and layout of t h i s progeny te s t w i l l therefore be described i n t h i s section. A high frequency of reproductive buds i n the e a r l y spring of 1968 resulted i n the decision to begin a p a r t i a l d i a l l e l c rossing program wiithin four company clone banks. A t o t a l of 21,034 f i l l e d seeds was extracted from the: cones of 244 successful crosses (Table 1). Of these crosses, 55 were ul t i m a t e l y included i n the progeny te s t (54 f u l l - s i b f a m i l i e s and:one polymix family) because: they contained a s u f f i c i e n t number of f i l l e d seeds. The p a r t i a l d i a l l e l scheme i s shown i n Appendix I . In addition to these f u l l - s i b : f a m i l i e s , 36 h a l f - s i b families-were included from open-pollinated seed c o l l e c t i o n s . Of these f a m i l i e s , 21 were from the: o r i g i n a l o r t e t s and 15 3 were from grafted clones of plus t r e e s . F i n a l l y , nine c o n t r o l f a m i l i e s were included—seven from randomly selected seedlots and two from bare root planting stock. Thus, a t o t a l of 100 f a m i l i e s comprised t h i s progeny t e s t . Subsequently, h a l f - s i b f a m i l i e s from plus- trees w i l l be termed h a l f - s i b s ( p . t . ) and h a l f - s i b f a m i l i e s from clone banks termed h a l f - s i b s ( c . b . ) . Table 1. Results of the 1968 p a r t i a l d i a l l e l crossing program. Company Location Number of Number of Number of seed crosses cones (C.) Empty F i l l e d ( F ) Total(T) T/C F/C Tahsis Co. L t d . Gold River 138 2492 6213 a 2.5 Crown Zellerbach Canada Ltd. Courtenay 33 465 6093 9939 16032 34.5 21.3 B r i t i s h Columbia Forest Products Caycuse 53 392 6569 2983 9552 24.4 7.6 Rayonier Canada Ltd. Gordon River 20 206 3701 1899 5600 27.2 9.2 Total 244 3555 21034 5.9 S o u r c e : S z i k l a i (19 71) a T h i s number was estimated. 16 Following 1969 spring germination--survival, phenology of bud s e t t i n g , e p i c o t y l length and t o t a l height were tabulated f o r each family i n August. These observations and measurements were completed by research associates and undergraduate students at the Faculty of Forestry (U.B.C ) and B.C.F.S. personnel ( S z i k l a i , 19 71). In the spring of 19 70, the seedlings were transplanted i n t o J i f f y - p o t s and i n the autumn mean family height was measured from a randomly selected sample of nine seedlings per family ((Sigurdson, 1971). Outplanting of the seedlings (1+1) at the three p l a n t a -t i o n s l i s t e d i n Table 2 commenced during the spring of 19.71 at a 3.05 by 3.05 metre (10 by 10 foot): spacing. As planned, each family was r e p l i c a t e d 16 to 18 times per plantation i n sin g l e - t r e e p l o t s . I n i t i a l l y the trees were to be planted i n long narrow r e p l i c a t i o n s c o n s i s t i n g of a si n g l e row of 100 tr e e s . However, the shape of the land a l l o c a t e d to the project forced the p a r t i t i o n of each r e p l i c a t i o n i n t o two sections. The a l l o c a t i o n of f a m i l i e s i n the plantations was s y s t e m a t i c — they were planted i n ascending :.numerical order and staggered by seven across r e p l i c a t i o n s . An i l l u s t r a t i o n of the f i n a l family layout within each of the three plantations i s shown i n Figures 1 and 2. Table 2. The three plantations of the Cooperative Progeny Test. Company Location Latitude Longitude Elevation (metres) Tahsis Co. Ltd. Gold River 4 9 ° 5 7 ' 1 2 6 ° 0 7 ' 3 3 5 Crown Zellerbach.-Canada Ltd. (C.Z.) Courtenay 4 9 ° 4 1 ' 1 2 5 ° 1 0 ' 4 5 5 B r i t i s h Columbia Forest Products ( B . C . F ; P . ) Caycuse 4 8 ° 4 8 » 1 2 4 ° 3 3 ' 4 2 5 18 Figure 1. Family p o s i t i o n within each plantation subsectiori' 4 1- ! A C 1 c 1 1 UL 77 £ iiil ^ * iZ Ai S XIII tiy 5 5 . 1 .7 m 1] 3+ 4° "•7 t'4 6" <-/ 7J to St 14 T/ i « li- 11 iS 41 i f t< t i 14- il 11 loo 1 lt 21 2t -X. 4-2 'M fci *1 7 s- S» u It 1 4 ' io n *3 5» >r 41 S-o *7 bj lo 7b %> «i 11 2 s- II K 71 3S 44 SI i-S t4 71 77 34 i « IS 7 t u 2s." J2 >f Us- S-2 sy ls- 7i 7> Si- 71 1? 4 7 ,i| *0 » t 33 4o ui s> to ii 7! 3) lt U |00 3-I i* » l »7 41 *I 61 •7 7I|. fo 17 1> 1 t ? '•«• 2« 3s 42 s-r t2 is 7S- 71 n (4 I 7 lo it =•5 *1 36 *3 1-7 ts '1 7t Si If 3 S " 11 'I 44 5-0 s7 bf 7» 77 S3 lo 1t 1 [ I Z II 5/ 3! 4s s-l ts- 71 U tt 11 17 1" 10 13 ;H 2( 32 *t S-2 S-1 tt 7i 71 ?s- 7i U t 11 14- a. *7 3; +o *7 S3 »• st 13 11 7 u 1 li" *| H >t 41 +? 1% 74 il tl 1» lOo 1 13 It *1 3ff 42 41 « 62 b1 7t- 12 ! ! li- 1 1 14-n *3 So 3t ii S-o it ti 7« 7b 13 »? lt 2. 10 15 It i f 3| '7 U4- s-l »/ 71 77 1» 11 3 il It 11 2=,- 32 11 0 5» s-2 ts- 72 7! is 11 V 4 12 ll 2.o 26 33 H >*b S3 *T (b 7i 71 U 5i f t y \3 18 *l >7 3+ Ho *7 S4 to 'J 74- !. l? 100 t '<(• i l . 22 2j Jo- 4-1 48 « (I 61! 7s- 11 5X n- 1 7 1? 20 l t 42 41 s-t (2 61 7t U «? V S it 21 2+ V So si 63 7« 77 S3 1« 1t 3 1 17 2} IS ?! 56 4<f s-l sS bi- 71 7S S4 11 SI 4 Iff IS 23 *t 5* SI »!• S"2 =7 ts" 72 71 SS n 18 $• if '? 24 >7 !3 V 41 S") t. tt 75 »' li 15 f-j b II 20 Iff i » 3+ n V *4 11 <>1 74- «l il t*- loo 7 ') » l 2t 2f 3s 1*2 ss- n 4! 7i- «* s» is I 1* 2i 2I > >t •*3 4? st tj 6/ It SJ si it 2 1 IS 2S 44 ?o '7 » t 7o 71 S1- lo 17 •5 10 It M 51 58 4s" s-l 55 ts- 71 7J SS 1l 12 4- 11 '7 23 So 33 T) •tl Sf tb 7« If «t l i if S 11 11 =t 31 H HO +7 S3 to t7 13 So s; 1) up t '3 '1 >7 S3 is VI 4! s4 t| tS 7+ •SI 3! 74 / 7 14 20 2-S >1 34 4i « t* 61 7S s» » l 1> 2 ! IS- *l 't -J? 43 To 03 7» 7t 4s 1» 16 1 1 It 21 30 3J H *7 t4 71 17 S4 ll 11 10 n "J 31 •A J( 4s- *J ts- 72 7« iss- 12 1» s- 1  IS i4 32. 4o 4t tt 73 1? « 7! f? t 12 '1 iS 3! 33 41 H-7 ST Co k! 71- « » •14 ISO 7 13 20 2t 31-V 4j il tl 7s- 81 St 1s- 1 f 14 21 27 3S it. 43 M *t u 6? 7t *f •It 2 1 IS 2t 2j ^t 4/ 4* 4-0 "7 03 7« 77 51 1. 77 ) 10 It =3 *) 3? 41 4s 51 s< ti- Tl 7S s*- 'II tt 1  '1 21*. 31 +l; i i ts- 7-2 7? 1J 'M S 11 iS is- 31 f| i-") to u '5 V it '13 I" b 13 '1 at >l US 4» » (.1 '/ 7*- SI ! / 1»- i 1 "* 20 »1 •*) IW ut *> U 75- Si 8! 1s ? i iy » l 25 3t 41 41 S-o St W 6| 7fc SJ SJ ib 1 It 22 1 HI SpacLng 10'xlO' i • I ! i I Source: Sziklai,(1971). Eighteen r e p l i c a t i o n s were included i n the Courtenay 'plantation. Family p o s i t i o n within each plantation — s u b s e c t i o n 2 a 1 9 51 M t4 1° T7 34 1* 1] 4- i». '1 -> »c »' ft "1 f f« (,? II 1! « ' l l It S" <l H 'I 51 t-° sr) I-) tt n 71 ft 6 u '? * V Ji IS "1 f , S t to 7) 1 » (7 IJ |i» 1 13 s» '» 3J "1 s» » tl If It- If ts 1* .' 5 i f tl '1 V- 4» 41 s) ri t2 t| 15 t» I? Is i 1 is ^ *8 »i *1 *7 t; ic it 15 It 1(, 5 te it 2j -<J Jt 4» *» i* S t W 1/ 77 i)i V u i| 17 M- 3» 37 >i) « i rt> t£ 1> 1« W <|i 18 y 12 lS « » 3S Hit t> SJ (• tt Ii Tl St l l 11 t 11 i j Jl. 3< >j US S3 S » 61 67 1* t« S) If I " 7 If 2" ») >} «> It s* «« 68 7s Jl !J li- 1 . 5 l¥ 2j i t » "I "J Ei" ii tl U U «1 It i 1 I' :** 5>~ 4i +3 Si tl tt 7° 17 83 l f 11 •> I" i7 *» »• 3f. H **1 E7 " tf l l n «t 1) Is H 11 U 4 »l >/ ,1 si si 6} tt I 1 1) sr It y B i| ^ >• )| * si t) In-t7 73 !» 15 lit t I) 20 i t 3> »} ul si t. tt" H Jl >7 1* '' 7 •« »l a7 )4 *. *7 ») ti tt t1 75- ii Ij Is- i 1 u " i f 3j 4f 41 cif fcj 17 7e 76 Si «? Ii 3 ) l l 1) >) l i *> «? W H HT II 77 St i« 17 t l» '"I ^ 3» >) M >• fl kt '1 1* 15 <r 11 1? s- i| i | >* * 11 \ fl *7 ts- 7. 73 7f «t l i 1f t ii ') 32 3} «J " SS tt II 7+ <» SI 1) !<• 7 13. >J il' it =1 S"l '1 l i is- $1 if If I g lit 2j 2S 3<|. H( V SS- 60 4 ! 7J It ft IS ! 1 IS « J l -55" 4> M Si 0| tT" 7<f. 77 9> If U 3 10 ' it 25 30 j t *3 4^  st fcz 7» 7S 78 n ll 17 <f " '7 24. 1| J7 sii SJ t> 71 71 . _ 7? JS- U 18 C 12 »J> M « f l *1 t 7 J 17 «t <)) 11 t 1317^  S3 J? H »* *1 lr 13 7! II ?1 H I" 7 1+ 2. 27 H 4-» ul i j t, tt 7* 11 U ti u 1 i 15 i l 25 3S- 41 t l t| t/ 7? So s? t i u 2 1 ' It 72 27 3C d i »1 62 t> 7t >l W lo '1 23 >. ?7 4.3 sr. s t t5 '? .77- «» : Jr 1i IJ 4 11 IJ 2+ 3| 75 <W SH tV 1« 7» 83 t t 12 ^ 12 31 2,7 4.5- s-i n ts- 71 t ) I t «7 13 l«° t 15 2C 2t >J to 4t 5-3 51 12 !» SS-1 It 2| '1 . 31- 41 47 St t. (7 13 ?l St t) 1t 2 t IS 22 IS 3s IVS SS H ts It l» ?| 1» U 5 I ll 23 2-] 3t 4.) HI yt 62 67 IS" 13 SJ i ll 1) f I' .') i t > 37 4<f >« tj 63 7« H ' * J1 11 H » 5| 3$ W Sr| fj H 7/ 77 » 1« i t i 1? <• 11 'I »t » ul s-i P ) ts 1» IS tt 1| In IIIS 7 n 2o -7 33 it. u| s-3 t> tt 13 7![ ?J 12-IS 1 2 IU 21 al 53 m uf lit tl (•/ 7f so JJ 13 1 'it 2 7 I f 21 21) 3S 4 i 47 S3- U (t 11 V «1 l t V V ' > It 23 70 5t 4) 6-0 St (,J> (7 Tt « 1 It 1r 17 •21 31 37 V fl C7 t* 1» -17 13 1l 1t : * 1 1 IS 2S * V s » ' " ' c s 71 Tl » 12 If 2t J J ? | m s) »1 tt 72 i i vi} n i I I S i i 7 5 Vi! £> 5 I ? 3 S fi i ! S Replications S z i k l a i ( 1 9 7 1 ) . 20 METHODS A. Data C o l l e c t i o n and Processinq The f i r s t coordinated e f f o r t to measure family performance within the plantations began during the autumn of 19 75. Company personnel measured each tree f o r t o t a l 19 75 height and height growth i n each of 19 75 and 19 74. Tree condition was also noted at t h i s time according to the c l a s -s i f i c a t i o n given i n Appendix VI. A preliminary examination of the data showed that an appreciable number of trees were i n poor condition (eg. severely browsed and/or forked). I t was therefore decided to include only those trees that were healthy or s l i g h t l y damaged (tree c l a s s i f i c a t i o n 2 or 3 i n Appendix VT) i n a l l the subsequent lea s t squares analyses. An attempt was made i n t h i s study to define uniform microsites ox blocks : within each p l a n t a t i o n . However, the r e s u l t s of t h i s work were abandoned because the blocks proved to be too small, often containing only a few of the 100 f a m i l i e s . I t was: therefore decided that only two large blocks would be defined i n each p l a n t a t i o n — t h e two subsections i l l u s t r a t e d i n Figures 1 and 2. Even with such large blocks, Families 3, 22, 29, 30, 51, 64, 81, 82 and 92 were not represented i n each block because a l l family members were dead or i n poor condition i n one of the blocks. 21 This created an "empty-cell" s i t u a t i o n that was d i f f i c u l t to handle with the computer program a v a i l a b l e . Therefore, when a family possessed no healthy representatives i n a p a r t i c u l a r block, the author accepted the t a l l e s t tree of those i n i t i a l l y excluded because of poor c o n d i t i o n . 4 B. Least Squares Analyses (1) Analysis of tree height and condition of family types and plantations The 100 f a m i l i e s were grouped i n t o the four family types included i n t h i s progeny t e s t — f u l l - s i b , h a l f - s i b (c.b.), h a l f - s i b (p.t.) and c o n t r o l . The mean height of each group was then estimated f o r 19 73, 19 74 and 19 75. In addit i o n , the mean height of each plantation was estimated during these three years. The following ANOVA-model was assumed:: Y. = JUL. +. F. + P . + FP. . + B w .. + FB., , . v + e , . v ljkm A* I j I J k(j) i k ( j ) mdjk). where Y. = tree height i n 1973, 1974 and 1975, ljkm 3 7 » ju^ = l e a s t square mean, F i = family type e f f e c t of the i family type, th Pj = plantation e f f e c t of the j plantation, 4 Variance/components were not c a l c u l a t e d i n the lea s t squares analyses because of time l i m i t a t i o n s and the complexity of such c a l c u l a t i o n s i n t h i s large experiment. 22 FP^^ = family type x plantation i n t e r a c t i o n , B k ( j ) = klock within plantation e f f e c t , F B i k ( j ) = f a r ni- 1-Y x block/plantation i n t e r a c t i o n , e , . v - r e s i d u a l mdjk) Only those trees deemed i n s a t i s f a c t o r y condition were included i n the a n a l y s i s . The number of trees included per family type and plantation therefore offered a second . approach i n which to evaluate family types and plantations. Chi-square t e s t s were performed to t e s t the s i g n i f i c a n c e of differences i n : 1) the number of trees included per family type and 2) the number of trees included per p l a n t a t i o n . The independence of these two f a c t o r s was also tested with a 4x3 contingency table.^ g (2) Analysis of parent performance i n 19 75. ( i ) h a l f - s i b parents. The same ANOVA model used i n the family type and planta-t i o n analysis was again run to evaluate parents of the h a l f -s i b f a m i l i e s . This time, however, a l e a s t squares estimate of mean 1975 height was obtained f o r each of the f a m i l i e s i n the 7 progeny t e s t . I t was assumed that family x block within 5 The reader i s refered to Dixon and Massey (1969) f o r a discussion of contingency t a b l e s , g In t h i s study the performance of a parent's progeny i s assumed to denote the performance of that, parent. 7 Family 100 was dropped from t h i s analysis because of computer program l i m i t a t i o n s . plantation i n t e r a c t i o n ^ F B i ] c ( j ) ) equalled zero, so that the model would reduce? to a siz e enabling analysis with the e x i s t i n g computer program. Assuming an equal pollen c o n t r i b u t i o n i n each of thei h a l f - s i b f a m i l i e s , ranking of the known parents according to mean family height would allow a comparison of parental breeding value. Obviously, the average breeding value of pollen from a natural f o r e s t stand should d i f f e r from that of pollen from a plus tree clone bank. The known parents were therefore compared separately, according to whether the un-known parent originated from a natural stand or a clone bank. ( i i ) f u l l - s i b parents The evaluation of parents from f u l l - s i b crosses demands a model which separates family performance i n t o maternal and paternal e f f e c t s . The following ANOVA model was assumed: Y. ., = F . + M'.. + P, + F P . , + MP.. + B v + e ., ljkmn x j k i k jk m(k) n(ijkm) where Y. ., = tree height i n 19 75 , 1 j kmn ' F ^ = maternal GCA of the i maternal parent, = paternal GCA of the j paternal parent, P K = plantation e f f e c t of the k pla n t a t i o n , . F P . , = maternal x plantation i n t e r a c t i o n , M P J ^ . =• paternal x plantation i n t e r a c t i o n , B . = block within plantation e f f e c t , m(k) ^ 1 e /. ., x = re s i d u a l n(ijkm) S p e c i f i c Combining A b i l i t y was not evaluated because the 24 computer program could not handle the empty c e l l s of the p a r t i a l d i a l l e l crossing design (Appendix I ) . In both the f u l l - s i b and h a l f - s i b parent analyses the parents were f i r s t ranked according to mean performance across a l l three, plantations^ i n 1975. Next, the parents were s t r a t i f i e d i n t o the 13 zones defined by the U.B.C. plus'tree lo c a t i o n map given i n Appendix IV. The objective was to determine i f the parents from a p a r t i c u l a r region ranked to-gether. I f so, geographic oxigin would o f f e r some predic t i o n of parental breeding value. The 19 75 t o t a l height of the parents was next examined within each plantation. The f i r s t objective was to determine those parents which performed c o n s i s t e n t l y i n each of the three plantations. Mean deviation of parent performance among plantations was used to determine these parents. The second objective was to f i n d those parents within each family type-which performed above t h e i r family type mean i n every planta-t i o n . These parents would be most promising because they possess genotypes capable of vigorous growth i n d i f f e r e n t environments. The t h i r d objective was to note extreme examples of genotype x environment i n t e r a c t i o n . 25 Q (3) Juvenile-"mature" c o r r e l a t i o n s The analysis of juvenile-"mature" c o r r e l a t i o n was divided i n t o two; sections according to the two types of data available:: 1. Family performance i n the: nursery during 1969 and 19 70., Thils data, presented by Sigurdson ((19 71):, was; obtained from a random sample of nine observations per family. 2. Individual tree: performance i n 1973, 1974 and 1975 based on the 19 75 height and growth measurements. A mean of eleven trees per family i n each plantation was deemed: i n s a t i s f a c t o r y condition (tree c l a s s i f i c a t i o n 2 or 3 i n Appendix VI) and included i n t h i s a n a l y s i s . ( i ) nursery performance The following multiple regression modei was assumed: Y = bQ + b^X^ + b 2 X 2 + b-jX^ + e where Y = lea s t squares mean family height i n a l l three plantations (19 75), X^ = mean family e p i c o t y l length ((1969), X 2 = mean family hypocotyl length (1969), X 3 = mean family height (1970), e =•• r e s i d u a l Sigurdson (19 71) l i s t s nursery measurements f o r 73 of the 100 famili e s - i n c l u d e d i n t h i s progeny t e s t . The r e s u l t s of t h i s analysis are therefore based on these. 73. f a m i l i e s . g The word "mature" i s put i n quotation marks because the height of Douglas-fir trees l e s s than f i v e years a f t e r planting i s not a mature measurement. 26 ( i i ) plantation performance The following model was assumed to analyze the height data taken from each tree i n the progeny t e s t : Y = b» + b^ X.. +; b„X~ + b_X_ +. b„X. + b^ X.. + 0 1 1 2 2 3 3 4 4 5 5 b 6 X 6 + b 7 X ? + e where Y = 19 75 tree height, 'XI = 19 74 tree height, X 2 =• 19 73 tree height, X 3 , X 4 = dummy variables s i g n i f y i n g p l a n t a t i o n , X 5,X g,X_ = dummy variables s i g n i f y i n g block within plantation, e = re s i d u a l The dummy variables X^ through X^ were included i n t h i s analysis to remove apparent c o r r e l a t i o n s i n tree height over the three years produced by block and plantation s i m i l a r i t i e s . In both the juvenile-"mature" analyses, d i f f e r e n t combinations of independent variables were used to determine those variables which explained the most v a r i a t i o n i n 19 75 2 height. The multiple c o e f f i c i e n t of determination (R ) was used to compare the contribution of the d i f f e r e n t v a r i a b l e s . (4) Interpretation of genotype x environment i n t e r a c t i o n s In an attempt to i n t e r p r e t genotype x environment i n t e r a c t i o n s , i t was hypothesized that the distance separating progeny from t h e i r parents would influence progeny perform-ance. The. greater the distance:, the greater the chance of environmental change and poor progeny performance through a lack of adaptation. To t e s t t h i s hypothesis the following model was assumed: Y = b Q + b^X^ + ^>2X2 + b 3 X 3 + e where Y = progeny x plantation i n t e r a c t i o n i n 19 75, X^ = l a t i t u d i n a l distance separating the progeny from t h e i r parents, = l o n g i t u d i n a l distance separating the: progeny from t h e i r parents, X^ = dummy variable s i g n i f y i n g i f progeny and parents are located i n d i f f e r e n t geographic seed zones, e = r e s i d u a l The dependent variable i n t h i s model was obtained from: 1) the previous f u l l - s i b parent performance analysis and 2) the previous h a l f - s i b parent performance a n a l y s i s . I f a parent was represented i n only one of these two groups, i t would contribute three observations—one observation f o r each of the three plantations. In cases where parents were represented i n both f u l l - and h a l f - s i b f a m i l i e s , the observa-tions from both were included f o r that parent i n t h i s analysi The distances separating the progeny from t h e i r parents were measured from the U.B.C. plus tree l o c a t i o n map (Appendix IV) 28 No attempt to evaluate e l e v a t i o n a l differences between parents and progeny was made, p r i m a r i l y because a l l three plantations were near the same elevation (Table 2 ) . C. Computer Programs Three computer programs were used i n t h i s study. They are l i s t e d as follows: 1 . UBC B M D 1 0 V — t h i s program i s a v a i l a b l e at the U.B.C. Computing Centre. I t was used i n the analysis of family type, plantation and parent performance. 2 . MREG—this program i s a v a i l a b l e at the Faculty of Forestry, U.B.C. I t was used i n the analysis of j u v e n i l e -"mature" c o r r e l a t i o n and genotype x environment i n t e r a c t i o n . 3. P L O T — t h i s program i s also a v a i l a b l e at the Faculty of Forestry. I t was used to plot the dependent var-i a b l e s on each of the quantitative independent variables to check the assumption of l i n e a r i t y . 29 RESULTS AND DISCUSSION A. Tree Height and Condition of Family Types and Plantations (1) Tree height i n 19 73, 19 74 and 19 75 The ANOVA r e s u l t s summarized i n Tables 3,4 and 5 i l l -u strate that a l l f a c t o r s i n the model except family type x block/plantation were highly s i g n i f i c a n t i n each of the three 9 years (oc= .01). Thus i t i s concluded that family type and plantation differences existed i n the progeny t e s t during 19 73, 1974 and 1975. The f i n d i n g that f a m i l i e s did not s i g n i f i c a n t l y i n t e r a c t with blocks (©<= .01) i s of importance. This i n d i c -ates that the within-block s i t e v a r i a t i o n i s r e l a t i v e l y homo-geneous. The ranking of the four family types according to mean height over a l l plantations remained consistent during 1973, 1974 and 1975 (Table 6). The Duncan Multi p l e Range Test demon-strated that a l l means d i f f e r e d from each other during 19 73 -Alpha (o<) i s used i n the text to denote the s i g n i f -icance l e v e l . In vthe tables the author has marked a l l F-s t a t i s t i c s as follows: 1. ** ( s i g n i f i c a n t at the .01 p r o b a b i l i t y l e v e l ) 2. * ( s i g n i f i c a n t at the .05- p r o b a b i l i t y l e v e l ) 3. N.S. ( s i g n i f i c a n t at greater than the .05 p r o b a b i l i t y l e v e l ) Table 3. Analysis of variance, of family type and planta-t i o n height i n 19 73 Source of Va r i a t i o n DF MS" F Family type (F) 3 21 ,,959..00' 78.12** Plantation (PX 2 5,4 78.10 19.49** Block/Plantation (B/P) 3 1,,9 70. 40 7.01** ExP 6 868.20 3V09** Fx(B/P) 9 652..80 2.32* Residual 3 ,,329 281.10 Table 4* Analysis: of variance of family, type and planta t i o n height i n 19 74 Source of V a r i a t i o n DF MS F Family type (El 3 34 ,;412 ..00 60.5 7** Plantation (P) 2 37,185 ..00 65.46** Block/Plantation tB/PX 3 2,683.60 4*72:** ExP 6 1,,95 7.6© 3.44** Ex(B/P) 9 974.60 1.72 N..S. Residual 3,329 568.10 Table 5.. Analysis- of variance: of family type and pXantia-t i o n height i n 19 75 Source of V a r i a t i o n DF' MS: F' Family type (Fl 3 60, .140. .00) 48.72** Plantation (P) 2 139,510..00: 113. .01** Block/Plantation (B/P) 3 8r079.40; 6.54** ExP 6 4,804.50 3.-89** Fx(B/P) 9 2,422.10 1.96* Residual 3,329 1,234.50, and 19 74 (cX = . 0 5 ) . i U However, i n 19 75, the co n t r o l f a m i l i e s and h a l f - s i b (p.t.) f a m i l i e s were not s i g n i f i c a n t l y d i f f e r e n t ( o< = .05). The use of t h i s multiple range t e s t to separate family type means which were c a l c u l a t e d across a l l plantations, i s not s t r i c t l y v a l i d because family type interacted with p l a n t a t i o n . However, the height of each family type i n each plantation was calcula t e d to determine the actual magnitude of the family type x plantation i n t e r a c t i o n s . These c a l c u l a t i o n s showed that: 1) a l l family types that were s i g n i f i c a n t l y d i f f e r e n t i n Table 6 ranked the same i n each plantation and 2) differences among family types were nearly consistent i n each plantation. Therefore, i t i s believed that these Duncan Test r e s u l t s are r e a l i s t i c . Table 6. Mean height i n centimetres of family type i n 19 73, 19 74 and 19 75 19 73 19 74 19 75 Family Type Mean Height (cm.) Duncan T e s t a Mean Height (cm.) Duncan T e s t a Mean Height (cm.) Duncan T e s t a H a l f - s i b (c.b.) F u l l - s i b H a l f - s i b (p.t.) Control 52.58 48.16 41.34 35.40 1 ' 1 \ 1 76.46 70.39 61.04 56.29 1 1 I 1 109.40 101.30 88.14 84.12 1 1 Family types bracketed by the same l i n e are not s i g n i f i c a n t l y d i f f e r e n t at the 0.05 p r o b a b i l i t y l e v e l . Subsequently, the Duncan Multiple Range Test i s termed the Duncan Test. 32 In theory, those f a m i l i e s with two plus tree parents (the f u l l - s i b s and presumably the h a l f - s i b s (c.b.)) should rank f i r s t . H a l f - s i b (p.t.) f a m i l i e s , p o l l i n a t e d by unselected parents, should rank intermediately, followed l a s t l y by the co n t r o l f a m i l i e s . The ranking of the f u l l - s i b s and h a l f - s i b s (c.b.) above the controls was most encouraging and substantiates the above theory. I t i s s u r p r i s i n g that the h a l f - s i b (c.b.) f a m i l i e s performed s i g n i f i c a n t l y above the f u l l - s i b f a m i l i e s . Rossibly, t h e i r greater mean height was a t t r i b u t a b l e to an average parental breeding value which exceeded that of the f u l l -s i b s ' . The v a l i d i t y of t h i s theory i s d i f f i c u l t to determine because only four of the 15 known h a l f - s i b (c.b.) parents were r e p l i c a t e d as f u l l - s i b parents-. The mean breeding value of these four did not exceed the f u l l - s i b parent mean, tending to disprove the above theory. The 1975 height performance of the h a l f - s i b s (p.t.) was disappointing when compared to that of the f u l l - s i b s and h a l f -sibs (c.b.). Hybrid vigour o f f e r s a possible explanation of t h i s height performance d i f f e r e n c e . The means of the f u l l - s i b and h a l f - s i b (c.b.) f a m i l i e s have resulted from the crossing of i n d i v i d u a l s from a l l o p a t r i c populations; whereas, the h a l f - s i b f a m i l i e s (p.t.) were derived from sympatric populations. Assuming that no e p i s t a s i s e x i s t s , t h i s h e t e r o t i c e f f e c t would be a t t r i b u t a b l e to:. 1) d i r e c t i o n a l dominance and 2) d i f f e r i n g gene frequencies among populations (Falconer, 1960). Additio n a l evidence of hybrid vigour has been demon-strated i n a r a c i a l crossing program of Douglas-fir conducted by the B.C.F.S.. During 1968, 13 B r i t i s h Columbia plus trees were used as both seed and pollen parents i n a crossing program which included crosses with pollen parents from Washington, Oregon and C a l i f o r n i a . A f t e r two years nursery growth, 11 of the 13 plus trees produced r a c i a l cross progenies s i g n i f i c a n t l y t a l l e r than the progenies from the same parent crossed with other B.C. plus trees (Orr-Ewing; 1971, 1973). The mean height of the three plantations i s given i n Table 7. During 19 73, l i t t l e d i f f e r e n c e existed among planta-t i o n s . In 19 74, growth i n the Caycuse and Tahsis plantations surpassed the growth i n the Courtenay p l a n t a t i o n . During 19 75, growth i n the Tahsis and Courtenay plantations was comparable. However, the dramatic growth of the Caycuse plantation (40 cm.) was r e f l e c t e d i n i t s height s u p e r i o r i t y at the end of the 19 75 growing season. (2) Tree condition i n 19 75 The number of healthy trees included i n the le a s t squares analysis i s shown according to family type and planta-t i o n i n Table 8. I t i s evident that the h a l f - s i b (c.b.) and f u l l - s i b f a m i l i e s possessed a higher f r a c t i o n of healthy trees than the h a l f - s i b (p.t.) and co n t r o l fam-i-l-i-es. A Chi-square 34 Table 7. Mean plantation height i n centimetres i n 1973, 1974 and 1975 Plantation 1973 19 74 1975 Mean Height (cm.) Duncan T e s t a Mean Height (cm.) Duncan T e s t a Mean Height (cm.) Duncan T e s t a Caycuse Gold JRiver Courtenay 46.00 46.00 41.12 I 1 71.93 68.59 57.61 i i i 112.70 92.20 82.29 1 1 1 i Mean Yearly Height (cm.) 44.3 7 66.04 95.73 Plantations bracketed by the same l i n e are not s i g n i f i c a n t l y d i f f e r e n t at the 0.05 p r o b a b i l i t y l e v e l . analysis showed these family type differences were highly .z 11 s i g n i f i c a n t ( X = 33.83, d.f. = 3 ) . I t can therefore be concluded that, as of 1975, the h a l f - s i b (c.b.) and f u l l - s i b f a m i l i e s possessed a higher f r a c t i o n of trees i n s a t i s f a c t o r y c o n d i t i o n . These r e s u l t s would i n d i c a t e that improved tree s u r v i v a l and health can be expected of f a m i l i e s originating, from two plus tree parents. The second Chi-square analysis demonstrated that differences i n the number of trees included per plantation A l l Chi-square c a l c u l a t i o n s i n t h i s section as shown i n Appendix V. 35 were also highly s i g n i f i c a n t ("X= 106.75, d.f. = 2). To inv e s t i g a t e these plantation differences f u r t h e r , tree condi-t i o n was p a r t i t i o n e d i n t o the four classes given i n Table 9. No great d i f f e r e n c e was indicated i n the 'percent dead' or •percent other* columns. However, the differences i n the •percent forked or browsed' column were large. The small percentage of forked or browsed trees i n the Courtenay planta-t i o n was undoubtedly a t t r i b u t a b l e to the f a c t that t h i s plantation was the only s i t e fenced to prevent deer browsing. These r e s u l t s demonstrate that fencing i s an e f f e c t i v e method of increasing progeny te s t s u r v i v a l and vigour on Vancouver Island. The c a l c u l a t e d Chi-square s t a t i s t i c of the 3x4 contingency table i n v o l v i n g family types and plantations was not s i g n i f i c a n t (y*= 8.60, d.f. = 6). I t can therefore be concluded that these two f a c t o r s were independent i n 19 75 ( i . e . the e f f e c t of family type was consistent within each of the three p l a n t a t i o n s ) . B. Parent Performance (1) Height of progeny from h a l f - s i b parents i n 19 75 The analysis of height performance f o r each of the f a m i l i e s i n the progeny t e s t demonstrated that a l l f a c t o r s i n the model were highly s i g n i f i c a n t i n 1975 (Table 10). 36 Table 8 . Number oft trees included i n the family type and 1 plantation analysis Plantation Fami] r Ly Type Number per Plantation Half-sib, (c.b.) F u l l - s i b H a l f - s i b (p>t.) Control Gold River 1 8 9 / 2 5 6 0 . 7 3 8 6 6 6 / 8 6 4 0 . 7 7 1 175/33:6 0 . 5 2 1 8 2 / 1 4 4 0 . 5 6 9 1 1 1 2 / 1 6 0 0 0 . 6 9 5 Courtenay 2 4 1 / 2 8 8 0 . 8 3 6 8 1 0 / 9 7 2 0 . 8 3 X 2 7 8 / 3 7 8 0 . 735 1 2 7 / 1 6 2 0 . 7 8 4 1 4 5 6 / 1 8 0 0 0 . 8 0 9 Caycuse 1 5 0 / 2 5 6 0 . 5 8 6 4 7 4 / 8 6 4 0 . 5 4 9 1 3 8 / 3 3 6 0 . 4 1 1 6 6 / 1 4 4 0 . 4 5 8 8 2 8 / 1 6 0 0 0 . 5 1 8 Number per: Family Type 5 8 0 / 8 0 0 0 . 7 2 5 1 9 5 0 / 2 700: 0 . 7 2 2 5 9 1 / 1 0 5 0 0 . 5 6 3 2 7 5 / 4 5 0 • : o . 6 i i 3 3 9 6 / 5 0 0 0 0 . 6 7 9 c e l l key: number of trees included / t o t a l number of trees planted f r a c t i o n of ..trees included r Table 9 . Summary of tree condition i n 1975 per plantation,. Plantation % Trees S a t i s f a c t o r y % Trees Dead %, Trees Forked and/or Browsed % Trees other Gold River 6 9 . 5 1 8 . 4 1 1 . 2 0 . 9 Courtenay 8 0 . 9 1 3 . 3 1 .6 4 . 2 Caycuse 5 1 . 8 2 3 . 5 2 1 . 9 2 . 8 Mean % 6.7.4 1 8 . 4 1 1 . 6 2 . 6 37 Table 10. Analysis of variance of family height i n 1975 Source of Vari a t i o n DF MS F Family (F) 98 5 ,690.40 5.08** Plantation (E) 2 129,460.00 115.47**= Block/(P) 3 23/827.00 21.25** FxP 196 1,861.70 1.66** Residual 3 ,079 1,121.20 The predicted means f o r each of the 99 f a m i l i e s are pre-sented i n Appendix I I . The means f o r the three plantations were comparable to the previous family type and plantation analysis and are therefore not presented again. The ranking of h a l f - s i b (p.t.) family parents according to mean 19 75 family height over a l l plantations, demonstrated a substantial height range (Table 11). The highest ranked family (parent 223) was 101.03 cm. as compared to 54.41 cm. from the lowest ranked family (parent 445). Most means, however, f e l l within a 20 cm. range. The Duncan Test produced 12 f i v e groups of s t a t i s t i c a l l y s i m i l a r means. Investigation of parent performance within each of the plantations demonstrated that the higher ranked parents were more va r i a b l e among plantations. For example, the mean _ This t e s t i s not s t r i c t l y v a l i d because the f a m i l i e s interacted with the plantations. I t i s however accepted because of the large number of means i n t h i s a n a l y s i s . Table 11. Mean 19 75 height i n centimetres of h a l f - s i b f a m i l i e s from open-pollinated plus t r e e s . ^ Family Matearnal Region Number Mean Height Giver Mean per Plantation (cm.) Parent of A l l Plantations Progeny (cm.) Mean ^ a Dune an Gold Courtenay Caycuse Mean 1 Test River Plantation Deviation 93 . 223* 11 41 101.03 84.78 86.95 131.38 20.23 91 172 8 33 97.36 70.75 76.66 144.51 31.49 87 114* 8 31 96.35 86.41 83.81 118.84 14.99 84 70* 13 42 96.25 94.02 79.85 114.89 12.42 94 224* 11 36 93J.14 82.22 84.94 112..25 12.47 85 76 6 2-3 /92.55 77.26 76.81 123.58 20:69 82. 63 7 26 91.29 71.94 80.33 121.61 20.21 90 165 12 19 90.47 88..4 5 68.22 114.75 16.18 79 34 12 34 89.93 77.62 78.8.4 113.31 15.59 80 43 7 29 89.57 96.28 84.78 87.66 4.47 95 95 13 33 88.94 76.06 75.06 115.12 17.45 78 28 13 3!0 88.22 80.00 89 ..29 95.35 5.47 81 49 8 18 87.21 67.94 81.97 111.72 16.34 89 162 12 35 82.58 79.20 70.98 9 7.59 10.00 92 177 6 27 81.83 104.88 67.31 73.30 15 . 3:7 83 69 12 22 79.07 73. 74 77.37 86.11 4.69 88 . 153 13 31 78.59 83.75 73.38 78.64 3.47 96 158 8 27 77.02 57.89 60.62 112.56 23.69 97 351 7 28 75.50 66.81 51.26 108.44 21.96 86 96 13 10 68.07 78.00 55.17 71.05 8.60 98 445 9 19 54.41 51.61 66.15 45.50 7.82 Mean 85.68 78.55 74.78 103.72 Parents bracketed by the same l i n e are not s i g n i f i c a n t l y d i f f e r e n t at the 0.05 p r o b a b i l i t y l e v e l . •Parent which performed above the h a l f - s i b (p.t.) mean i n each plantation. deviation of the top nine parents a l l exceeded 12 cm.; whereas, the mean deviation of several of the lower parents was less than 8 cm. The higher ranked parents over a l l plantations can therefore be expected to be more variable among plantations. Parents 223,114,70 and 224 belonged to f a m i l i e s which performed above the h a l f - s i b (p.t.) mean i n every p l a n t a t i o n . \ The author believes that these parents are the most promising i n t h i s family type, because they show a c o n s i s t e n t l y good height performance i n each p l a n t a t i o n . I n t e r e s t i n g l y , parent 172 ranked second according to o v e r a l l plantation performance; however, i t was not included i n the above group because i t performed below the h a l f - s i b (p.t.) mean at Gold River. Parent 177 offered a s t r i k i n g example of parent x plantation i n t e r a c t i o n . I t ranked f i r s t i n the Gold River p l a n t a t i o n — 8 cm. above the next performer. Yet, i t s poor performance i n the other two plantations resulted i n a low o v e r a l l ranking. The ranking of parents of h a l f - s i b (c.b.) f a m i l i e s , according to mean family height over a l l plantations, again demonstrated a wide range of parental performance (Table 12). The Duncan Test separated the means int o four groups of 13 s t a t i s t i c a l l y s i m i l a r means. l^T h i s t e s t i s not s t r i c t l y v a l i d because f a m i l i e s interacted with plantations. Table-12. Me an 1975 height i n centimetres of h a l f — s i b f a m i l i e s from open—pollinated clone banks Mean Height Over Mean Height per Plantation (cm.) A l l Plantations Family Maternal Region Number (cm.) Parent n-F Progeny Mean Dunean a Gold Courtenay Caycuse Mean Test River Plantation Deviation 55 b 36* 12 40 126.17 131.74 108.25 138.53 11.95 56 55 6 39 118.93 100.03 110.56 146.20 18.18 59 36* 13 41 116.92 109.32 97.32 144.12 18.13 60 45 7 39 116.28 101.92 103.27 143.64 18.24 58 134 3 38 115.47 95.50 94.15 156.76 27.53 70 356 13 30 113.82 120.98 83.92 136.55 19.92 63 93 13 40 113.13 130.56. 87.50 121.32 17.08 68 226 3 33 113.09 139.45 84.61 115.22 18.99 57 110 8 42 106.73 83.53 91.75 144.92 25.46 62 166 7 37 105.97 114.86 86.22 116.86 13.18 65 208 1 33 105.36 98.89 100.62 116,58 7.48' 67 220 6 40 105.19 99.09 92.95 123.53 12.23 69 235 1 38 104.63 95.08 105.08, 113.73 6.37 61 160 8 38 93.26 96.62 79.91 103.26 8.90 66 215 1 29 92.11 97.47 87.58 91/29 3.57 64 207 1 24 76.91 73.31 87.69 69.72 7.19 Mean 10 7.75 105.52 93.84 123.89 Parents bracketed by the same l i n e are not s i g n i f i c a n t l y d i f f e r e n t at the 0.05 p r o b a b i l i t y l e v e l . Family 55 was not an open-pollinated f a m i l y — i t was a h a l f - s i b polymix family. Parent which performed above the h a l f - s i b (c.b.) mean i n each plantation. 4 1 The higher ranked parents again tended to be more vari a b l e among plantations than the lower ranked parents. One exception was parent 36 (Family 55) which ranked highest over a l l plantations and also exhibited a low mean plantation d e v i a t i o n . This i n d i v i d u a l was- also the parent of the: only two f a m i l i e s (Family 55 and 59) which performed above the h a l f -s i b (c.b.) mean i n each plantation. Parent 36 therefore appears to be the most promising i n d i v i d u a l of t h i s family type at this. time. No s t r i k i n g examples of parent x plantation i n t e r -action were: noted within t h i s h a l f - s i b group. I t should: be stressed that, when ranking the maternal parents of the: open-pollinated families, an equal paternal contribution was assumed because the pollen parents were unknown. This assumption i s probably v i o l a t e d when seed i s c o l l e c t e d from plus; trees because of provenance differences between pop-u l a t i o n s . In addition, seed f o r the h a l f - s i b (c.b.): f a m i l i e s was c o l l e c t e d from four d i f f e r e n t clone banks and the paternal contribution would have therefore probably d i f f e r e d among these f a m i l i e s . However, i t i s believed that the s i m p l i f y i n g assumption of an equal pollen contribution i s required u n t i l precise estimates of mean pollen parent breeding values are obtained f o r the s p e c i f i c provenances and clone banks that contributed pollen to t h i s pro:geny t e s t . No d e f i n i t e geographic trends were detected which influenced parent performance i n e i t h e r of the" h a l f - s i b groups. In general, the best and worst ranked parents were spread throughout the geographic range of plus tree s e l e c t i o n . The rather poor performance of those h a l f - s i b (c.b.,) parents from region 1 i n Appendix IV was noteworthy. These trees were from a marginal Douglas-fir population on northern Vancouver Island which appears to be g e n e t i c a l l y i n f e r i o r . At t h i s e a r l y stage of the progeny te s t i t i s d i f f i c u l t to make recommendations concerning the roguing of parents based on the height of t h e i r h a l f - s i b f a m i l i e s . Probably, only those parents that performed poorly i n each of the three plantations could be s a f e l y excluded at t h i s time ( i e . h a l f - s i b (p.t.) parents 445 and 96; h a l f - s i b (c.b.) parents 207 and 215). (2) Height of progeny from f u l l - s i b parents i n 19 75 When a l l 54 f u l l - s i b f a m i l i e s were included i n the analysis a singular matrix.resulted and the analysis was impossible. Further i n v e s t i g a t i o n showed that the presence of six f a m i l i e s (38, 39, 43, 45, 46 and 47) was responsible f o r the high c o r r e l a t i o n , and these f a m i l i e s were therefore ex-cluded from the an a l y s i s . The ANOVA summary i n Table 13 demonstrates that a l l main e f f e c t s were highly s i g n i f i c a n t (cx = .01). However, of the parent x plantation i n t e r a c t i o n s , only the maternal i n t e r a c t i o n was s i g n i f i c a n t (oC - .05). I t i s therefore concluded that the paternal parents performed c o n s i s t e n t l y within each pla n t a t i o n . 43 Table 13. Analysis of variance of maternal and paternal GCA i n 19 75 Source of Vari a t i o n DF MS P Maternal GCA (P) 26 3,733.70 3.18** Paternal GCA (M) 12 5,903.30 5.03** Plantation (P) 2 5,055.40 4.31** Block/ (P) 3 15,435.00 13.15** FxP 52 ' 1,571.30 1.34* MxP 24 1,287.00 1.10 N.S. Residual 1,620 1,173.40 (i ) maternal GCA i n 19 75 Although the maternal GCA ranged from 30.48 cm. (parent 61) to 68.13 cm. (parent 125), the Duncan Test produced only three homogeneous groups of means (Table 14). One of these groups included the lowest 25 of the t o t a l 27 maternal 4. 14 parents. Parents 125, 162, 60 and 118 performed above the mean maternal GCA i n each p l a n t a t i o n . These four parents also ranked within the top f i v e parents according to GCA over a l l planta-t i o n s . Although the o v e r a l l GCA ranking of parent 62 i s high, i t s poor performance i n the Courtenay plantation (3 7.02 cm.) d i s q u a l i f i e d i t from t h i s group. Examples of parent x plantation i n t e r a c t i o n show that parents 35 and 63 both ranked high i n the Gold River environment and low i n the more southerly Caycuse environment. Conversely, parent 145 was low ranked at Gold River and mid ranked at Caycuse. — This Duncan Test i s not s t r i c t l y v a l i d because of s i g n i f i c a n t maternal parent x plantation i n t e r a c t i o n . Table-14. GCA i n centimetres-of maternal parentss i n 1975 GCA Over A l l GCA per Plantation (cm.) Maternal Region Number Plantations (cm.) Parent of Progeny GCA Duncan Test a Gold River Courtenay Caycuse Mean GCA dev-i a t i o n among plantations 125* 8 45 68.13 84.20 50.55 69.64 11.72 162* 12 42 66.61 58. 76 43.06 98.04 20.94 60* 3 80 63.31 73.36 57.16 59.38 6.71 62 7 47 59.68 60.99 3^ 7.02 81.03 15.11 118* 8 180 5 7.23 50.23 54.69 66.79 6.3 7 196 11 84 ' 53.66 43.28 31.01 86.69 22.02 82 10 J 94 53.041 50.21 33.81 75.10 14. 71 167 1 40 50^50 39.-10 33.62 78.78 18 .,85 87 13 43 50. m 46.64 49.11 54.34 2.87 175 8 20 49.91 43.93 32.24 73.56 15. 77 224 11 46 48.97 58.60 47.38 40.96 6.41 36 13 125 48.66 56il0 46.55 43.33 4.96 233 7 36 46.99 52.56 49.11 39.31 5.12 25 12 24 46.49 54. 77 30.39 54.32 10.74 114 8 39 46.41 43.61 51. 74 43.85 3.56 35 12 42 41.87 65.62 41.86 18.24 15.80 223 11 43 41.56 44.24 32. 78 47.65 5.85 55 ;6 112 40.29 35.78 40.47 44.62 3.01 153 13 112 40.^ 8 7 34.04 43.80 44. 73 4.54 28 13 72 40.. 72 46.60 29 . 70; 45.87 7.35 172 8 74 38.68 2 7..66 39.19 4.9 .18 7.34 63 7 30 35.22 57.43 33.27 14.96 14.81 45 7 112 33.34 39.62 35.64 24.77 5.72 232 3 36 32.61 39.92 42.95 14.96. 11.77 160 8 91 31.35 40.45 25.84 27.77 6.06 145 8 35 3:i. 28 11.84 30.01 5 2.00 13.81 61 8 38 30.48 49.07 32.33 10.00 13.64 Mean 46.29 48.54 39.90 50.44 cL Parents bracketed by the same l i n e are not s i g n i f i c a n t l y d i f f e r e n t at the 0.05 p r o b a b i l i t y l e v e l . •Parent which performed above the mean maternal GCA i n each plantation. ^ 45 ( i i ) paternal GCA i n 19 75 The Duncan Test showed f i v e sets of s t a t i s t i c a l l y -s i m i l a r paternal GCA means (Table 15). The lowest group of s i m i l a r GCA values contained only six of the t o t a l 13 values; a s i t u a t i o n easier to i n t e r p r e t than the Duncan Test r e s u l t s f o r the maternal GCA means. This s i t u a t i o n was undoubtedly a t t r i b u t a b l e to a larger average number of progeny per pater-nal parent. An i n v e s t i g a t i o n of paternal GCA within each plantation had no s t a t i s t i c a l basis because the paternal GCA x plantation i n t e r a c t i o n was not s i g n i f i c a n t . A b r i e f examination showed that four parents (118, 45, 28 and 92) performed above the mean paternal GCA i n each p l a n t a t i o n . These four were also found within the top ranking four parents according to GCA over a l l plantations. As with the h a l f - s i b parents, no d e f i n i t e geographic trends were noted which influenced f u l l - s i b parent perform-ance over a l l plantations. A good example was the parents from region 8 (Appendix IV) which occured at the top and bottom of both the maternal and paternal GCA rankings. As with the h a l f - s i b parents, the roguing of f u l l - s i b parents based on the GCA*. r e s u l t s i s d i f f i c u l t at t h i s e a r l y stage of the progeny t e s t . The author believes that no maternal parent should be excluded at t h i s time because the Duncan Test showed the lowest ranked parent (61) to be s t a t i s t i c a l l y s i m i l a r Table 1 5 . GCA i n centimetres of paternal parents i n 19 75 Paternal Region Number-GCA Over A l l Plantations (cm.) GCA per Plantation (cm.) Parent of Progeny GCA Duncan3 'Test Gold River Courtenay Caycuse Mean GCA dev-i a t i o n among plantations 118* 45* 28* 92* 60 35 70 1 5 7 134 63 1 7 7 175 1 3 3 8 7 13 13 3 12 13 6 3 7 6 8 13 258 468 157 108 37 187 238 35 36 36 60 76 51 6 4 . s s -e s . 9 8 5 8 . 5 3 •-5 4 . 1 6 ' 5 1 . 0 7 4 9 . 8 0 4 7 . 2 7 4 4 . 1 9 4 4 . 0 6 4 4 . 0 0 3 6 . 4 7 3 2 . 3 9 1 5 . 6 3 6 4 . 1 5 6 7 . 9 4 5 6 . 3 3 4 9 . 8 4 5 9 . 7 6 3 7 . 4 7 5 2 . 2 6 4 2 . 1 2 3 7 . 6 9 4 8 . 3 8 4 7 . 1 6 3 7 . 2 0 3 8 . 1 1 4 9 . 8 6 4 8 . 0 7 4 9 . 1 8 4 4 . 5 5 3 5 . 7 1 3 0 . 1 4 4 5 . 0 7 4 2 . 7 1 3 7 . 6 4 3 5 . 3 5 • 3 3 . 1 6 3 5 . 7 8 3 2 . 9 3 7 9 . 73 7 5 . 9 5 7 0 . 0 8 6 8 . 0 9 5 7 . 74 8 1 . 7 9 4 4 . 4 8 4 7 . 7 4 5 6 . 8 6 4 8 . 3 0 2 9 . 0 8 2 4 . 1 8 - 2 4 . 1 5 b 1 0 . 1 0 1 0 . 6 1 7 . 7 0 9 . 2 9 1 0 . 2 4 2 1 . 3 3 3 . 3 3 2 . 3 7 8 . 5 3 5 , 7 8 7 . 1 3 5 . 4 7 2 6 . 5 2 Mean 4 6 . 6 3 4 9 . 1 1 4 0 . 0 1 5 0 . 7 6 aParents bracketed by the same l i n e are not s i g n i f i c a n t l y d i f f e r e n t at the 0 . 0 5 p r o b a b i l i t y l e v e l . •Parent which performed above the mean paternal GCA i n each plantation. The presence of a negative GCA value i s not uncommon i n t h i s type of lea s t squares a n a l y s i s . 47 to the t h i r d ranking parent (60). Of the paternal parents, the author f e e l s that only the lowest ranking parent (133) could s a f e l y be excluded at t h i s time, because i t performed poorly within a l l three plantations. (3) Maternal e f f e c t s i n 1975 Six parents contributed both male and female gametes i n the crossing design (Appendix I ) . In these s i x cases a GCA estimate was obtained f o r both pollen and egg c o n t r i b u t i o n , thus allowing an i n v e s t i g a t i o n of maternal e f f e c t . Table 16 demonstrates that a trend towards a negative maternal e f f e c t was present i n 1975. Parent 45 offered the most s t r i k i n g example of a lower maternal contribution (-30.64 cm.). The three r e c i p r o c a l crosses shown i n Table 17 allowed the te s t of a d i f f e r i n g maternal c o n t r i b u t i o n . Two crosses were s i g n i f i c a n t l y d i f f e r e n t ( Oc = 0.05). The lower maternal contribution of parent 45 was undoubtedly responsible f o r these s i g n i f i c a n t d i f f e r e n c e s . The small number of parents r e p l i c a t e d i n both sexes prevent one from drawing d e f i n i t e conclusions concerning maternal e f f e c t s i n Douglas-fir. The author b e l i e v e s , however, that the negative trends evidenced here warrant further i n v e s t i g a t i o n . 48 Table 16. - Maternal effects; i n centimetres during 1975 Parent Maternal GCA (cm.). Paternal GCA (cm.) Maternal E f f e c t (cm.) F 45 33i. 34 63.98 -30.64 21.93** 118 5 7.23 64.58 - 7.35 1.54 N..S. 60 635. 31 51.07 12.81 1.45 N.S. 28 40.72: 58.53 -17.81 4.82* 35 41.87 49.80 - 7.93 1.59 N.S. 63 35.22 44.00 - 8.78 1.98 N.S. Table 17. Reciprocal cross differences i n centimetres during 19 75. Par ent Family Height Difference (cm.) F A B AxB (cm.) BxA (cm.) 118 63 101.29 99.81 1.48 0.03 N.S. 118 45 126.97 94 ..21 32.76 18.82**^ 28 45 10:6.17 91.16 15.01 4.01* 49 C. Problems i n the Family Type, Plantation and Earent Analyses Two of the f i v e assumptions which allow v a l i d s t a t i s -t i c a l inference were v i o l a t e d i n the family type, plantation and parent analyses. These two v i o l a t i o n s concern the normal d i s t r i b u t i o n of the unexplained or r e s i d u a l v a r i a t i o n and the a l l o c a t i o n of the f a m i l i e s within the plantations. The d i s t r i b u t i o n of the within c e l l v a r i a t i o n was examined by randomly s e l e c t i n g two f a m i l i e s from each planta-t i o n and constructing a histogram height f o r each family. These d i s t r i b u t i o n s are drawn i n Appendix I I I . In most cases the d i s t r i b u t i o n s did not follow the c l a s s i c bell-shaped curve. The extreme performers were often too frequent. 1^ As Wright (19 75) notes, the systematic a l l o c a t i o n of f a m i l i e s within the plantations could create problems. Envir-onmental s i t e conditions tend to be c o r r e l a t e d across c o n t i g -uous plots.. Therefore, f a m i l i e s always adjacent i n t h i s systematic design should experience higher environmental cor-r e l a t i o n s than those further apart. As a r e s u l t , differences between family means become confounded with environmental e f f e c t s . In addition to these assumption v i o l a t i o n s , the UBC BMD10V computer program presented two problems. F i r s t l y , i t ^ A B a r t l e t t ' s Test of 25 randomly chosen f a m i l i e s was also performed. This t e s t was unable to detect unequal within c e l l variance (oc= .05). 50 was not programmed, to p r i n t the inverse matrix which would allow one to estimate the sampling error about the least squares c o e f f i c i e n t s (the predicted means);. I t was therefore impossible: to test a d d i t i o n a l hypotheses concerning the predicted means once the computer analysis had been completed. Secondly, a f u l l i n t e r p r e t a t i o n of n o n s i g n i f i c a n t f a c t o r s (e.g. paternal GCA x plantation i n t e r a c t i o n ) was impossible because the program only entered each f a c t o r l a s t i n t o the:, model when determining i t s c o n t r i b u t i o n . I t could therefore not be determined i f the f a c t o r remained nonsign i f i c a n t i f entered e a r l i e r . D. Juvenile-"Mature" Correlations (1) Nursery measurements As shown i n Table 18, the three family measurements i n the nursery combined to explain a s i g n i f i c a n t amount of the v a r i a t i o n i n 1975 family height (R 2 = .397). The P a r t i a l -F t e s t indicated that 1970 height and 1969 e p i c o t y l length each added a s i g n i f i c a n t contribution ( oC ~ .01). However, 1969 hypocotyl length was not s i g n i f i c a n t . The:, nonsignificance of hypocotyl length i s f u r t h e r demonstrated by the small 2 drop i n R (.397 to .363) when t h i s variable was excluded from the maximum model. The r e l a t i v e l y large c o n t ri bution 51 Table 18. C o r r e l a t i o n of 1975 family height with 1969 and 19 70. family measurements i n the: nursery Variable(s) 2 R co n t r i b u t i o n E a r t i a l - F Maximum model 0.39751 1969 e p i c o t y l length (X1> -entered f i r s t -entered l a s t 0. 230.3 7* * 0.0.7569**' 8.7** 1969 hypocotyl length (X 2) -entered f i r s t -entered l a s t 0.03354 N.S. 0.0.1946 N.S.. 2.2 N.S. 1970 t o t a l height (X 3) -entered f i r s t -entered l a s t 0.29375** 0.13626** 15.6 * • 19 70 t o t a l height + 1969 e p i c o t y l length 0.36397** 41. 7** """The l e a s t squares equation was as follows: Y = 50.614 + 0.547 X a - 1.384 X 2 + 0.114 X 3 + e. Excluding 1969 hypocotyl length the equation became, Y = 29.649 + 0.526 X^ + 0.113 X 3 + e. The: dependent v a r i a b l e Y was measured i n centimetres. A l l independent variables were measured i n millimetres. of e p i c o t y l length was s u r p r i s i n g , since t h i s measurement was recorded less than four months a f t e r germination. The author believes e p i c o t y l length i s therefore worthy of future consideration when older or mature data becomes a v a i l a b l e . Family s e l e c t i o n , based: only on these three nursery measurements, i s inadvisable. Assuming the c o r r e l a t i o n be-tween the family measurements i n the nursery and i n the 2 plantation remains at thxs l e v e l i n the future, the R f i g u r e i l l u s t r a t e s that only 40 percent of the v a r i a t i o n i n the mature t r a i t can be explained. Therefore^, s e l e c t i o n at the nursery stage could only produce 40 percent of the genetic gain possible from s e l e c t i o n at maturity. (2) Early plantation measurements The l e a s t squares analysis summarized i n Table 19, demonstrated that 19 75 tree height was highly c o r r e l a t e d with 1974 and 1973 tree heights (R 2 = .926). The P a r t i a l -F t e s t s i ndicated that a l l f a c t o r s i n the model were s i g n i f -i c a n t (<X = .01). The va r i a b l e that explained by f a r the greatest amount of v a r i a t i o n i n 19 75 height was 19 74 height. The c o r r e l a t i o n between 19 74 and 19 73 heights was 2 high (R = .817). This high c o r r e l a t i o n was evidenced by 2 the large drop xn R contributxon when exther v a r i a b l e was entered i n t o the model once the other was present (Table 19). 53. Table 19. C o r r e l a t i o n of 19 75 tree height with 19 74 and 1973 tree h e i g h t s 1 Variable(s) 2 R contribution P a r t i a l - F Maximum model 0.92616** 19 74 tree height (X 1> -entered f i r s t -entered a f t e r block and plantation -entered l a s t 0.88712** 0.81747* * 0.23930* *• 11008.7** 19 73 tree, height (x^). -entered f i r s t -entered a f t e r block and plantation -entered l a s t 0.63026** 0.59227** 0.0.1410** 648.5** 19 74 + 19 73 tree height -entered f i r s t -entered l a s t 0.90536** 0.83037** 38200.8** Plantation (X 3,X 4) -entered f i r s t -entered l a s t 0.08155** 0.01935** 890.2** Block (X 5,X 6,X ?):; -entered f i r s t -entered l a s t 0.02089** 0.000631** 28.1** The l e a s t squares equation was as follows:-Y = 5.002 + 1.785 X a - 0.611 X 2 - 6.855 X 3 -0.561 X 4 + 1.266 X 5 + 0.045 x 6 + 1.249 X ? + e. A l l variables were: measured i n centimetres. The dummy variables were coded, x 3 X4 X5 X6 x y Gold River block 1 1 0. 1 0 0 block 2 1 0 -1 0 0 Courtenay block 1 0 1 0 1 0 block 2 0 1 0 -1 0 Caycuse block 1 -1 -1 0 0 1 block 2 -1 —1 0 0 -1 54 The e f f o r t to remove apparent c o r r e l a t i o n s between the height variables: caused by block and plantation sim-2 i l a r i t i e s proved worthwhile. The R values dropped an aver-age of s i x percent once the contribution of blocks- and planta-tions was removed. Therefore, an important source of v a r i a -t i o n , that would have produced overestimates of c o r r e l a t i o n among the height v a r i a b l e s , has been removed i n t h i s model. From these r e s u l t s i t can be concluded that 19 74 height gave a good estimate of 19 75 height. Therefore, i f the previous year's height i s a v a i l a b l e , other j u v e n i l e data w i l l probably add l i t t l e i n the model. E. Interpretation of Genotype x Environment Interactions: The results: i n Table 20 show that the; attempt to i n t e r p r e t genotype x environment i n t e r a c t i o n was: unsuccessful because- the: combined co n t r i b u t i o n of a l l f a c t o r s i n the. model was not s i g n i f i c a n t (oc= .05). Both geographic distance separating progeny from t h e i r parents and. t r a n s f e r across seed zones therefore appear of l i t t l e value f o r p r e d i c t i n g progeny performance at a given l o c a t i o n . Wright (1975) anticipated that attempts to i n t e r p r e t genotype x environment i n t e r a c t i o n would be unsuccessful. He stated that he has r a r e l y been able to i n t e r p r e t t h i s i n t e r -action s u c c e s s f u l l y i n cases where the plantations have been located: within 150 miles of each other. 55 Table. 20. Summary of genotype x environment i n t e r a c t i o n analysis; Variables (s). 2 R Contribution P a r t i a l - F Maximum model L a t i t u d i n a l d i s t a n c e Longitudinal distance Seed zone t r a n s f e r 0.0.05 20 0.00491 0.00112 0.00009 0.338 N.S. 0.788 N.S. 0.056 N..S'. 0.002 N.S. F. H e r i t a b i l i t y Estimation No attempt was made i n t h i s study to derive; h e r i t -a b i l i t y estimates from the within-family c o r r e l a t i o n s i n t h i s progeny t e s t . Work i n t h i s area i s undoubtedly of future i n t e r e s t . However, the author believes that the following problems could bias h e r i t a b i l i t y estimates. F i r s t l y , the objectives of t h i s progeny t e s t were to evaluate the growth performance of progeny from selected' plus trees, not to estimate genetic parameters and h e r i t -a b i l i t i e s . The parents were therefore selected only from the phenotypic extremes i n the population set by the plus tree s e l e c t i o n c r i t e r i a . As a r e s u l t , the parents of t h i s progeny t e s t would be phenotypically, and l i k e l y g e n e t i c a l l y , more s i m i l a r than a random sample of parents. In turn v a r i a -t i o n among the f u l l - and h a l f - s i b f a m i l i e s would be reduced and h e r i t a b i l i t y underestimated.. Secondly, the addi t i v e variance c a l c u l a t e d from open-pollinated h a l f - s i b f a m i l i e s would l i k e l y be biased" upward (Namkoong, 1966; Stonecypher, 1966). This i s pr i m a r i l y a t t r i b u t a b l e to the f a c t that proximally located" trees are l i k e l y to repeatedly p o l l i n a t e the known mater-nal parent. The progeny would therefore contain an addi-t i o n a l likeness due to inbreeding and h e r i t a b i l i t i e s would be overestimated. Estimates of h e r i t a b i l i t y i n t h i s pro-geny t e s t should therefore be r e s t r i c t e d to c a l c u l a t i o n s i n v o l v i n g f u l l - s i b f a m i l i e s . T h i r d l y , the systematic a l l o c a t i o n of f a m i l i e s within the plantations could bias h e r i t a b i l i t y estimates. Each family, or tree with s i n g l e - t r e e p l o t s , i s surrounded by the same common core of f a m i l i e s . The trees of each family therefore experience s i m i l a r competition which should increase within-family c o r r e l a t i o n s and h e r i t a b i l i t y e s t i -mates. Mather and Jinks (1971) discuss^ the importance of randomization i n experiments designed to estimate genetic parameters. They recommend that randomization be employed from the time of seed germination. 1 6 Falconer (1960) notes that t h i s argument does not apply to parent-offspring regressions.. The variance among parents' i s reduced to the same extent as the parent-offspring covariance. Thus, the slope of the regression l i n e and h e r i t a b i l i t y remain unaltered. 57 CONCLUSIONS AND RECOMMENDATIONS The conclusions drawn from t h i s work are summarized below: 1. The excellent performance of the f u l l - s i b f a m i l i e s , and the h a l f - s i b f a m i l i e s from clone banks, demonstrates that meaningful e a r l y genetic gains are possible from the genetic manipulation of Douglas-fir. 2. The poor e a r l y performance of the h a l f - s i b f a m i l i e s from o;pen-pollinated plus trees, i n d i c a t e s that l i t t l e genetic gain can i n i t i a l l y be expected from crossings between trees from sympatric populations. 3. Differences among both the h a l f - s i b and f u l l - s i b parents were s i g n i f i c a n t . Therefore, further s e l e c t i o n within these two groups should produce a d d i t i o n a l genetic gain. 4. Nursery measurements were s i g n i f i c a n t l y c o r r e l -ated with height performance a f t e r seven growing seasons. These measurements are therefore worthy of future considera-t i o n when older data i s a v a i l a b l e . 5. In only one case did the parents from the same geographic area rank c l o s e l y i n progeny performance across a l l plantations. Thus, the geographic l o c a t i o n of a plus tree o f f e r s l i t t l e i n d i c a t i o n of i t s breeding value and the r e s u l t s from progeny t e s t i n g remain important. 58 6.. The geographic distance which separates progeny from t h e i r parents? is;no t correlated: with progeny height i n each; p l a n t a t i o n T h e r e f o r e , , otherr factors: must be: i n v e s t i -gated! before: one can s u c c e s s f u l l y predict progeny/ performance at a. s p e c i f i c l o c a t i o n . The author * s: recommendations; concerning:; future manage-ment and analysis of thx's progeny t e s t are: as follows;;: 1., A l l trees should: be: c l e a r l y tagged to prevent: future i d e n t i f i c a t i o n problems. Many of: the o r i g i n a l stakes are; now l o s t or damaged: and; future tree measurement w i l l probably become more d i f f i c u l t ifc no action i s taken at pre-sent. 2 . Thinning should be undertaken as: soon as; competi-t i o n begins; to af f e c t : tree: vigour. This: would hopefully mini-mize: a d d i t i o n a l within-family c o r r e l a t i o n s caused by the systematic design. 3 . The larger blocks used to remove the within planta-t i o n v a r i a t i o n i n this; analysis; were; note s a t i s f a c t o r y . Future analyses should make: a more r e a l i s t i c attempt to remove s i t e v a r i a t i o n 4 . A standard form f o r assessing tree performance should be used i n the future;.. A form d i r e c t l y readable by keypunchers; would minimize t r a n s c r i p t i o n e r r o r s . 59 5*. The great importance of juvenile-mature cor-r e l a t i o n i n tree-, improvement work makes i t e s s e n t i a l that full', and accurate j u v e n i l e recorders be maintained. Records from the 19 75 plantation measurement should therefore be stared: f o r future use.. 60 SUMMARY The author believes that two findings i n t h i s study e s p e c i a l l y demonstrate the value of t h i s progeny t e s t . F i r s t l y , the low performance of some plus tree progenies show that plus tree s e l e c t i o n , based s o l e l y on i n d i v i d u a l phenotypic values, i s not a r e l i a b l e i n d i c a t o r of parental breeding value. Other more accurate indices:, such as mean progeny performance, are therefore required to achieve maximum genetic gains from each c y c l e of s e l e c t i o n . Secondly, the excellent performance of crosses between clones of geograph-i c a l l y separated plus trees demonstrates the importance of heterosis to juv e n i l e height growth. I f t h i s trend continues to maturity, hybrid seed c o l l e c t e d from clone?banks and seed orchards, would possess a d d i t i o n a l genetic s u p e r i o r i t y over seed c o l l e c t e d from superior phenotypes:; i n natural stands or seed production areas. When one considers the increasing demands on B r i t i s h Columbia's f o r e s t resources, progeny t e s t results-such as these become t r u l y valuable to future f o r e s t management i n t h i s province. 61 LITERATURE CITED Dixon, W. J., and Massey, F. J . , 1969. Introduction to S t a t i s t i c a l A n a l y s i s . 3rd ed. McGraw-Hill Book Co., New York. 638pp. Falconer, D. S., 1960. Introduction to Quantitative Genetics. The Ronald Press, New York. 365pp. G i l b e r t , N., 1967. Additive combining a b i l i t i e s f i t t e d to plant breeding data. Biometrics 23: 45-49. , 19 73. Biometrical I n t e r p r e t a t i o n . Clarendon Press, Oxford. 125pp. Harvey, W. R., 1960. Least-squares Analysis of Data. ARS H-4, U.S.D.A. 157pp. Heaman, J. C , 1967. A Review of the Plus Tree Selection Program f o r Douglas F i r i n Coastal B r i t i s h Columbia. B.C.F.S. Research Note No. 44. Hicks, C. R.,. 1964. Fundamental Concepts i n the Design of , Experiments. Holt, Rinehart and Winston, New York. 293 pp. Kerlinger, F. N. and Pedhazur, E. J . , 1973. M u l t i p l e Regression i n Behavioral Research. Holt, Rinehart and Winston, New York. 5 34 pp. Ledig, T., 1974 An analysis of methods f o r s e l e c t i o n of trees from wild stands. Forest Science 20: 2-16. Lerner, I. M., 1950. Population Genetics and Animal Improvement. Cambridge Un i v e r s i t y Press. 342pp. _, 1958. The Genetic Basis of S e l e c t i o n . Wiley and Sons, New York. 298 pp. L i , J . C , 1964. S t a t i s t i c a l Inference II (The multiple regression and i t s r a m i f i c a t i o n s ) . Edwards,Brothers Inc., Ann Arbor. 5 75pp. 62 Lush, J . II., 1945. Animal Breeding Elans. Iowa State U n i v e r s i t y Press. Ames, Iowa. 443 pp. Mather, K. and Jinks, J . L., 1971. Biometrdcal Genetics. 2nd ed. Co r n e l l U n i v e r s i t y Press. 382 pp. Namkoong, G. 1966. Inbreeding e f f e c t s on estimation of genetic additive variance. Forest Science 12: 8-13. 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? f o r evaluating breeding systems such as seedling oxchards. Silvae Genetica 15: 76-84. Orr-Ewing, A. L. , 1958. Plus?-tree s e l e c t i o n f o r Douglas F i r Seed Orchard. B.C.F.-Si. Research Review. 33pp. , 19 7.1. In t e r - and I n t r a s p e c i f i c Crosses within the Genus Eseudotsuga. B.C.F".S. Research Review. 14-17.pp. , 19 73. Inter- and I n t r a s p e c i f i c Crosses within the Genus Pseudotsuqa. B.C.F.S. Research Review, p. 21. . Orr-Ewing, A. L. and S z i k l a i , 0., 1960. Plus-tree Selection f o r Douglas F i r Steed Orchards. B.C.F.S. Research Review. 26-27 pp. O v e r a l l , J . E. and Spiegel, B, K., 1969. Concerning least: squares analysis of experimental data. Psychological B u l l e t i n 72: 311-322. Peterson, R., 1970. Animal Breeding Notes. Mimeograph. Uni v e r s i t y of B r i t i s h Columbia. 160 pp. Pirchner, F., 1969. Population Genetics i n Animal Breeding. Translated by Pirchner and Krosigk. Freeman and Co;.., San Francisco. 274 pp. Searle, S.R., 1966. Matrix Algebra f o r the B i o l o g i c a l Sciences. John Wiley and Sons, New York. 296 pp. Sigurdson, L. C. , 1971. Early progeny te s t r e s i t s of Douglas-fir. KiS.F. t h e s i s . Faculty of Forestry, U.B'.C. 30 pp. and Appendix. Snyder, E. B., 1972. Glossary f o r Forest Tree Improvement Workers. Revised ed. U.S.D.'.A., Forest Service, Southern Forest Experiment Station. 22 pp. 63 Sprague, G. F., 1967. Quantitative Genetics of Plant Improvement. Plant Breeding A Symposium held at Iowa, Editor K. J . Frey. Iowa State U n i v e r s i t y Press. Ames, Iowa. 430 pp. Stonecypher, R. W., 1966. Estimates of Genetic and Environmental Variances and Covariances i n a Natural Population of L o b l o l l y Pine (Pinus Taeda L . ) . Technical B u l l e t i n No. 5, Southlands Experiment Forest. Woodlands Department, International Paper Co., Bainbridge, Georgia. 167 pp. S z i k l a i , 0. Working Plan of Cooperative Progeny Test of Douglas-Fir. Unpublished information report, 19 71. U n i v e r s i t y of B r i t i s h Columbia. Wright, J . W. Correspondence with Dr. 0. S z i k l a i , Professor, U n i v e r s i t y of B r i t i s h Columbia. 23 September, 1975. , 19 76. Introduction to Forest Genetics. Academic Press, New York. 463 pp. 64 Appendix I. Family numbers of the p a r t i a l d i a l l e l c ross. 1 .28 35 36 45 60 63 70 92 118 133 134 15 7 175 177 213 220 353 25 28 35 36 45 55 6a 61 62 63 82 87 114' 118 125 145 153 160 162 16.7 172 175 176 181 196 215 218 223? 224 232 233 281 356 3 1 48 14 49 15 19 21 3 3 20. 22 50 1 .': 10 36 2 44 53 11 42 41 40 54 3 4 8 9 12 13 5 34 37 16 23 51 17 24 26 27 18 25 39 43 29 28 30 32 - 38 46 35 52 6 7 47 45 1 The: numbers: in the margins signify the plus tree Source* S z i k l a i b t l f 71 35 t h i n t h e b o d Y signify the f u l l - s i b family. Appendix; II.. Summary of, 19 75 family heights i n centimetres Family O r i g i n Maternal Parent Paternal Parent Mean Height Across A l l Plantations: (cm.) Mean Height per Plantation (cm.) Gold River Courtenay Caycuse: 1 c . z ; . 55 45 114.38 116.74 103.77 122.63 2 t i 60. 45 125.59 136.70: 112.44 127.63 35 i t 1.14 45 110.51 109.20 106.42 115.88 4 118 45 126.97/ 125.87 .105.84 148.61 5 I I 125 45 13 2..15 149.69 105.26 141.50 6, I I . 232 45 96.62 105.4 7' 97.65 86.74 7/ H 233 45 111-04: 118.20 103.81 111 ..13" 8 I I . 118! 63 101.29 96.21 96.62 111^04 9 t i 118 92 106.7.3 86.80 109.05 1241..36 10 t i 55 118 100;. 86 87.98 95.34 119.25 11- i t 63: 118 99.81. 118.97' 89 . 75 90.71 12 i t 1.18 134 101.22. 85.10 98.94 119^.61 13 t t 118 175 86..09 82.20 9'7-. 75 78.30 14 t i 28 45 10.6.17 111.90 88.45' 118.16" 15 t i 36. 45 107.9 7 109.15 107:. 76 107.01 16: i t 153 45 108.23 101..58 98.501 124.59 17 t i 160 45 92.88 96..60 75.58 106.46 18 t t 172 45 95.82 90. 75 89.22 107.83 19 I I 36 92 102.83 111..09 94.58 102.80 20 t t 45. 92 92.65 90.14 86.363 101.46 21 i t 36.: 118 118 ..9 9 123.52 100.08 133.44 22 t i 45 118 94 ..21 9:1.28 96.5 7? 94,78 23 i n 153' 118 - 105 ..88 105 .38 100•JO 111.46 24 i t 160 118 9 5..63 111.76 82.63 92.50 25 i f 172 118 10.9.48 91.63 101.29 135.54 26 B.CF'.J?. 162 70 113.89 108.39 94 . 75 13:8.55 2 7 t i 16:7? 70 97495 , 89.23 85.31 119.31 Appendix I I . Continued Family O r i g i n Maternal Parent; Paternal Parent Mean Height Across A l l P1 antations (cm.) Mean Height per - Plantation (cm.) Gold River Courtenay/ Caycuse 28 B.C.F.P. 196 70 97.91 89.36 83.31 121.07 29 it 175 . 133 65.50 79.33 71.80 45.39 30 ii 196 1331 69.40 79.02 7.0.56 58.61 31 I I 25 175 78.93 89.50 72.80 74.50 3?2 I I 196 175; 93.11. 85-.24: 72.70 121.39 33; ti 45J 28 91.16 1021.0.4 87.91 83.54 34 I I 145 28 89.91 65.87 85.81 118.077 35- tt 223 28 103.16 101.24 91.57 116.68 3:6 I K 55 15 7 87.48 78.35 92.81 91.29 37 ti 1531 28 100.28 76.47 105.59 118.77 38 Tansis 215 355 80.38 78.6 7 72.38 90.10 39 it 176 36 96.55 93.22' 74.93 121..50 40; ir 82 177 85.51 90.50 74.52 91.50 41 it 82: 70 102.99 104.02 84.89 120.07 42 it 82 60 104.24 107.70 76.14- 128.87 43 I I 181 36 102.90 100.29 84 .95 123.44 44; it 60 177 102.07 120.25 96.25 89.53 45 it 281 220 101.61 116.55 84 .02 104.24 46 ti 218 220 109 .19 107.35 100.11 120.11 47 356 , 213 106.60 107.32 974,73 114.76 48 Rayonier 28 35 89.09 82.32 61.81 123.12 49 35 35 94.74 103.67 81.62 98.94 50 I I . 6i: 35 80.3.0; 84V08 69.10 87. 73 51 it 160 35 82.29 73.89 67.24 105.75 52 it 224 35 98.71 93.48 84 .14 118.52 53 ti 62 70 106.92 110.66 88.72 121.39 54 it 87 70 977.46 102.88 94.77 94.77 Appendix; II,. Continued; Family O r i g i n Maternal Parent Paternal Parent Mean Height Across A l l Plantations (cm.) Mean Height per Plantation (cm.) Gold River Courtenay Caycuise 55 Tan sis; 36 unknown 1 2 6 . 1 7 1 3 1 . 7 4 1 0 8 . 2 5 1 3 8 . 5 3 56 C.Z.. 55 1 1 8 . 9 3 1 0 0 . 0 3 1 1 0 . 5 6 1 4 6 . 2 0 57 t ! 110 1 0 6 . 7 3 8 3 . 5 3 9 1 . 7 5 1 4 4 . 9 2 58 t l 134 1 1 5 . 4 7 9 5 . 5 0 9 4 . 1 5 1 5 6 . 7 6 59 l l r 26 t i ( 1 1 6 . 9 2 1 0 9 . 3 2 9 7 . 3 2 1 4 4 . 1 2 60 I t 45 t i 1 1 6 . 2 8 1 0 1 . 9 2 1 0 3 . 2 7 1 4 3 . 6 4 6 1 ' I I 160- 9 3 . 2 6 9 6 . 6 2 7 9 . 9 1 1 0 3 . 2 6 62 I t 166 105-977 1 1 4 . 8 6 8 6 . 2 2 1 1 6 . 8 6 63 i t 9 3 I I . 1 1 3 . 1 3 1 3 0 . 5 6 8 7 . 5 0 1 2 1 . 3 2 64 Tahsis 2 0 7 7 6 . 9 1 7 3 . 3 1 877.69 6 9 . 7 2 65 i t 20.8 1 0 5 . 3 6 9 8 . 8 9 1 0 0 . 6 2 1 1 6 . 5 8 66 I I 215 9 2..11 9 7 . 4 7/ 8 7 . 5 8 9 1 . 2 9 6 7 t i 220 105.1.9 9 9 . 0 9 9 2 . 9 5 1 2 3 . 5 3 68 i t 226 1 1 3 . 0 9 1 3 9 . 4 5 8 4 . 6 1 1 1 5 . 2 2 69 i t 235 1 0 4 . 6 3 9 5 . 0 8 1 0 5 , 0 8 1 1 3 . 7 3 70; i.i- 356 1 1 3 . 8 4 1 2 0 . 9 8 8 3 . 9 2 1 3 6 . 5 5 71 c z . unknown 8 9 . 8 8 9 7 . 5 7 6 7 . 1 6 1 0 4 . 9 2 72 I I " 8 8 . 5 9 9 9 . 5 4 6 5 . 4 2 1 0 0 . 8 3 73 Rayonier 70:.34 7 1 . 7 4 6 2 . 9 1 7 6 . 3 9 74 B.C.-P.P. 7 8 . 8 0 6 8 . 4 8 6 4 . 3 9 1 0 3 . 5 4 75 Tahs-is; t t t t 8 6 . 3 7 9 4 . 0 7 6 7 . 1 3 9 7 . 9 2 76 i t 81. .23 7 1 . 2 7 7 6 . 2 3 9 6 . 2 0 77 i f 9 1 . 6 3 7 1 . 3 8 8 4 . 2 5 1 2 4 . 2 7 78 B.C.F. S 28 8 8 . 2 2 8 0 . 0 0 -' 8 9 . 2 9 9 5 . 3 5 79 t t 34 8 9 . 9 3 7 7 . 6 2 7 8 . 8 4 1 1 3 . 3 1 80 i t 4 3 8 9 . 5 7 9 63.. 2 8 8 4 . 7 8 8 7 . 6 6 8 1 i t 49 8 7 . 2 1 6 7 . 9 4 8 1 . 9 7 1 1 1 . 7 2 Appendix IX.« Continued: Family Origins Maternal!. Pa rent-Paternal Parent Mean Height Across A l l Plantations (cm.). Mean Height; per Plantation (cm.). Gold River Courtenay Caycuse 82 B . C.:F . S . e s unknown 9 1 . 2 9 7 1 . 9 4 8 0 . 3 3 1 2 1 . 6 1 83 69 7 9 . 0 7 7 3~. 74 7 7 . 3 7 8 6 . 1 1 84 70 It ! 9 6 . 2 5 94 ..02 7 9 . 8 5 1 1 4 . 8 9 85 " 76 ir 9 2 . 5 5 7,7.26 7 6 . 8 1 1 2 3 . 5 8 86 96 68.0:7 7 8 . 0 0 5 5 . 1 7 7 1 . 0 5 8 7 114 t t - 9 6 . 3 5 8 6 . 4 1 8 3 . 8 1 1 1 8 . 8 4 88 153 7 8 . 5 9 83 . 75 7 3 . 3 8 7 8 . 6 4 89 162- 8 2 . 5 8 7 9 . 2 0 7 0 . 9 8 9 7 . 5 9 90 165 in 9 0 . 4 7 8 8 . 4 5 6 8 . 2 2 1 1 4 . 7 5 9 1 172 in 977.36 7 0 . 7 5 7 6 . 6 6 1 4 4 . 5 1 92 Itr 1.777 8 1 . 8 3 1 0 4 . 8 8 677 .31 73.3a 93 I I - 223 1 0 1 . 0 3 8 4 . 7 8 8 6 . 9 5 1 3 1 . 3 B 94 " 224 9 3 . 1 4 82 . 2 2 8 4 . 9 4 1 1 2 . 2 5 95 II 95 8 8 . 9 4 7 6 . 06 7 5 . 6 5 1 1 5 . 1 2 : 96 1 5 8 7 7 . 0 2 5 7 . 8 9 6 0 . 6 2 1 1 2 . 5 6 97 3 5 1 7 5 . 5 0 66: .81 5 1 . 2 6 1 0 8 . 4 4 98 445 : I I 5 4 . 4 1 5 1 . 6 1 6 6 . 1 5 45 .50* 99 b.r. unknown 8 1 . 7 2 7 9 . 6 0 7 0 . 9 4 9 4 . 6 3 s i g n i f i e s a 27+0 bare root, family which; was selected: from each company's spring; planting; stoeJc. Appendix I I I . . D i s t r i b u t i o n of the within c e l l v a r i a t i o n two randomly selected f a m i l i e s from each plantation 60 80 100 120 140 160 180 200 1975 tree height (cm.) 220 65 85 105 125 19 75 tree height (cm.) 65 85 105 125 145 1975 tree height (cm.)' Appendix I I I . Continued 55 75 95 115 135 155 1975 tree height (cm.) 4 cy 3 c ue 2 0) 1 0 75 95 115 135 155 1975 tree height (cm.) 175 4 3 u c ue 2 o* re 1 0 45 65 85 105 125 145 1975 tree height (cm.) Appendix IV. U n i v e r s i t y of B r i t i s h Columbia plus tree location map Source: Forestry 302 lecture notes- (1966) 72 Appendix V. Chi-square analyses of t r e e condition among; family types;, and plantations 1. Chi-square^ c a l c u l a t i o n f o r family type differences Family/ Type: Observed (OX ExpectedL CE).a (0-E.X2/E. F u l l l - s i b 1950; 1833.84 7.36 Half-sib. (c.b..) 580 543.3:6 2.47 , Ha l f - s i b (p.t.) 591 713.16 20.93 Control 275 305 ..64 3..07 Total 3396 3396.00 Xl= 33.83 b The; expected, numbers of trees; were ca l c u l a t e d 1 from; the number of trees planted within family type times mean s u r v i v a l of a l l family types (0..6792X. V A..01 = 11.34, d.f. • = 3 2. Chi«-square: c a l c u l a t i o n f o r plantation differences Plantation Ob served! (0) Expected"; ( E ) A (O-E;)2/E Gold River 1112 1086.72. 0.59 Courtenay 1456; 1222.56 44;. 5 7 Caycuse 828 1086.72 61.59 To t a l 3396 3396.00 J2= 106.75 b ^ h e expected, numbers of trees were; c a l c u l a t e d from t h e number of trees planted: within plantation times mean s u r v i v a l i n a l l plantations (0.6792);. b ^ j b i = 9 - 2 1 » d ' , f # = 2 73 Appendix V. Continued 3. Contingency table c a l c u l a t i o n to determine i f family type and plantation e f f e c t s are independent Plantation Family Type Observed(OX Expected (E) (0--E))2/E Gold River F u l l - s i b 666 63B.51 1.18 H a l f - s i b (c.b.) 189 189.91 0.00 ; H a l f - s i b (p.t.) 175 193.52 1.77 Control 82 90.04 0.71 Courtenay F u l l - s i b 810 836.04 0.81 H a l f - s i b (c.b.X 241 248.67 0.24 H a l f - s i b (P, t e ) 2 78 25 3.39 2.39 Control 12 7 117.90 0.70 Caycuse F u l l - s i b 4 74 475.44 0.00 H a l f - s i b (c.b.) 150 141.41 0.52 H a l f - s i b (p.t.) 138 144.10 o:.26 Control 66 6 7.. 05 0.02 T o t a l 3396 3:356.o Xz= 8.60 a / X ^ 0 1 = 16.81, d.f. = (plantations - 1)* (family types - 1) ( 74 AppendxxVT.. Tree condition c l a s s i f i c a t i o n used i n t h i s 1 study. Class Tree Condition 0 Dead or missing 1 Weak—probably w i l l not survive 2 S l i g h t l y browsed or forked 3 Healthy 4 Weak and damaged 5 Badly browsed 6 Badly forked 7 Badly forked and browsed; c h l o r o t i c 8 Data not understandable 9 Poor microsite Those trees c l a s s i f i e d as 2 or 3 were included i n the l e a s t squares analyses. 

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.831.1-0075299/manifest

Comment

Related Items