Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Genetic variability in Douglas-fir based on molecular genetic markers and morphological traits Ponoy, Bundit 1993

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

Item Metadata

Download

Media
831-ubc_1993_fall_phd_ponoy_bundit.pdf [ 10.24MB ]
Metadata
JSON: 831-1.0075159.json
JSON-LD: 831-1.0075159-ld.json
RDF/XML (Pretty): 831-1.0075159-rdf.xml
RDF/JSON: 831-1.0075159-rdf.json
Turtle: 831-1.0075159-turtle.txt
N-Triples: 831-1.0075159-rdf-ntriples.txt
Original Record: 831-1.0075159-source.json
Full Text
831-1.0075159-fulltext.txt
Citation
831-1.0075159.ris

Full Text

GENETIC VARIABILITY IN DOUGLAS-FIR BASED ON MOLECULARGENETIC MARKERS AND MORPHOLOGICAL TRAITSbyBUNDIT PONOYB.Sc.(Forestry), Kasetsart University, THAILAND, 1980M.Sc.(Forestry), Kasetsart University, THAILAND, 1983A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDoctor of PhilosophyinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF FOREST SCIENCESWe accept this thesis as conformingto the required standard.THE UNIVERSITY OF BRITISH COLUMBIAMay 1993© Bundit Ponoy, 1993In presenting this thesis in partial fulfillment of the requirements for an advanced degreeat the University of British Columbia, I agree that the Library shall make it freelyavailable for reference and study. I further agree that permission for extensive copying ofthis thesis for scholarly purposes may be granted by the Head of my Department or by hisor her representatives. It is understood that copying or publication of this thesis forfinancial gain shall not be allowed without my written permission. Department of Forest ScienceThe University of British ColumbiaVancouver, CanadaDate: 26May 1993DE-6 (2/88)ABSTRACTGenetic variability in Douglas-fir (Pseudotsuga menziesii (Mirb) Franco) inBritish Columbia was assessed both by analyzing molecular genetic markers at thepopulation level and by evaluation of quantitative traits in parent trees.Genetic variability in the natural populations of Douglas-fir in B.C. selected fromthree geographic regions (coastal, transitional, and interior) was assessed by analyzingrestriction fragment length polymorphisms (RFLP) for chloroplast (cpDNA) andmitochondrial (mtDNA) DNA, and random amplified DNA fingerprinting (RADF)markers for the nuclear genome. Genetic diversity and the degree of populationdifferentiation for cpDNA and mtDNA were estimated at 3-hierarchical levels. Amongthe three regions, highest diversity of cpDNA was observed in the transition populations,although it was not significantly different from the other two regions. In mtDNA, morepolymorphism with no significant difference was observed in interior populations thancoastal and transition populations. Haplotype frequencies were found to be betterparameters for genetic diversity estimates than allelic frequencies with organellar DNA.Nuclear genomes of populations in two geographic regions, coastal and interior, wereevaluated using RADF markers. Allelic frequencies were calculated at each locus andused to estimate genetic diversity and the degree of population differentiation. Higherlevels of within population genetic diversity were obtained than have been reported instudies using allozyme markers. With all three genomes, higher genetic diversities withinpopulations than among populations were observed in all three regions. Severalcombined evolutionary forces are likely to be responsible for the current genetic make-upof B.C. Douglas-fir populations.Quantitative traits were evaluated in 100 families of 17-year old coastal Douglas-fir progenies in progeny tests at three locations, including Caycuse, Courtenay, and GoldRiver. The progeny tests used a systematic single tree plot-design and included fouriiAbstractfamily types --full-sib, half-sib from clone bank (C), half-sib from original plus tree (P)and control. Full-sib and half-sib (C) families had higher survival percentages and fastergrowth rates than half-sib (P) and control families However, lower wood densities wereobserved in full-sib and in half-sib (C) families. Individual tree narrow sense heritabilityestimates in growth traits (0.116-0.234) for half-sib (C) and half-sib (P) families werecomparable whereas heritability estimates in full-sib families ranged from 0.068 to 0.102for growth traits. Moderately high heritabilities in wood density were observed in full-sib(0.256) and in half-sib (C) (0.494) families and lower heritability estimates were observedin half-sib (P) (0.189).Genetic gains were estimated for height, diameter, and volume at differentselection intensities. About 18-30% gain for volume was obtained when selection wasmade at 10% of the best parents in each family type. Age-age correlations when carriedout for height growth, it was determined that field performance at years 5-6 could predictrelative height growth at 17 years. However, when assessed over several age classesregression of Loge of age ratio (LAR) estimates for predicting relative height growthsuggested very low correlation (r2=0.148) and this may not be promising in futureselection.iiiTABLE OF CONTENTSAbstract^ iiTable of Contents^ viList of Figures viiiList of Tables^ xAcknowledgements xiChapter 1 Introduction and Literature Review^ 1^1.1 Introduction^ 11.2 Literature Review 3Chapter 2 Genetic Diversity of Douglas-fir in B.C. Based on Random Amplified^DNA Fingerprinting^ 262.1 Introduction^ 262.2 Materials and Methods 292.2.1 Plant material^ 292.2.2 Isolation of DNA from megagametophytes^292.2.3 DNA Amplification 322.2.4 Polyacrylamide gel electrophoresis (PAGE)^342.2.5 Silver staining of amplified DNA in PAGE 342.2.6 Data analysis^ 352.3 Results and Discussion 362.3.1 Variation and genetic diversity in Douglas-fir populationsusing RADF^ 362.3.3 Partitioning of genetic variability within and betweenDouglas-fir populations^ 392.3.5 Effect of altitudinal position on genetic variation^442.4 Conclusions^ 472.5 Literature Cited 49^Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.0^543.1 Introduction^ 543.2 Materials and Methods^ 573.2.1 Plant material 57ivTable of Contents3.2.2 Isolation and restriction of genomic DNA^ 603.2.3 Electrophoresis and alkaline transfers of genomic DNA.^613.2.4 Labeling cpDNA probe and hybridization 623.2.5 Data analysis^ 643.3 Results and Discussion 653.3.1 Chloroplast DNA variation^ 653.3.2 Estimates of cpDNA diversity 713.4 Conclusions^ 763.5 Literature Cited 77Chapter 4 Mitochondrial DNA Variation of Douglas-fir in B.0^824.1 Introduction^ 824.2 Materials and Methods^ 844.2.1 Plant material. 844.2.2 Isolation, restriction, electrophoresis and transfer ofgenomic DNA^ 844.2.3 MtDNA probes and hybridization. ^ 854.2.4 Data analysis 854.3 Results and Discussion^ 864.3.1 Mitochondrial DNA variation^ 864.3.2 Estimates of mtDNA diversity 924.4 Conclusions^ 954.5 Literature Cited 96Chapter 5 Genetic Parameters of 100 Families of 17-Year Old CoastalDouglas-fir Progenies^ 995.1 Introduction^ 995.2 Materials and Methods 1015.2.1 Background ^ 1015.2.2 Data collection and analyses^ 1065.2.2.1 Analysis of tree growth characteristics and conditionof family types 1065.2.2.2 Analysis of parent performance^ 107Half - sib families 107Full - sib families 107vTable of Contents5.2.2.3 Estimation of heritability^ 1095.2.2.4 Genetic gain^ 1105.2.2.5 Age-age correlation 1105.3 Results and Discussion 1125.3.1 Survival and growth performance^ 1125.3.2 Growth performance and condition of family types^ 1165.3.3 Parent performance^ 1255.3.3.1 Half-sib (C) families^ 1255.3.3.2 Half-sib (P) families 1275.3.3.3 Full-sib families 1295.3.4 Genetic gain^ 1385.3.5 Age-age correlation 1435.4 Conclusions 1465.5 Literature Cited^ 148Chapter 6 Conclusions and Recommendations^ 150Appendices^ 157viLIST OF FIGURES2.1^Approximate locations of coastal and interior Douglas-fir populationsused in evaluation of genetic diversity^ 302.2^Effect of altitude on genetic diversity of coastal Douglas-fir populations^463.1^Approximate locations of coastal, transitional, and interior Douglas-firpopulations used in studying cpDNA variation of the species in B.C.^583.2^Lodgepole pine restriction site map^ 635.1^Mating design for controlled pollination families of coastal Douglas-fir,including 33 maternal parents, 17 paternal parents and 54 crosses^ 1025.2^Location of coastal Douglas-fir plus trees and three progeny test plantations ^ 1035.3^Single-tree plot planting design of coastal Douglas-fir progeny test^ 1055.4^Growth characteristics of four different family types of coastal Douglas-firprogenies^ 1185.5^Combining ability and ranking of maternal and paternal parents based onheight^ 1335.6^Combining ability and ranking of maternal and paternal parents based ondiameter^ 1345.7^Combining ability and ranking of maternal and paternal parents based onvolume^ 1355.8^Combining ability and ranking of maternal and paternal parents based onpilodyne measurement^ 1365.9^Regression of age-age correlation for height growth of coastal Douglas-firprogenies to predict age ratio for selection^ 145viiLIST OF TABLES1.1 Species included in the genus Pseudotsuga ^ 31.2 Inheritance of organelle DNA in conifers and some forest tree species^ 132.1 Description of population location and sources of seed samples usedfor studying of genetic variation at molecular level in Douglas-fir^ 312.2 List of arbitrary primers showing sequence and G+C content andnumber of segregating loci used in the study^ 332.3 Genetic variability at 97 loci in all populations of Douglas-fir in B.C. 382.4 Genetic variability in Douglas-fir based on RADF and isozyme markers ^ 392.5 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimated from 97 RADF loci^ 402.6 Genetic diversity statistics (Nei, 1973) of Douglas-fir based on RADF andisozyme markers^ 432.7 Effect of elevation on genetic diversity of coastal Douglas-fir populations ^ 453.1 Sources of needle and seed samples used for studying of cpDNA variation inDouglas-fir^ 593.2 Polymorphisms observed in cpDNA of Douglas-fir after digesting with5 different enzymes and hybridizing with 4 different probes sets ^ 673.3 Haplotype frequency of coastal Douglas-fir populations based onmtDNA probes^ 683.4 Haplotype designation for cpDNA analysis^ 693.5 Haplotypes observed from the RFLP analysis of cpDNA^ 703.6 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimation based on allelic frequencies of cpDNA^ 713.7 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimation based on haplotype frequencies of cpDNA^ 72viiiList of Tables4.1 Polymorphisms observed in mtDNA of Douglas-fir after digesting with5 different enzymes and hybridizing with 3 different probes^874.2 Haplotype frequency of coastal Douglas-fir populations based onmtDNA probes^ 894.3 Haplotype designation for mtDNA analysis^ 904.4 Haplotypes observed from the RFLP analysis of mtDNA^ 914.5 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimation based on allelic frequencies of mtDNA^ 924.6 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimation based on haplotype frequencies of mtDNA^ 935.1^Site information for three coastal Douglas-fir progeny test plantations^ 1045.2 Structure of the analysis of variance and variance components estimatesfor half-sib and full-sib families^ 1085.3 Coefficients of variance components for variance estimates^ 1085.4 Analysis of variance for survival of 100 families 17-year old coastalDouglas-fir progenies across three test sites ^ 1125.5 Analysis of variance for survival of 100 families 17-year old coastalDouglas-fir progenies across three test sites ^ 1135.6 Analysis of variance of growth variables of 100 families of coastalDouglas-fir progenies^ 1145.7 Estimation of growth performance of 100 families of Coastal Douglas-firprogenies at three test sites^ 1155.8 Comparison of height growth of coastal Douglas-fir progenies over threetest sites from 1973 to 1989 ^ 1155.9 Analysis of variance and growth characteristics of 4 family types of coastalDouglas-fir progenies^ 1165.10 Mean height of 4 family types in 1973, 1974, 1975, 1977and 1989^ 119ixList of Tables5.11 Estimation of growth performance of 4 family types of coastal Douglas-firprogenies for Caycuse plantation^ 1205.12 Estimation of growth performance of 4 family types of coastal Douglas-firprogenies for Courtenay plantation^ 1215.13 Estimation of growth performance of 4 family types of coastal Douglas-firprogenies for Gold River plantation^ 1225.14 Summary of family types ranking of 17-year old coastal Douglas-firand average growth characteristics for all three plantations ^ 1245.15 Analysis of variance of growth variables of half-sib (C) families^ 1255.16 Genetic and phenotypic variances, phenotypic means, and heritabilitiesof growth variables of half-sib (C) families^ 1265.17 Analysis of variance of growth variables of half-sib (P) families ^ 1275.18 Genetic and phenotypic variances, phenotypic means, and heritabilitiesof growth variables of half-sib (P) families^ 1285.19 Analysis of variance of growth variables for full-sib families ^ 1295.20 Genetic and phenotypic variances, phenotypic means, and heritabilitiesof growth variables of full-sib families ^ 1305.21 Analysis of variance of growth variable of maternal and paternalgeneral combining abilities for full-sib families ^ 1315.22 Summary of paternal and maternal parents ranking of coastal Douglas-firbased on general combining ability^ 1375.23 Genetic gain and percent gain of half-sib(C) family after selecting based onbest families at different intensities and traits^ 1395.24 Genetic gain and percent gain of half-sib (P) families after maternal parentselecting at different intensities and traits^ 1405.25 Genetic gain and percent gain of full-sib families after maternal parentselecting at different intensities and traits^ 1425.26 Correlation coefficients of height growth at different ages^ 143xACKNOWLEDGEMENTSI am very grateful for the guidance and assistance provided by my supervisorycommittee of Dr. J.E. Carlson, Dr. 0. Sziklai, Dr. Y.A. El-Kassaby, which has beeninvaluable in the culmination of this research. Thank is owed to Dr. G. Namkoong forthe final comment and correction of the thesis.I am deeply indepted to the ASEAN-Canada Forest Tree Seed Centre Project(CIDA) for their financial support of this research, Royal Forest Department (Thailand)for giving the opportunity to the study. A special thank goes to the former director of thecentre, Mr Pisal Wasuwanich, for his early support and enthusiasm for the study.Assistance in collecting data by Canadian Pacific Forest Product Ltd., Tahsis andBritish Columbia Forest Product Ltd. is greatly appreciated. I am also thankful for theassistance with the material collection provided by Mr. C. Heaman (Forest ScienceResearch Branch Station, Victoria, B.C.), Mr. Jack Woods (Cowichan Lake ResearchStation, Mesachie Lake, B.C.), and Mr. Frank Young (Kalamalka Research Station &Seed Orchard, Vernon, B.C.). Constructive reviews and comments of all or portion ofthis thesis and data analyses provided by Dr. Y.-P. Hong, are greatful. I also thank to allcolleagues in Dr. Carlson's lab for their advices and supports.I would like to thank my parents, brothers and sisters for their loves and supports.Finally, a special thank goes to Wannapa Prasirtsuk for her unfailing support during mytime at UBC.This thesis represents the efforts of many, however, the mistakes remain my own.xiCHAPTER 1INTRODUCTION AND LITERATURE REVIEW1.1 INTRODUCTIONDouglas-fir (Pseudotsuga menziesii (Mirb) Franco.) is an economically importanttree species which occurs throughout western North America (Fowells, 1965).Conventional methods for tree improvement and propagation of Douglas-fir generallyinvolve selection of phenotypically superior trees from natural stands (Silen, 1978)followed by preservation of genetic material, seed orchard establishment and selectionsthrough the evaluation of progeny testing (Heaman, 1967). Increases in the productivityof Douglas-fir have been achieved through plantation establishment and nitrogenfertilization since 1960 (Farnum et al., 1983). However, growth rate is a complex traitlikely under the control of many genes. Obviously, genetic improvement and propagationof Douglas-fir using seeds are slowed by its long life cycle. Growth traits usually requireone to several decades for adequate expression. As a result, searching for ways to reducethe number of years required for a generation of selection and breeding has beenattempted.At present, the application of molecular biological techniques, RFLPs andRAPDs, is being considered in forest tree improvement programs to assist in reducing thetime in advanced generation selection processes (Neale et al., 1992; Williams and Neale,1992). At the molecular level, the use of restriction endonuclease digestion of totalgenomic DNA followed by hybridization with a radioactively labeled probe yieldshybridizing fragments of sizes that may vary between genotypes. This form ofRestriction Fragment Length Polymorphism (RFLP) has become popular and has beenused extensively in genetic analysis. The RFLP procedure has provided much valuableinformation on genetic structure of species from studies such as inheritance, linkage1Chapter 1 Introduction and Literature Reviewanalysis, and genome mapping. Cloning of RFLP markers linked to traits has enabledindirect selection strategies for plant improvement programmes.Another recent development, the polymerase chain reaction (PCR) hasrevolutionized many standard molecular biological techniques. One variation on theoriginal PCR procedure generates a specific class of molecular markers termed RandomlyAmplified Polymorphic DNAs (RAPDs) (Williams et al., 1990). This new developmenthas the advantages of being technically simple, quick to perform, requires only smallamounts of not high quality DNA and involves no radioactivity. RAPDs are well suitedfor application in plant breeding, population genetics and genetic diversity. Currently,RAPD markers are gaining wide acceptance and the procedure has been optimized for usein many applications. There are several names and acronyms given to this single-primerDNA amplification procedure, all of which are all well suited to the technique: Randomlyamplified polymorphic DNAs (RAPDs) (Williams et al., 1990), arbitrarily primedpolymerase chain reaction (AP-PCR) (Welsh et al., 1990), DNA amplificationfingerprinting (DAF) (Caetano-Anolles et al., 1991), and random amplified DNAfingerprinting (RADF) (Hong, unpublished method).In this thesis, the author assessed several different ways to study genetic variationof Douglas-fir in B.C. The first objective of the study was to evaluate the variation in thespecies using molecular markers. Molecular genetic techniques, RFLP and RADF, wereused to analyze genetic diversity of Douglas-fir in B.C. from three geographic regions,including coastal, transitional, and interior. The two cytoplasmic genomes, chloroplastand mitochondria, were evaluated using RFLP markers and the nuclear genome wasinvestigated through RADF analysis. Another objective of the thesis was the evaluationof selected plus trees in a 17-year-old progeny test of coastal Douglas-fir from differentfamily types by the quantitative genetics approach.2Chapter 1 Introduction and Literature Review1.2 LITERATURE REVIEW1.2.1 Douglas-firThe genus Pseudotsuga consists of eight species (Table 1.1) distributed inWestern North America and Eastern Asia (Chengde, 1981; Hermann, 1982). Within thegenus only Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) is widely distributed or ofmajor economic importance. It is a dominant component of forests throughout WesternNorth America (Fowells, 1965).Table 1.1 Species included in the genus Pseudotsuga (cf. National Council on GeneResources, 1982; Hermann, 1982)Common Name^Scientific Name^RemarksSubject SpeciesDouglas-firWild RelativesBig cone Douglas-firFormosan Douglas-firJapanese Douglas-firChinese Douglas-firForret's Douglas-firSpecies under EvaluationShort -needle Douglas-firPseudotsuga menziesii^North AmericaPseudotsuga macrocarpa^North AmericaPseudotsuga wilsoniana^TaiwanPseudotsuga japonica^JapanPseudotsuga sinensis^ChinaPseudotsuga forrestii ChinaPseudotsuga brevifolia^ChinaPseudotsuga gaussenii^China3Chapter 1 Introduction and Literature ReviewDouglas-fir was first introduced into England and it has been planted in manyparts of continental Europe (Frothingham, 1909) and elsewhere. Two species ofPseudotsuga are native to western North America but only one species (Pseudotsugamenziesii) with two forms, coastal and interior, is found in Canada (Hosie, 1969). Itsnative range, extends from latitude 19°N (central Mexico) to 55°N (central BritishColumbia) over a distance of nearly 3,200 kilometers with elevations from sea level toover 3,200 m (Schmidt, 1960) and longitudes from 98° to 125°W (On-Ewing, 1975) ornearly 1,600 kilometers from the pacific ocean inland. The coastal form of Douglas-fir,which is the largest and most important, grows on the islands and mainland of the westcoast, whereas the interior form is distributed throughout the Rocky Mountain area fromSouthern Alberta and central British Columbia south through the western United Statesinto Mexico.Most systematic studies in Pseudotsuga have centered on cytology because of theunusual karyotype of Douglas-fir, whose 13 chromosomes comprise one of the fewdeviations from the haploid component of 12 typical of the Pinaceae (Christiansen, 1963;Thomas and Ching, 1968; Doerksen and Ching, 1972; El-Kassaby et al., 1983). Sziklaiet. al. (1985) used synthetic karyotype for numerical studies of similarities among speciesin the genus. Based on canonical analyses of arm lengths, their studies indicated thatDouglas-fir could be closely related to either a North American (Pseudotsugamacrocarpa (Vasey) Mayr) or an Asian species (Pseudotsuga gaussenii Flous).Douglas-fir (Pseudotsuga menziesii) contains a somatic chromosome number of2n=26 consisting of five metacentric, six submetacentric and two telocentric pairs(Colangeli et al., 1982.). Pseudotsuga menziesii has two recognized varieties, P.menziesii var menziesii (coastal) and P. menziesii var glauca (interior). Although the twovarieties do interbreed, they differ in both growth rate and size at maturity and also differ4Chapter 1 Introduction and Literature Reviewin botanical and morphological characteristics (Hermann and Lavender, 1990). Theinterior form is stockier and seldom exceeds 45 m. in height compared to the 50 to 70 m.and occasionally even 100 m. of coast form. It has a short tapering trunk and crown withlong limbs. The coastal form commonly pioneers into wetter parts of the coast forestregion, being less tolerant of shade than most other species with which it associates. Theinterior form, on the other hand, is more shade tolerant and more cold hardy (Hermannand Lavender, 1990). Chemical (von Rudloff, 1972; Li and Adams, 1989) andcytological (El-Lakany and Sziklai, 1971; De-Vescovi and Sziklai, 1975) investigationshave also shown differences both between and within the two varieties.1.2.2 Genetic variation studiesPlants respond to the environment in two ways. First, individual plants mayrespond by changes in morphological, physiological and biochemical characteristics.This phenomenon is often called "phenotypic plasticity" (Bradshaw, 1965). Plasticity isan aspect of phenotype that may be under genetic control and modifiable by selection(Khan and Bradshaw, 1976; Jain 1978). Populations of plants may change in geneticcomposition--genotypes with a high fitness for the environment will be maintained, whileless suitable genotypes will disappear. In specialized and stable environments, geneticdifferentiation may lead to maintenance of only a limited number genotypes, with a highfitness for the environment. In a more variable and less predictable environment, a largergenetic variation will be maintained on with genetic plasticity in the populations (Kuiper,1990).Genetic variation in plants generally increases as life forms progress from shorterto longer-lived species. A multidimension matrix for classifying population structurewithin life forms can be defined according to distribution, density, mating system, and5Chapter 1 Introduction and Literature Reviewpollen vectors. Long generation time and wind-pollinated outcross mating systems tendto have high levels of genetic variation, mostly within populations (Hamrick et al., 1992).In contrast, annual herbaceous species, primarily selfing, have lower levels of variation,and it lies mostly between populations. (Hamrick et al., 1992; Woodruff and Gall, 1992).Genera and species can differ greatly in their variation patterns. This issometimes referred to as their genetic architecture. Genetic variation can be assessed atvarious levels; obviously, variation among species is the highest level within a genus.Variation within species is the level with which tree breeders are most concerned (Zobeland van Buijtenen, 1989). The within-species differences can be subdivided further intogeographic variation, stand-to-stand variation, variation among individual trees andvariation within individual trees (Zobel and Talbert, 1984).The determination of the variation present in natural stands or plantations is a firststep in studying a species. Genetic correlations are valuable for multi-trait and indirectselection methods. The latter situation is quite general, because selection is rarelyconducted in the same environmental conditions as production. The genetic correlationbetween the same trait measured in two environments is widely used to express genotypex environment interaction. Heritability is the proportion of the total phenotypic variancedue to gene variance and is defined in two ways with different purposes (Crow, 1976).Heritability in the broad sense is an attempt to measure the importance of all geneticvariances as a ratio of the total variance, whereas heritability in narrow sense is a ratio ofadditive genetic variance to total variance. Most heritability estimates in the forestgenetics are for narrow sense heritability, because most tree improvement programs areaimed at improving general combining ability and utilize only the additive geneticvariance. However, the heritability estimates are applied only to a particular population6Chapter 1 Introduction and Literature Reviewgrowing in a particular environment and at a particular time (Cotterill and Dean, 1988;Zobel and Talbert, 1984).Such methods used for evaluating genetic variation based on phenotypiccharacters, or the expressed products of genes have proven to be very useful. However,they do not provide a complete picture of genetic stability. These characters may not beexpressed uniformly, as their expression may be related to environmental or physiologicalfactors (Potter and Jones, 1991). Therefore, both plant morphology and patterns ofexpressed proteins of genetically identical plants may be different if the plants are grownin different environments, or they are analyzed at different stages of growth and whendata from different geographical areas are compared. In addition, variation in anindividual's phenotype may be determined not only by the genotype and environment ofthat individual but also by maternal effect (Roach and Wulff, 1987), due to unequalgenetic or dosage contributions from each parent to the phenotype of its offspring. Threedifferent classes of maternal effects--cytoplasmic genetic, endosperm nuclear, andmaternal phenotypic--were classified by Roach and Wulff, (1987). First, cytoplasmicgenetic maternal effects are derived from the fact that organelles--plastids, andmitochondria--can be directly transferred from maternal plant to the offspring duringovule formation and development. This transmission is independent of nuclear genes andcontributes to heritable variation in both qualitative and quantitative traits in plants.Second, plant endosperm always contains more doses of maternal than paternal genes.The endosperm contains enzymes important for germination and is also the source ofnutrients for the developing embryo; therefore, the maternal parent has a more importantrole in determining this nutrient source. Third, the phenotype resulting from theenvironment or genotype of the maternal parent may be influenced by structure orphysiology. The tissue surrounding the developing embryo and endosperm are all7Chapter 1 Introduction and Literature Reviewmaternal. These tissues eventually form seed structures that are important determinantsof seed dormancy, dispersal, and germination. The variation in these traits can eveninfluence the mature phenotype of an individual. Different plant characters aredifferentially sensitive to maternal conditions and these may be expressed at differentstages of development (Wulff and Bazzaz, 1992). The most direct study of unequalcontribution by maternal and paternal to the phenotype of the offspring is throughreciprocal crosses. Any different of reciprocal pairs will be due to a maternal or paternaleffect and maternal effects will be indicated if there are greater differences betweenmaternal families than paternal families. However paternal effects have been found onlyin a few species (Tilney-Bassett, 1975). And maternal effects can have a significanteffect on the phenotype of an individual, therefore, this type of variation will be importantfor evolutionary and ecological study (Roach and Wulff, 1987).1.2.3 DNA variation in plantsDNA resides in three separate compartments in the plant cell--the nucleus,chloroplast and mitochondria. Most of the genetic information is contained in the nucleusand this had been thought to be the only site of hereditary information. About 25 yearsago it was realized that both chloroplast and mitochondria contain their own DNA, withimportant functions in photosynthesis and respiration (reviewed in Strauss et al., 1989).In addition to these functions, organelle DNA plays a role in other valuable characters,such as mutant genes in chloroplasts with enhanced resistance to herbicides, whereasDNA in the mitochondria can cause male sterility. There have been numerous studies ofgenetic variability in nuclear DNA in plants through gene products such as morphology,allozymes, and terpenes; however, little was known about variability in organelle DNAuntil molecular genetic methods were developed recently.8Chapter 1 Introduction and Literature Review1.2.3.1 Nuclear DNA (nDNA)Nuclear DNA of plants reveals an enormous variation in amount of DNA perhaploid genome (Mantell et al., 1985). There is no obvious correlation between the sizeof a plant's genome and its biological complexity, and it would be expected that thenumber of different genes is similar in all higher plants (Boffey, 1991). In conifers, thenuclear genome contains roughly 10 -11 grams DNA per haploid nucleus or approximately10 10 nucleotide base pairs (Miksche, 1971). The bulk of the plant nuclear genome iscomposed of repeated sequences, often with very high copy numbers. In pea, forexample, only about 15% of the nDNA consists of sequences present as single copies orin low copy number. The nuclear genome contains sequences that have different rates ofevolution (Schaal et al., 1991) and gene linkages on chromosomes appear to bemoderately conserved even across diverse conifer genera (Conkle, 1992). Lapitan (1992)concluded that low copy sequences are generally conserved in genomes of relatedspecies, whereas repeated sequences that may be closely juxtaposed to such low copysequences are rapidly turning over; and thus genomes have mechanisms that allow somesequences to undergo amplification, depression, and deletion, while maintaining thecomposition and order of neighboring low copy sequences. In Douglas-fir, El-Lakanyand Sziklai (1971) reported that the nuclear volume and DNA content correlatedsignificantly with latitude of seed sources. Coastal provenances had larger nuclearvolume and greater amounts of DNA than inland ones, and they also concluded thatcoastal and inland Douglas-fir may belong to different populations with the same patternof variation.In conifers, major barriers to the application of genetic engineering are the lack ofdefined loci (Sederoff and Ledig, 1985), and knowledge about genetic controlmechanisms (Olsen, 1988). The existence of many genes that affect valuable properties9Chapter 1 Introduction and Literature Reviewof trees, usually as a group of additive genetic factors, has been inferred by techniques ofquantitative genetics. Using isozyme markers, Ledig and Conkle (1983) have reported on45 to 60 loci and have linkage maps for many of these. However, none of the isozymeloci have been directly related to valuable traits. Recently. DNA-based genetic markershave been developed to overcome some limitations of isozymes, including restrictionfragment length polymorphism (RFLP) and polymerase chain reaction (PCR) assays. Thecontribution of nDNA markers would permit population genetic analyses of any region ofa genome. To date, few nuclear RFLP analyses have been reported in forest trees(reviewed in Wagner, 1992).1.2.3.2 Chloroplast DNA (cpDNA)Almost all eukaryotic cells contain both nuclear and mitochondrial DNA, butplant cells have a third genome, chloroplast DNA, contained in plastids. Although someof the sequences found in plastid genes are also located in the nucleus (Stern and Palmer,1984) and in the mitochondrial genome (Kemble et al., 1983), it seems that most genesare expressed only in plastids (Boffey, 1991). Some of the genes are of particular interestto the biotechnologist, because they could be manipulated to improve crop yield, orresistance to a herbicide (Grierson and Covey, 1988.) In view of the major roles playedby plastid-encoded proteins in photosynthesis, it follows that the plastid genome isessential for plant survival (Boffey, 1991). A typical chloroplast genome consists of adouble-stranded circular DNA, 120 - 200 kbp in length depending on the species (Palmer,1985). Chloroplast genomes have conservative rates of both structural change andnucleotide substitution, which facilitates interpretation of restriction fragment phenotypesin terms of mutational changes (Palmer, 1988). The rate of chloroplast DNA evolution isabout half that of nuclear DNA (Schaal et al., 1991). The degree of diversity of cpDNA10Chapter 1 Introduction and Literature Reviewin angiosperm species is less than that of animal mitochondrial DNA or plant nuclearDNA, which may limit the utility of cpDNA analysis in studies of angiosperm populationstructure (Palmer, 1987). However, conifer chloroplast genomes for which data areavailable differ from those of angiosperms; they lack a large inverted repeat region andprobably contain dispersed repeats (Lidholm et al., 1988; Strauss et al., 1988). Both ofthese characteristics have been associated with increased chloroplast genome variabilitydue to structural rearrangements (Palmer and Thompson, 1981). Studies at theintraspecific or population level using cpDNA are relatively few, and many have foundlittle cpDNA diversity. In Pinus torreyana, chloroplast genome variation was studied intwo separate populations, San Diego and island Santa Rosa, and no variation was detected(Waters and Schaal, 1991). However, chloroplast genes have been successfully employedfor phylogenetic analysis of many plants (Baum and Bailey, 1991; Terachi et al., 1990;McDonald and Mabry, 1992) including tree species (Strauss and Doerksen, 1990).Moreover, cpDNA can provide a powerful tool for examining hybridization andintrogression, since the molecule is not disrupted by sexual recombination andintrogression can be detected by presence of a variant chloroplast type from anotherspecies. Wagner et al., (1987) studied cpDNA variation in Pinus banksiana and P.contorta in a region of sympatry. The result showed that cpDNA variation was found athigh levels in these pines and was sufficient for studies at the population level.1.2.3.3 Mitochondrial DNA (mtDNA)Mitochondria are the organelles inside the cells of higher organisms which directenergy production through the electron transport chain. In return for supplying the entirecell with enzymes and structures necessary for energy metabolism, mitochondria receivea steady supply of nutrients for their own growth and reproduction (Cann, 1989).11Chapter 1 Introduction and Literature ReviewMitochondrial DNA of higher plants is much larger than mtDNA of humans or yeast, andcan exist as a heterogeneous population of different sized circular molecules within oneplant (Newton, 1988). Each mitochondrion may contain two to four copies of its owngenome, and a single cell can have many hundreds of mitochondria largely independentof events dictated by the cell nucleus. Their division is not tied to cell division, in eithersomatic or germ cells (Cann, 1989). Recent investigations suggest that plantmitochondrial genomes of many species contain cpDNA sequences (Stern and Palmer,1984). Moreover, mtDNA sequences as well as cpDNA have been found in nuclear DNA(Kemble et al., 1983).The mitochondrial DNA codes for RNA components of mitochondrial protein-synthesizing machinery and for a limited number of proteins, including factors associatedwith cytoplasmic male sterility (CMS) in maize, and Sorghum (Bailey-Serres et al.,1986). There is also evidence that mutations of different mitochondrial genes causeCMS, although in a few cases CMS may not be caused by mitochondrial gene mutations(reviewed in Fragoso et al., 1989). Moreover a number of other commercially interestingproperties, such as resistance to plant pathogens, are also located on mtDNA.In forest trees, the use of mtDNA markers in population genetics is still limited.Little is known of the amount and distribution of mtDNA diversity in natural populations.However, some reports are beginning to appear on population differentiation of mtDNA,such as in Pinus species (Hong et al., in press) and many more are expected to bereported in the near future.1.2.3.4 Mode of inheritance in organelle genomes.For beginning any study of DNA evolution, either nuclear DNA or organelleDNA, it is important to understand how these genomes are inherited (Table 1.2).12Chapter 1 Introduction and Literature ReviewTable 1.2 Inheritance of organelle DNA in conifers and some forest tree speciesSpecies^cpDNA^mtDNA^ReferencesCoast Redwood(Taxodicaceae)Douglas-fir(Pinaceae)Incense Cedar(Cupressaceae)Jack Pine(Pinaceae)Loblolly Pine(Pinaceae)Longleaf Pine(Pinaceae)Larch(European, Japanese)Poplarpaternal^paternalpaternal^maternalpaternal^paternalpaternalpaternal^maternalpaternalmaternalmaternal^maternalNeale et al., (1989)Marshall, and Neale, (1992)Neale et al., (1986)Neale et al., (1991)Wagner et al., (1989)Dong et al., (1992)Neale and Sederoff, (1989)Wagner et al., (1992)Dong et al., (1992)Wagner et al., (1992)Szmidt et al., (1987)Mejnartowicz, (1991)Rajora et al., (1992)Rajora and Dancik, (1992)Spruce^paternal^maternal^Stine et al., (1989)(Englemann, blue, white)^Stine and Keathley, (1990)Szmidt et al., (1988)Sutton et al., (1991)Many reports have shown that mitochondrial genomes appear to be alwaysmaternally inherited in angiosperms. In conifers, mtDNA has been studied in only a fewspecies, such as loblolly pine, coast redwood (Neale et al., 1988), and coastal Douglas-fir(Marshall and Neale, 1992). Loblolly pine and coastal Douglas-fir showed maternal13Chapter 1 Introduction and Literature Reviewinheritance of mtDNA, whereas redwood showed the unexpected result of paternalinheritance. In Douglas-fir, the inheritance of cpDNAs was studied by Neale et al.,(1986) and the results showed paternal inheritance This result is in contrast to thematernal or biparental transmission of cpDNA observed in angiosperms.1.2.4 Systematic and evolutionary studiesDNA sequence markers have several advantages over other characters, such asallozymes and terpenes (reviewed in Strauss et al., 1992). Most DNA sequence variationoccurs in non-coding regions of DNA (Olmstead et al., 1989). In non-coding regions, itmay be assumed that all sites evolve at the same rate (Li and Graur, 1991), and it allows avariety of evolutionary scales to be studied. In addition, a large number of independenthomologous characters can be identified.In phylogenetic studies, the evolutionary relationships among groups oforganisms are illustrated by means of a phylogenetic tree. Two methods for inferringphylogenetic relationships from molecular data are commonly used, including cladisticsand phenetics. Cladistics is the study of the pathways of evolution of organisms. Thebranching sequence of phylogenetic tree expresses such ancestor-descendant relationshipsamong groups of organisms and is called a cladogram. Cladatic methods attempt to inferphylogenies and build trees by analyzing gains or losses of individual characters. Atypical representative of the cladistic approach is the maximum parsimony method (Fitch,1977). On the other hand, phenetics, or distance method is the study of relationshipsamong a group of organisms on the basis of degree of molecular, phenotypic oranatomical similarities between them, and a phylogenetic tree expressing pheneticrelationships is called a phenogram (Li and Graur, 1991). The unweighted pair groupmethod with arithmetic mean (UPGMA) (Sokal, and Micherner, 1958; Sneath and Sokal,14Chapter 1 Introduction and Literature Review1973) is a typical phenetic method. Although, phenogram may serve as an indicator ofcladistic relationships, it is not necessarily identical to the cladogram. If there is a linearrelationship between the time of divergence and the degree of genetic or morphologicaldivergence, the two types of trees may become identical (Li and Graur, 1991).Of the three genomes, nuclear, chloroplast, and mitochondrial, the chloroplastgenome has proved to be the most useful for phylogenetic analyses due to its high copynumber, often 5,000 genome copies per cell (Olmstead et al., 1989). Polymorphisms incpDNA within species is often very low, but sometimes can be considerable (Soltis et al.,1989) and the polymorphisms may be high beyond species limits among very closelyrelated species (Doyle et al., 1989). In genetic diversity studies, patterns of cpDNA andmtDNA variation are not necessarily congruent with patterns of morphological orallozyme variation (Moore et al., 1991). The patterns of geographic variation in cpDNAcould differ from those of mtDNA. Mutation is the ultimate source of genetic variation(Hartl, 1980) and nucleotide substitution in mtDNA is less than one-third that in cpDNA.The slower rate in mtDNA evolution than in cpDNA is probably due to a lower mutationrate (Wolfe et al., 1987). Thus it is possible that geographic structure has evolveddifferently in relatively young populations with regard to mtDNA and cpDNA. Withhigh levels of nucleotide substitution, restriction site analysis is most appropriate forphylogenetically closely related organisms (Olmstead et al., 1989). However, the samespecies in different geographic regions may not be closely related based on RFLPanalysis. As reviewed in Sytsma (1990), several sections of Oncidium in Brazil, forexample, are phylogenetically more closely related, based on cpDNA, to other genera inthe same region than they are to other sections of Oncidium with similar floramorphologies occurring in other geographical areas. Also, the present geographicdistribution of individual species often shows little correlation with the degree of cpDNA15Chapter 1 Introduction and Literature Reviewdifferentiation among species (Szmidt et al., 1988). In conifers, phylogeneticrelationships have been studied using RFLP markers in several species, such as cpDNAin California closed-cone pines (Hong, 1991; Hong and Strauss, in press), cpDNA,mtDNA, and nuclear rDNA in Pseudotsuga spp.(Strauss et al., 1990), and in Pinus spp.(Strauss and Doerksen, 1990), and nuclear rDNA in Pinus spp. (Govindaraju et al.,1992).16Chapter 1 Introduction and Literature Review1.3 LITERATURE CITEDBailey-Serres, J., L.K. Dixon, A.D. Liddell, and C.J. Leaver. 1986.^Nuclear-Mitochondrial Interactions in Cytoplasmic Male-Sterile Sorghum. Theor. Appl. Genet. 73:252-260.Baum, B.R., and L.G. Bailey. 1991. Relationships among Native and Introduced NorthAmerican Species of Hordeum, Based on Chloroplast DNA Restriction-SiteVariation. Can. J. Bot. 9:2421-2426.Boffey, S.A. 1991. DNA. In Rogers, L.J. (Ed), Methods in Plant Biochemistry. AminoAcids Proteins and Nucleic Acids. Volume 5. Academic Press Limited. pp. 147-169.Bradshaw, A.D. 1965. Evolutionary Significance of Phenotypic Plasticity in Plants.Adv. Genet. 23:115-155.Caetano-Annoles, G., B. J. Bassam, and P.M. Gresshoff. 1991. DNA AmplificationFingerprinting: A Strategy for Genome Analysis. Plant Molecular Biolygy Reporter 9:294-307.Cann, R.L. 1889. Cytoplasmic Inheritance. In Yearbook of Science & Technology.McGraw-Hill. pp. 86-89.Chengde, C. 1981. A Brief Introduction to the Chinese Species of the GenusPseudotsuga. Davidsonia 12:15-17.Christiansen, H. 1963. On the Chromosomes of Pseudotsuga macrocarpa andPseudotsuga menziesii. Silvae Genet. 12:124-127.Colangeli, A.M., Y.A. El-Kassaby, and 0. Sziklai. 1982. Karyotye Analysis of theGenus Pseudotsuga. In Proceedings of the IUFRO Joint Meeting of WorkingParty on Genetics About Breeding Strategy Including Multiclonal Varieties.Federal Republic of Germany, September 6-10,1982. (Abstract).Conkle, M.T. 1992. Genetic Diversity—Seeing the Forest Through the Trees. NewForests 6:5-22.Cotterill, P.P., and C.A. Dean. 1988. Changes in the Genetic Control of Growth ofRadiata Pine to 16 Years and Efficiencies of Early Selection. Silvae Genet.37:138-146.Crow, J.F. 1976. Genetics Notes. Seventh Edition. Burgess Publishing Company.Minneapolis, Minnesota.17Chapter 1 Introduction and Literature ReviewDe-Vescovi, and O. Sziklai. 1975. Comparative Karyotype Analysis of Douglas-fir.Silvae Genet. 24:68-73.Doerksen, A.H., and K.K. Ching. 1972. Karyotypes in the Genus Pseudotsuga. For. Sci.18:66-69.Dong, J., D.B. Wagner, A.D. Yanchuk, M.R. Carlson, S. Magnussen, X.-R. Wang, andA.E. Szmidt. 1992. Paternal Chloroplast DNA Inheritance in Pinus contorta andPinus banksiana: Independence of Parental Species or Cross Direction. J. Hered.83:419-422.Doyle, J.J., J. Doyle, and A.H.D. Brown. 1989. The Limit of Chloroplast DNA inPhylogeny Reconstruction: Polymorphism and Phylogeny in the B Genome ofGlycine. Am. J. Bot. 76:239 (suppl.)El-Kassaby, Y.A., A.M. Colangeli, and O. Sziklai. 1983. A Numerical AnalysisofKryotypes in the Genus Pseudotsuga. Can. J. For. Res. 61:536-544.El-Lakany, M.H.,and O. Sziklai. 1971. Intraspecific Variation in Nuclear Chlaracteristicsof Douglas-Fir. Advancing Frontiers of Plant Sciences. 28:363-378.Farnum,P., R. Timmis, and J.L. Kulp. 1983. Biotechnology of Forest Yield. Science219:649-702.Fitch, W.M. 1977. On the Problem of Discovering the Most Parsimonious Tree. Am.Natur. 111:223-257.Fowells, H.A. 1965. Silvic of Forest Trees of the United States. USDA Handb 271.Fragoso, L.L, S.E. Nichols, and C.S. Levings III. 1989. Rearangements in MaizeMitochondrial Genes. Genome. 31:160-168.Frothingham, E.H. 1909. Douglas-fir: A Study of the Pacific Coast and RockeyMountain Forms. USDA. Forest Service-Circular 150.Govindaraju, D., P. Lewis, and C. Cullis. 1992. Phylogenetic Analysis of Pines UsingRestriction Fragment Length Polymorphisms.  Pl. Syst. Evol. 179:141-153.Grierson, D., and S.N. Covey. 1988. Plant Molecular Biology. Second Edition. Blackie,Glasgow.Hamrick, J.L., M.J.W. Godt, and S.L. Sherman-Broyles. 1992. Factors InfluencingLevels of Genetic Diversity in Woody Plant Species. New Forest 6:95-124.18Chapter 1 Introduction and Literature ReviewHartl, D.L. 1980. Principles of Population Genetics.^Sinauer Associates, Inc.Sunderland, Mass.Heaman, J.C. 1967. A Review of the Plus Tree Selection Programme for Douglas-fir inCoastal British Columbia. Research Note No. 44. B.C. Forest Service, Victoria,Canada.Hermann, R.K. 1982. The Genus  Pse^ Historical ^and No erm -1 lature.Forest Research Laboratory, Oregon State University, Corvallis.^SpecialPublication 2a. 29 p.Hermann, R.K. and D.P. Lavender. 1990. Pseudotsugar menziesii (Mirb.) Franco. InSilvic of North America. Volume 1. Conifers. USDA. Forest Service.Agriculture Handbook 654. pp. 527-540.Hong, Y.-P. 1991. Chloroplast DNA Variability and Phylogeny in the California ClosedCone Pines. Ph.D. Thesis. Oregon State University.Hong, Y.-P., and S.H. Strauss. Chloroplast DNA Analysis of Phylogenetic Relationshipsin the California Closed-Cone Pines (Pinus radiata, P. muricata, and P.attenuata). Molecular Phylogenetics & Evolution (in press)Hosie, R.C. 1969. Native Trees of Canada. Seventh Edition. Canadian Forestry Service.Department of Fisheries and Forestry.Jian, S.K. 1978. Inheritance of Phenotypic Plasticity in Soft Chess, Bromus mollis L.(Graminae) Experientia 34:835-836.Khan, M.A., and A.D. Bradshaw. 1976. Adaptation to Heterogeneous Environment. II.Phenotypic Plasticity in Response to Specing in Linum. Aust. J. Agric. Res.27:519-531.Kemble, R.J., R.J. Mans, S. Gabay-Laughnan, and J.R. Laughnan. 1983. SequencesHomologous to Episomal Mitochondrial DNAs in the Maize Nuclear Genome.Nature 304:744-747.Kuiper, P.J.C. 1990. Analysis of Phenotypic Responses of Plants to Change in theEnvironment in Term of Stress and Adaptation. ACTA. Bot. Neerl. 39:217-227.Lapitan, N.L.V. 1992. Organization and Evolution of Higher Plant Nuclear Genomes.Genome 35:171-181.Ledig, F.T., and M.T. Conkle. 1983. Gene Diversity and Genetic Structure in NarrowEndemic, Torrey Pine (Pinus torreyana Parry ex Carr). Evolution 37:79-85.19Chapter 1 Introduction and Literature ReviewLi, P., and W.T. Adams. 1989. Range-wide Patterns of Allozyme Variation in Douglas-fir (Pseudotsuga menziesii). Can. J. For. Res. 19:149-161.Li, W.-H., and D. Graur. 1991. Fundamentals of Molecular Evolution.  SinauerAssociates, Inc. Publishers. Sunderland, Massachusetts.Lidholm, J., A.E. Szmidt, J.-E. Hallgren, and P. Gustafsson. 1988. The ChloroplastGenomes of Conofers Lack One of the rRNA-Encoding Inverted Repeats. Mol.Gen. Genet. 212:6-10.Mantell, S.H., J.A. Mathew, and R.A. Mckee. 1985. Principles of Plant Biotechnology.Blackwell, Oxford.Marshall, K.A., and D.B. Neale. 1992. The Inheritance of Mitochondrial DNA inDouglas-fir (Pseudotsuga menziesii (Mirb.) Franco). Can  J. For. R. 22:73-75.McDonald, J.A., and T.J. Mabry. 1992. Phylogeny Systematics of New World Ipomoea(Convolvulaceae) Based on Chloroplast DNA Restriction Site Variation. Pl. Syst. Evol. 180:243:259.Mejnartowicz, L. 1991. Inheritance of Chloroplast DNA in Populus. Theor. Appl.Genet. 82:477-480.Miksche, J. 1971. Intraspecific Variation of DNA per Cell Between Picea sitchensisProvenances. Chromosoma (Berl) 32:343-352.Moore, W.S., J.H. Graham, and J.T. Price. 1991. Mitochondrial DNA Variation in theNorthern Flicker (Colaptes auratus, Ayes.). Mol. Biol. Evol. 8:327-334.National Council on Gene Resources. 1982. Douglas-fir Genetic Resources. AnAssessment  aiNCalin California. California Gene Resources Program,National Council on Gene Resources, Berkeley, California.Neale, D.B., and R.R. Sederoff. 1989. Paternal Inheritance of Chloroplast DNA andMaternal Inheritance of Mitochondrial DNA in Loblolly Pine. Theor. Appl. Genet. 77:212-216.Neale, D.B., K.A. Marshall, and R.R. Sederoff. 1988 Inheritance of Chloroplast andMitochondrial DNA in Conifers. In Jan-Eric Hallgren (Ed),  Molecular Geneticsof Forest Trees. Proceedings of the Frans Kempe Symposium in lima,  June 14-16, 1988. pp. 89-100.20Chapter 1 Introduction and Literature ReviewNeale, D.B., K.A. Marshall, and R.R. Sederoff. 1989. Chloroplast and MitochondrialDNA are Paternally Inherited in Sequoia sempervirens D. Don Endl. Proc. Natl. Acad. Sci. USA. 86:9347-9349.Neale, D.B., K.A. Mashall, and D.E. Harry. 1991. Inheritance of Chloroplast andMitochondrial DNA in Incense-Cedar (Calocedrus decurrensis). Can. J. For. Res.21:717-720.Neale, D.B., N.C. Wheeler, and R.W. Allard. 1986. Paternal Inheritance of ChloroplastDNA in Douglas-fir. Can. J. For. Res. 16:1152-1154.Neale, D.B., M.E. Devey, K.D. Jermstad, M.R. Ahuja, M.C. Alosi, and K.A. Marshall.1992. Use of DNA Markers in Forest Tree Improvement Research. New Forests6:391-407.Newton, K. 1988. Plant Mitochondrial Genomes: Organization, Expression andVariation. Arum. Rev. Plant Physiol. Plant Mol. Biol. 39:503-532.Olmstead, R.G., R.K. Jansen, H.J. Micheals, S.R. Downie, and J.D. Palmer. 1989.Chloroplast DNA Phylogenetic Studies in th Asteridae. In Kawano, S. (Ed),Biological Approaches and Evolutionary Trends in Plants Academic Press,London. pp. 119-134.Olsen, W.C. 1988. Molecular Biology in Forestry Research: A Review. In Valentine,F.A. (Ed.), Forest and Crop Biotechnology. Progress and Prospect. Springer-Verlag. pp. 315-334.Orr-Ewing, A.L. 1975. Breeding in Pseudotsuga in Coastal British Columbia. InProceedings of the Fifteenth Meetings of the Canadian Tree ImprovementAssociation: Part 1. Petawawa Forest Experiment Station, Chalk River. August18-22, 1975. pp. 181-182.Palmer, J.D. 1985. Evolution of Chloroplast and Mitochondrial DNA in Plants andAlgae. In Maclntyre, R.J. (Ed), Molecular Evolutionary Genetics. Plenum Press.New York. pp. 131-240.Palmer, J.D. 1987. Chloroplast DNA Evolutionand Biosystematic Used of ChloroplastDNA Variation. Am. Natur. 130, S6 - S29.Palmer, J.D. 1988. Isolation and Structural Analysis of Chloroplast DNA. In Weissbach,A. and H. Weissbach (Eds), Methods for Plant Molecular Biology. AcademicPress. pp. 105-124.21Chapter 1 Introduction and Literature ReviewPalmer, J.D., and W.F. Thompsom. 1981. Rearrangement in the Chloroplast Genomes ofMung Beanand Pea. Proc. Natl. Acad. Sci. USA.  78:5533-5537.Potter, R.H., and M.G.K. Jones. 1991. Molecular Analysis of Genetic Stability. InDodds, J.H. (Ed.), In Vitro Methods for Conservation of Plant Genetic Resources.London: Chapman and Hall. pp. 71-91.Rajora, O.P., and B.P. Dancik. 1992. Chloroplast DNA Inheritance in Populus. TheorAppl Genet. 84:280-285.Rajora, O.P., J.W. Barett, B.P. Dancik, and C. Strobeck. 1992. Maternal Transmission ofMitochondrial DNA in Interspecific Hybrids of Populus. Curr. Genet. 22:141-145.Roach, D.A., and R.D. Wulff. 1987. Maternal Effects in Plants. Ann. Rev. Ecol. Syst.18:209-235.Schaal, B.A., S.L. O'Kane, Jr., and S.H. Rogstad. 1991. DNA Variation in PlantPopulations. TREE. 6:329-333.Schmidt, R.L. 1960. Factor Controlling the Distribution of Douglas-fir in Coastal BrisishColumbia.uarOtery Journal of Forestry 54:156-160.Sederoff, R.R., and F.T. Ledig. 1985. Increasing Forest Productivity and Value ThroughBiotechnology. In Forest Potentials Productivity and Value.  Proceedings of aSymposium Held at Tacoma, Washington. Weyehaueser Science Symposium.Volume 4. pp. 253-276.Silen, R.R. 1978. Genetic of Douglas-fir.  USDA For. Serv. Res. Pap. Wo-35. 34 p.Sneath, P.H.A., and R.R. Sokal. 1973. Numerical Taxonomy. W.H. Freeman, SanFrancisco.Sokal, R.R., and C.D. Michener. 1958. A Statistical Method for Evaluating SystematicRelationships. Univ. Kansas Sci. Bull. 28:1409-1438.Soltis, D.E., P.S. Soltis, and B.D. Ness. 1989. Chloroplast DNA Variation and MultipleOrigins of Autopolyploidy in Heuchera micrantha (Saxifragaceae). Evolution43:650-656.Stern, D.B., and J.D. Palmer. 1984. Extensive and Widespread Homologies BetweenMitochondrial DNA and Chloroplast DNA in Plants. Proc. Natl. Acad.Sci. USA.81:1946-1950.22Chapter 1 Introduction and Literature ReviewStine, M., and D.E. Keathley. 1990. Paternal Inheritance of Plastids in EngelmannSpruce x Blue Spruce Hybrids.  J. Hered. 81:443-446.Stine, M., B.B. Sears, and D.E. Keathley. 1989 Inheritance of Plastids in InterspecificHybrids of Blue Spruce and White Spruce. Theor. Appl. Genet. 78:768-774.Strauss, S.H., and A.H. Doerksen. 1990. Restriction Fragment Analysis of PinePhylogeny. Evolution 44:1081-1096.Strauss, S.H., A.H. Doerksen, and J.R. Byrne. 1990. Evolutionary Relationships ofDouglas-fir and Its Relatives (Genus Pseudotsuga) from DNA RestrictionFragment Analysis. Can. J. Bot. 68:1502-1510.Strauss, S., D.B. Neale, D.B. Wagner. 1989. Genetics of the Chloroplast in ConifersBiotechnology Research Reveals Some Surprises. J. Forestry 87:11-17.Strauss, S.H., J. Bousquet, V.D. Hipkins, and Y.-P. Hong. 1992. Biochemical andMolecular Genetic Markers in Biosystematic Studies of Forest Trees. NewForests. 6:125-158.Strauss, S.H., J.D. Palmer, G.T. Howe, and A.H. Doerksen. 1988. Chloroplast Genomesof Two Conifers Lack a Large Inverted Repeat and Extensively Rearranged. Proc.Natl. Acad. Sci. USA. 85:3898-3902.Sutton, B.C.S., D.J. Flanagan, J.R. Gawley, C.H. Newton, D.T. Lester, and Y.A. El-Kassaby. 1991. Inheritance of Chloroplast and Mitochondrial DNA in Picea andComposition of Hybrids from Introgression Zones. Theor Appl. Genet. 82:242-248.Sytsma, K.J. 1990. DNA and Mophology: Inference of Plant Phylogeny. TREE. 5:104-110Sziklai, 0., Y.A. El-Kassaby, and R.K. Scagel. 1985. Relationship of Pseudotsugamenziesii with Other Pseudotsuga Species Inferred from KaryotypeReconstruction. In Ruetz, W., and J. Nather (Eds), Proceedings of the IUFROWorking Party on Breeding Strategy for Douglas-Fir as an Introduced Species.Vienna, Austria. June 1985. pp. 127-142.Szmidt, A.E., T. Alden, and J.-E.. Hallgren. 1987. Paternal Inheritance of ChloroplastDNA in Larix. Plant Mol. Biol. 9:59-64.Szmidt, A.E., A. Sigurgeirsson, X.-R. Wang, J.-E. Hallgren, and D. Lindgren. 1988.Genetic Relationships among Pinus species Based on Chloroplast DNA23Chapter 1 Introduction and Literature ReviewPolymorphism. In Hallgren, J.-E. (Ed.), Proc. Fra s Kem s e S M s lecularGenetic of Forest Trees. timed, Sweden, pp. 33-47.Terachi, T., Y. Ogihara, and K. Tsunewaki. 1990. The Molecular Basis of GeneticDiversity Among Cytoplasms of Triticum and Aegilops. 7. RestrictionEndonuclease Analysis of Mitochondrial DNAs from Polyploid Wheats and TheirAncestral Species. Theor. Appl. Genet. 80:366-373.Thomas, G., and K.K. Ching. 1968. A Comparative Karyotype Analysis of Pseudotsugamenziesii (Mirb.) Franco, and Pseudotsuga wilsoniana (Hayata). Silvae Genet.17:138-142.Tilney-Bassett, R.A. 1975. Genetics of Variegated Plants. In Birky, Jr, C.W., P.S.Pearlman, and T.J. Byers (Eds.), Genetics and Biogenesis of Mitochondria andChloroplasts. Columbus: Ohio State Univ. Press. pp. 268-308.von Rudloff, E. 1972. Chemosystematic Studies in the Genus Pseudotsuga. I. Leaf OilAnalysis of the Coastal and Rocky Mountain Varieties of the Douglas-fir. Can. J.Bot. 50:1025-1040.Wagner, D.B. 1992. Nuclear, Chloroplast, and Mitochondrial DNA Polymorphisms asBiochemical Markers in Population Genetic Analyses of Forest Trees. NewForests 6:373-390.Wagner, D.B., D.R. Govindaraju, C.W. Yeatman, and J.A. Pitel. 1989. PaternalChloroplast DNA Inheritance in a Diallel Cross of Jack Pine (Pinus banksianaLamb.). ..eol. 80:483-485.Wagner, D.B., G.R. Furnier, M.A. Saghai-Maroof, S.M. Williams, B.P. Dancik, and R.W.Allard. 1987. Chloroplast DNA Polymorphisms in Lodgepole Pine and JackPines and Their Hybrids. Proc. Natl. Acad. Sci USA. 84:2097-2100.Wagner, D.B., W.L. Nance, C.D. Nelson, T. Li, R.N. Patel, and D.R. Govindaraju. 1992.Taxonomic Patterns and Inheritance of Chloroplast DNA Variation in a Survey ofPinus echinata, Pinus elliottii, Pinus palustris and Pinus taeda. Can. J. For. Res.22:683-689.Waters, E.R., and B.A. Schall. 1991. No Variation is Detected in the ChloroplastGenome of Pinus torreyana. Can. J. For. Res. 21:1832-1835.Welsh, J., and M. McClelland. 1990. Fingerprinting Genomes Using PCR with ArbitraryPrimers. Nucleic Acids Res. 18:7213-7218.24Chapter 1 Introduction and Literature ReviewWilliams, C.G. and D.B. Neale. 1992. Conifer Wood Quality and Marker-AidedSelection: A Case Stusy. Can. J. For. Res. 22:1009-1017.Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski, and S.V. Tingey. 1990. DNAPolymorphisms Amplified by Arbitrary Primers are Useful as Genetic Markers.Nucleic Acids Res. 18:6531-6535.Wolfe, K.H., W.-H. Li, and P.M. Sharp. 1987. Rates of Nucleotide Substitution VaryGreatly among Plant Mitochondrial, Chloroplast, and Nuclear DNAs. Proc. Natl. Acad. Sci. USA. 84:9054-9058.Woodruff, D.S., and G.A.E. Gall. 1992. Genetic Conservation. Agriculture, Ecosystemsand Environment 42:53-73.Wulff, R.D., and F.A. Bazzaz. 1992. Effect of The Parental Nutrient Regime on Growthof the Progeny in Abutilon Theophrasti (Malvaceae). Am. J. Bot. 79(10):1102-1107).Zobel, B., and J. Talbert. 1984. Applied Forest Tree Improvement. John Wiley andSons. New York:Zobel, B.J., and J.P. van Buijtenen. 1989. Wood Variation; Its Causes and Control. Springer-Verlag. Berlin Heidelberg.25CHAPTER 2GENETIC DIVERSITY OF DOUGLAS-FIR IN B.C. BASED ONRANDOM AMPLIFIED DNA FINGERPRINTING2.1 INTRODUCTIONIn conifers, the long generation time of most species means that manymorphological descriptors can only be assessed at maturity and genetic studies at thepopulation level are usually based on quantitative traits. During the past two decadesallozymes have been developed and widely used as genetic markers in forest genetics andapplied breeding studies dealing with population genetic structure, mating systems, andphylogeny (Guries and Ledig, 1982; Milton, 1983; Wheeler et al., 1983; Neale andAdams, 1985; El-Kassaby et al., 1987; Li and Adams, 1989; Moran and Adams, 1989).Monoterpene levels in vegetative tissue have been used as biosystematic markers at bothinterspecific (Adams and Simmons, 1987) and intraspecific levels (Forde and Blight,1964). More recently, the application of molecular techniques, such as RestrictionFragment Length Polymorphisms (RFLPs) has proven to be a powerful tool in the studyof genetic variation. Because molecular markers are not subject to environmentalinfluences, they provide an opportunity to examine more precisely the genetic basis oftraits and the relationships between taxa (Bernatzky and Tanksley, 1989).Since the polymerase chain reaction (PCR) was introduced in 1985 (Mullis andFaloona, 1987), it has been applied and modified to suit many aspects of moleculargenetic studies. Two groups have reported that PCR using single primers is veryeffective in revealing genetic polymorphisms (Welsh and McClelland, 1990; Williams etal., 1990). One such technique is known as Random Amplified Polymorphic DNA(RAPD). The RAPD marker system varies from standard PCR in that polymorphisms aredetected using a single primer of 8 to 10 bases in length and of "arbitrary" sequence26Chapter 2 Random Amplified DNA Fingerprintingrather than allele specific primer pairs of 20 -30 bases, uses more temperature cycles, upto 45, and uses annealing at 35-37°C rather than 42 - 55°C. RAPDs have been shown tobe effective in identifying useful polymorphisms in both repetitive DNA and low-copyDNA sequences (Williams et al., 1990). Random amplified DNA fingerprinting (RADF)is a modification of the RAPD technique in which amplification products are separated bysize on a polyacrylamide gel and visualized by silver staining (Hong and Carlson,unpublished). RADF has several advantages over RFLP and RAPD, such as betterseparation of amplified products, higher sensitivity of DNA detection, and lesser expenseby preserving gels for permanent record instead of photographs. RADF is very similar tothe DNA Amplification Fingerprinting technique of Caetano-Annoles et al., (1991) withthe exception that RAPD amplification conditions are adhered to. RADF can produce alarge number of genetic markers, since an unlimited number of loci can be detected.Direct evaluation of genetic diversity is possible with RADF by analyzing haploidgenomes, such as megagametophyte DNA of conifer seeds.The RAPD (or RADF) technique for amplifying genomic DNA from haploidtissue in conifers seems ideal for the study of genetic structure at the molecular level.The segregation of RAPD markers in conifers has been demonstrated in Douglas-fir andwhite spruce (Carlson et al., 1991). Tree-specific genetic linkage maps have also beenconstructed in white spruce using megagametophyte DNA (Tulsieram et al., 1992). Inagricultural plants, RAPD markers have been applied to cocoa improvement programmesfor characterization and fingerprinting of cocoa genotypes (Wilde et al., 1992).To date both RFLP fingerprinting (Gilbert et al., 1991; Milgroom et al., 1992;Triggs et al., 1992) and RAPD marker (Chalmers et al., 1992) studies have revealedgenetic variation at the population level. Mendelian inheritance (i.e. 1:1 segregation ofcorresponding markers) of DNA fingerprinting markers for an individual indicates that on27Chapter 2 Random Amplified DNA Fingerprintingaverage an individual received half of its segregating DNA fragments from each parent.The probability that two unrelated individuals will present the same multilocus pattern isgenerally many orders of magnitude smaller than the reciprocal of the population size(Jeffreys et al., 1985). Fingerprints generated by the RADF technique usingmegagametophyte DNA, which is strictly maternal, can be utilized to assess geneticdiversity in conifers. PCR-based methods, such as RAPDs and RADF, are an alternativeto RFLPs which may reduce costs and allow molecular genetic markers to be morewidely adopted (Strauss et al., 1992).In this study, genetic variation in coastal and interior regions was assessed usingRADF marker analysis of Douglas-fir from two geographic regions in British Columbia,coastal, and interior. Haploid genomic DNAs from megagametophyte tissues of Douglas-fir seeds were used as genetic material to determine genetic variation of the species and todetermine phylogenetic relationships among populations in different geographic regions.Effect of altitude of origin on genetic variation of coastal Douglas-fir populations wasalso assessed because there have been several studies which show genetic variationamong altitudes. Finally, effect of sample sizes on genetic differentiation estimates wasinterpreted using computer simulation. Optimum sample size for evaluating geneticdifferentiation of populations was evaluated based on the assumption of constantobserved allelic frequencies from different sample sizes.2.2 MATERIALS AND METHODS2.2.1 Plant materialSeed samples of Douglas-fir were selected from the same individuals used inorganellar RFLP analysis (Chapters 3 and 4), representing two different geographic28Chapter 2 Random Amplified DNA Fingerprintingregions in B.C., coastal vs. interior. Six populations with four individuals each wereselected in each geographic region (Figure 2.1 and Table 2.1). For interior Douglas-firsamples, seeds from some individuals used in the organellar RFLP analysis were notavailable; therefore, other seed lots from the same populations were sampled. In addition,eleven individuals of coastal Douglas-fir from Washington were sampled to be used as anoutgroup for phylogenetic analysis.2.2.2 Isolation of DNA from megagametophyte.The method used for isolation of haploid DNA from the megagametophyte ofDouglas-fir seed was a modification of the CTAB method (DeVerno et al., 1989; Wagneret al., 1987). Seeds from each individual were imbibed in distilled water for at least 4hours to overnight prior to DNA extraction. Under a dissection-microscope at lowmagnification, seed coat and underlying layer were removed and the embryo wasdiscarded, using a dissecting forceps and disposable scalpel blade. Themegagametophyte was retained and transferred to a 1.5 ml microfuge tube containing 30111 CTAB wash buffer (50 mM Tris HC1 pH 8.0, 25 mM EDTA, 0.35 M sorbitol, 0.1% 13-mercaptoethanol). The megagametophyte was homogenized using a motorized pelletpestle. Eight volumes of CTAB wash buffer was added to the tube and mixed and 1/5volume of 5% sarkosyl was also added and then mixed using a vortex mixer. Themixture was incubated for 3-5 minutes at room temperature and then 1/7 volume of 5 M29Chapter 2 Random Amplified DNA FingerprintingFigure 2.1 Approximate locations of coastal and interior Douglas-fir populationsused in evaluation of genetic diversity (Al, Bl, C1 - coastal lowerelevation; A2, B2, C2 - coastal higher elevation; Interior -- CP - CentralPlateau, CT - Cariboo Transition, MR - Mt Robson, MI - Mica, SA -Shuswaps Adams, WK - West Kootenay).30Chapter 2 Random Amplified DNA FingerprintingTable 2.1 Description of population location and sources of seed samples used forstudy of genetic variation at molecular level in Douglas-firPopulations Locations Sources of materialsLatitude(°N)Longitude(°W)Elevation(m)Coastal Cowichan Lake ResearchAl 48°-49° 123°-125° 450-600 Station, Mesachie Lake, B.C.A2 48°-49° 123°-125° > 600 Forest Service ResearchB1 49°-50° 125° 0-150 Branch, Victoria, B.C.B2 48°-49° 123°-125° 150-300Cl 490-500 125°-127° 0-150C2 49°-50° 125°-127° 150-300Interior Kalamalka Research Station &CP 53°-54° 122°-124° 610-900 Seed Orchard, Vernon, B.C.CT 51°-52° 121°-122° 760-1000SA 50° 118°-119° 650-1250WK 49° 116°-117° 659-900MR 52°-53° 119°-120° 732-1140MI 51° 188° 585-68531Chapter 2 Random Amplified DNA FingerprintingNaC1 was added and mixed gently. 1/10 volume of 8.6% CTAB in 0.7 M NaCl wasadded and mixed gently and incubated for 15 minutes in a 65°C water bath with periodicmixing . RNA was eliminated by digesting with 2 ul of RNase free DNase (1 mg/u1) for15 minutes in a 37°C water bath. The tube was removed from the water bath and left tocool at room temperature for 5 - 10 minutes before extraction of protein with equalvolume of phenol: chloroform: isoamyl alcohol (25: 24: 1) by inverting the tube gentlyand spinning in a microcentrifuge at maximum speed (1,300 rpm or 14,926 g) for 10minutes. The aqueous phase containing DNA (upper part) was transferred to a new 1.5ml microfuge tube, then excess phenol was removed by adding 1 ml of diethyl ether tothe tube followed by mixing gently for 10 seconds and spinning in the microcentrifugefor 1 minute. The upper layer (ether) was removed by aspiration and DNA precipitatedby addition of 1 ml ice cold absolute ethanol and incubation at -20°C for 1 hour toovernight or 30 minutes at -70°C. The precipitated DNA was pelleted at maximum speedin a microcentrifuge at 4°C for 30 minutes and the pellet was washed in cold 70% ethanoland dried in the Speedvac for 5 - 10 minutes. The DNA pellet was redissolved in 160 piof T/E (10 mM Tris/1 mM EDTA pH 8.0). The concentration of DNA was determinedby gel electrophoresis against DNA standards. Finally, the DNA was prepared at thedesired concentration for use in PCR amplification by dilution with T/E.2.2.3 DNA Amplification.The method used for amplification of DNA was a modification of the RAPDprocedure reported by Williams et al., (1990) in which a single arbitrary primer was usedto yield amplification fragment polymorphisms (AFPs) of genomic DNA. Reactionmixtures for PCR (25 vil) contained 2 ul of dNTPs (A,C,G,T), 2.3 mM MgC1 2 , 0.3 uMprimer, 2 ng genomic DNA, 2.5 pi of 10X PCR buffer (supply with enzyme), and 2.532Chapter 2 Random Amplified DNA Fingerprintingunits of Amplitaq DNA Polymerase Stoffel fragment and overlay with 50 mineral oil.The reaction mixtures were usually prepared without template DNA prior to use. Thetemplate DNAs were denatured at 94°C for 7 minutes under mineral oil overlay to ensurethe DNAs were denatured and to avoid over exposing enzyme to heat before starting thereaction. Then tubes were taken to ice immediately and twenty microliters of reactionmixtures was added to the denatured Douglas-fir template DNA and mixed gently with amicropipetter. The amplification reactions were carried out in a DNA thermal cycler(N801-0150, Perkin Elmer/Cetus) using the following conditions: 94°C for 2 minutesbefore starting the cycle, 94°C for 1 minute (denaturing), 36°C for 1 minute (annealing),72°C for 2 minutes (extension), for 45 cycles. After the last cycle, another 10 minutes at72°C for extension was performed and then the reactions were soaked at 4°C until takenout of the cycler. The amplification products were kept in a freezer (-20°C) until theywere analyzed. In this study, twenty-five primers of 10 nucleotides in length were testedin amplification of megagametophyte DNAs of Douglas-fir. The amplification productsfrom these primer tests were separated by electrophoresis in 1.5% agarose gels, followedby staining with ethidium bromide (1µg/ml) and visualizing and photographing under UVradiation. The five primers (Table 2.2) which showed the maximum number ofinformative amplification products were chosen for the study on genetic diversity.Table 2.2 List of arbitrary primers showing sequence and G+C contentand number of segregating loci used in the studyPrimer Sequences # G+C # of loci203 CACGGCGAGT 7 22266 CCACTCACCG 7 20278 GGTTCCAGCT 6 15295 CGCGTTCCTG 7 25297 GCGCATTAGA 5 1533Chapter 2 Random Amplified DNA Fingerprinting2.2.4 Polyacrylamide gel electrophoresis (PAGE).Polyacrylamide gel electrophoresis was conducted in 1.5 nun thick slab gels of4% polyacrylamide and 7 M urea. The ratio of acrylamide to cross-linker N,NMethylenebisacrylamide was 19:1. Gels were cast onto the glass sandwich plates and runon a vertical gel electrophoresis system (BRL model V 16-2). The gels and theelectrophoresis running buffer (TBE) contained 0.83 M boric acid, 0.1 M Tris-base and0.1 M EDTA pH 8.3. The gel was prerun for 1 hour before loading samples. Each tubeof amplification products was mixed with 10 ill loading buffer (50% sucrose, 0.6%Orange G, 0.1% Xylene Cyanol F.F.) before loading onto the gels. The samples wereloaded into each well of the gel, using a long narrow bore pipet tip. A 100 by ladderDNA standards (BRL) were also loaded on the outside lanes for each gel. Controlamplification reactions without DNA templates were also conducted and run on each gelto test for DNA contamination. Gels were run at 60 volts overnight or at 200 volts for 3hours.2.2.5 Silver staining of amplified DNA in PAGE.After electrophoresis, the separated PCR amplification products were visualizedby silver staining Silver staining was performed by a modification of Bassam et al.,(1991). The procedure was performed in the following sequence: the gel was (1)immersed in 10% glacial acetic acid for 10 minutes with slow shaking, (2) rinsed indistilled water for 2 minutes (optionally, repeated rinsing two more times), (3) silverstained in 200 ml of silver nitrate solution (0.1% AgNO 3 , 0.018% HCOH) for 30 minuteswith slow shaking, (4) rinsed briefly in distilled water, (5) reduced in 200 ml solutioncontaining 3% anhydrous sodium carbonate, 0.018% HCOH, and 0.0002% Na 2CO3 5H2O, with slow shaking and then when the desired image intensity had developed,34Chapter 2 Random Amplified DNA Fingerprintingreduction was stopped with 10% glacial acetic acid. All of the processes were performedat room temperature. To preserve the gel for permanent record, the gel was immersed in50% ethanol for 5 - 10 minutes, then placed between 2 layers of cellophane membraneson a glass or plastic plate with all four margins firmly sealed with cellophane membranesand with binder clips and air-dried at room temperature. The dried gels could be keptindefinitely in a clear sealed plastic bag for later analysis (either manually or by imageanalysis).2.2.6 Data analysis.Mendelian inheritance (1:1 segregation of polymorphic markers; absence andpresence of amplification products) in progeny of RADF polymorphic markers used inthis study was assumed based on the results of segregation analyses in the spruce geneticlinkage mapping study (Tulsieram et al., 1992, Hong and Carlson in preparation).Genetic loci assignment of amplification fragment polymorphisms was done based on thefragment size of the observed bands on the gel, relative to the DNA size markers (100 byladder). In this way, amplification products for individuals run on different gels could becompared. Presence of the amplification product was designated as allele A and absenceas allele B for each segregating loci. Five megagametophyte DNAs were analyzed pereach individual tree in each population. This results in a 6.3% probability ofmisidentifying a heterozygote as a homozygote, based on the assumption of Mendelianinheritance. Genotypes at all segregating loci were recorded according to each individualand used for calculating population genetic parameters--allelic frequencies, mean numberof alleles per locus, percentage of polymorphic loci, and expected heterozygosity.Observed mean heterozygosity was obtained by counting the number of heterozygousindividuals in each population. All analyses were conducted with the BIOSYS-1 release35Chapter 2 Random Amplified DNA Fingerprinting1.7 software (Swofford and Selander, 1989). The gene diversity statistic (Nei andChesser, 1983) was estimated using the computer package GENESTAT-PC version 2.1(Whitkus, 1988).2.3 RESULTS AND DISCUSSION2.3.1 Variation and genetic diversity in Douglas-fir populations using RADFAllelic polymorphism was detected in RADF markers upon polyacrylamide gelelectrophoresis and silver staining of DNA. Based on the observation of 1:1 segregationratios of RAPD markers in a spruce linkage mapping study using megagametophyte DNA(Tulsieram et al., 1992), Mendelian inheritance of polymorphic RADF markers wasassumed. Using RADF markers based on 5 primers, 97 loci were observed and geneticvariability at 97 loci for all populations were calculated (Table 2.3). Intrapopulation chi-square analysis to test for departure from Hardy-Weinburg equilibrium for each locus(data not shown) indicated that on average allelic distributions did not deviate fromequilibrium expectations (a=0.05).Percent polymorphic loci within populations ranged from 78.4% (B1 and C2) to88.7% (WK) (Tables 2.3). In general, high levels of heterozygosity have been frequentlyobserved in conifers (Hamrick et al, 1981). In this study, the observed heterozygositiesfor all populations ranged from 0.459 (Al) to 0.665 (WK) (average 0.523), whereasexpected heterozygosities were between 0.305 and 0.440 (average 0.380). Thesefrequencies were not significantly different between coastal and interior Douglas-firpopulations. In Douglas-fir mean heterozygosities determined with allozymes werepreviously reported as 0.15 for coastal and 0.18 for interior populations in B.C., whereasthe populations in California and Eastern Colorado exhibited almost twice the36Chapter 2 Random Amplified DNA Fingerprintingheterozygosity estimated for the populations in B.C. (Table 2.4). Li and Adams (1989)estimated the mean expected heterozygosity of Douglas-fir from throughout the range ofthe species to be 0.137. In this study, the average heterozygosity estimates using RADFmarkers are higher than those based on allozyme markers perhaps because RADFmarkers represent a broader and more direct coverage of the genome. However,allozymes are gene products generated from only a small portion of the genome, i.e.protein coding sequences. Furthermore, Li et al., (1985) found that the nucleotidesubstitutions per silent site occur five times more frequently than amino acid replacementsubstitution per nonsilent site (in isozymes). Proteins are tolerant of some amino acidsubstitutions and those proteins can keep essentially the same electric net charge,structure and function against changes in the amino acid sequence, which results inundetectable changes by electrophoresis (Bowie et al., 1990; Hartl, 1980). Based onthese observations Nei (1986) estimated that about 33.3% of nucleotide substitutionsseem to result in amino acid changes and about 30% of those are detectable byelectrophoresis. This estimate predicts that the resolution of electrophoresis is capable ofdetecting only about 10% of nucleotide substitutions in allozyme coding sequences.Therefore, RADF markers which cover more of the genome and are more sensitive tonucleotide substitutions are better for studying genetic diversity at the population levelthan are allozymes.37Chapter 2 Random Amplified DNA FingerprintingTable 2.3 Genetic variability at 97 loci in all populations of Douglas-fir in B.C.PopulationMeansamplesize perlocusMean no.ofalleles perlocusPercentageof Mean heterozygosityPolymorphiclociObserve&(±SE)Expected2(±SE)Coastal, Al 4.0 1.8 79.4 0.459 0.305(0.036) (0.022)Coastal, A2 4.0 1.8 84.5 0.575 0.414(0.034) (0.020)Coastal, B1 4.0 1.8 78.4 0.461 0.365(0.034) (0.022)Coastal, B2 4.0 1.8 77.3 0.485 0.350(0.036) (0.022)Coastal, Cl 4.0 1.8 80.4 0.539 0.386(0.036) (0.022)Coastal, C2 4.0 1.8 78.4 0.523 0.374(0.037) (0.022)Interior, CP 4.0 1.8 79.4 0.454 0.355(0.033) (0.021)Interior, CT 4.0 1.8 82.5 0.482 0.370(0.034) (0.021)Interior, SA 4.0 1.8 81.4 0.523 0.386(0.034) (0.021)Interior, WI( 4.0 1.8 88.7 0.665 0.440(0.035) (0.019)Interior, MR 4.0 1.8 82.5 0.575 0.403(0.036) (0.021)Interior, MI 4.0 1.8 83.5 0.536 0.409(0.034) (0.021)Mean 81.4 0.523 0.3801 Based on a direct count of heterozygotes2 An unbiased estimate based on Hardy-Weinberg expectation (Nei, 1978).38Chapter 2 Random Amplified DNA FingerprintingTable 2.4 Genetic variability in Douglas-fir based on RADF and isozyme f markersLocationNo.ofpop.No. oflociMean no.of allelesper locusPercentageofPolymorphiclociMeanheterozygosity97Coastal, B.C. 1 6 (RADF) 2.00 96.9 0.43297Interior, B.C. 1 6 (RADF) 2.00 94.8 0.44097Coastal, Washington' (RADF) 1.90 83.5 0.375Coastal, B.C.2 11 21 2.23 0.150Interior, B.C.2 11 21 2.23 - 0.180Coastal, California2 9 11 3.17 74.2 0.332Interior, California2 1 17 1.78 100.0 0.330Eastern Colorado2 5 22 1.86 64.0 0.246f Source: National Council on Gene Resources (1982).1 This study, using RADF markers.2 Previous study, using allozyme markers.2.3.2 Partitioning of genetic variability within and between Douglas-fir populationsGenetic diversity statistics were calculated to determine the partitioning andnature of variation for total population (H T), within populations (Hs), among populations(DsT) and degree of population subdivision (GsT). The means across all 97 loci for allcomparisons are presented in Tables 2.5. With RADF markers, estimates of geneticdiversity within populations (H s) in each geographic region (coastal and interior) werealmost the same (H s 0.373 and 0.402). Low levels of genetic diversities amongpopulations (D sT) for both coastal (0.058) and interior (0.038) populations were observed.These values for among population genetic diversity were lower than for withinpopulations. The estimate of population subdivision (G sT), and degree of differentiation39Chapter 2 Random Amplified DNA Fingerprintingwas higher, but not significantly different, for coastal populations (0.134) than for interiorpopulations (0.086). However, genetic diversity of Douglas-fir in these two regionsshowed very similar patterns both within and among populations.Table 2.5 Hierarchical partitioning of genetic diversity (Nei and Chesser,1983) estimated from 97 RADF lociLevel of Analysis Hs(±SE)HT(±SE)DST(±SE)GST(±SE)Total population 0.384 0.443 0.060 0.135(0.277) (0.317) (0.038) (0.080)Coastal 0.373 0.431 0.058 0.134(0.252) (0.291) (0.054) (0.116)Interior 0.402 0.440 0.038 0.086(0.269) (0.298) (0.040) (0.086)Hs = gene diversity within populations; HT= total gene diversity; DsT= gene diversity amongpopulations (D sT = HT - Hs), GsT= proportion of total gene diversity resulting from geneticdifferentiation among populations (G sT = DsT/HT), SE estimated using Jackknifmg overpopulations (Weir, 1990).The distribution of genetic diversity in Douglas-fir observed in this study can beclassified as a typical pattern for long lived woody species, showing higher geneticdiversity within populations than among populations. The pattern suggests thatwidespread conifers with relative degree of habitat heterogeneity over the entire range orin certian portions of the species (Millar and Libby, 1991). This type of variation hasbeen found in several conifers, such as Incense-cedar (Libocedrus decurrens) (Harry,1987), and western white pine (Pinus monticola) (Steinhoff et al., 1983). In addition,outcrossing species tend to show most genetic variation within population with littledifferentiation among populations (Woodruff and Gall, 1992). Yeh and O'Malley (1980)used enzyme electrophoresis to detect variations in natural populations of Douglas-fir and40Chapter 2 Random Amplified DNA Fingerprintingfound moderately high diversity within populations (H s 0.155) with almost no differencesamong populations (GsT was only 0.026). This same pattern was observed by El-Kassabyand Sziklai (1982), and Moran and Adams (1989), and is similar to nuclear DNAdiversity observed in this study. In Sitka spruce (Yeh and El-Kassaby, 1980), Montereypine (Plessas and Strauss, 1986), and cedar elm (Sherman-Broyles et al., 1992), lowdegrees of population subdivision were also observed. Furthermore, Hamrick and Godt(1990) studied genetic diversities in plants based on life history and concluded that longlived woody perennials (GST 0.076) and outcrossing wind-pollinated plants (G sT 0.099)had considerably lower degree of population subdivision than wind dispersed seed plants(GST 0.143).In woody plants, genetic diversities based on allozyme markers vary widely.Within species genetic diversities in woody plants range from 0.0 to 0.40. (Hamrick etal., 1992) and within populations range from 0.0 to 0.35. Among gymnosperms, genusPseudotsuga showed higher genetic diversity than other pine species both within speciesand within populations (Hamrick et al., 1992). In addition, conifers have been shown toharbour large amounts of genetic diversity, but little geographic differentiation has beenfound in allelic frequencies at allozyme loci (Mouna and Szmidt, 1984) due to highdiversity within populations (Hs) and low level of genetic diversity among populations(DST). This is in agreement with the results observed here of high levels of geneticdiversity in Douglas-fir within populations and low levels of genetic diversity amongpopulations revealed by RADF. The organization of genetic diversity of Douglas-firappears to reflect the relatively high degree of micro-habitat heterogeneity of populationsin both coastal and interior regions. However, genetic diversity estimates based onRADF markers in Douglas-fir populations were higher than those based on allozymemarkers (Table 2.6). RADF markers can detect variation in both silent and non silent41Chapter 2 Random Amplified DNA Fingerprintingamino acid sites and also in both coding and non coding regions over the whole genome.This may explain a higher degree of heterozygosity based on RADF markers thanallozyme markers being detected in this study (in 2.3.1). Because of the wide distributionof the species and gene flow via both pollen and seed dispersion, high levels of geneticdiversity in nuclear DNA in Douglas-fir in both regions can be maintained. However, thepopulations within each region were probably subdivided by random genetic drift(founder effect) when the populations expanded.4241,Table 2.6 Genetic diversity statistics (Nei, 1973) of Douglas-fir based on RADF and isozyme markersSpeciesNo. ofpopulationsNo.oflociNo. ofindividuals/popHT Hs GST Remark ReferenceCoastal Douglas-fir6 97 4 0.439 0.368 0.160^RADF This study6 2 bulk 0.003 Mejnartowitcz (1976) 19 4 10-bulk 0.026 Yang et al. (1977)11 21 bulk 0.159 0.155 0.026 Yeh and O'Malley (1980)27 10 0.190 0.177 0.068 elevation El-Kassaby and Sziklai (1982)22 27 31-72 0.178 0.177 0.001^breedingzonesMerkle and Adams (1987)43 20 15-bulk 0.177 0.164 0.071 Li and Adams (1989)12 stands 18 bulk 0.168 0.165 0.012^2 areas Moran and Adams (1989)Interior Douglas-fir6 97 4 0.445 0.389 0.124^RADF This study11 21 bulk 0.043 Yeh (1981)36 20 0.157 0.151 0.043 Northern Li and Adams (1989)24 20 0.087 0.077 0.122 Southern Li and Adams (1989)(all of the above studies were based on isozyme markers, with the exception of this study).Chapter 2 Random Amplified DNA Fingerprinting2.3.3 Effect of altitudinal position on genetic variationGenetic diversity of Douglas-fir populations in B.C. as estimated based on RADFgenotypic data at 97 loci shows that all of the populations studied contain a considerableamount of genetic variation. The estimates of genetic diversities are generally higherthan those observed in Douglas-fir using allozyme markers (Table 2.6.).The effect of elevation on the level of genetic variation was analyzed only in thecoastal Douglas-fir populations. Coastal Douglas-fir populations were selected based onseed stratification zones (A, B, and C) and the populations in each zone were subdividedby elevations which ranged from sea level to higher than 600 meters (Table 2.7). Manybiologists assume that there would be little differentiation within continuous populations(Mayr, 1963). However, Jain and Bradshaw (1966) mentioned that genetic differentiationdoes occur over short distances in many plants and differentiation can occur withincontinuous stands of long-lived woody perennials (Endler, 1977). As studied in Norwayspruce, Bergmann and Gregorius (1979) reported that genetic variation decreased withincreased altitudinal position. This trend was also found in ponderosa pine (Mitton et al.,1980). Nevertheless, in this study no general trend toward decreased genetic variationwith increased altitudinal positions in coastal Douglas-fir populations was observed(Figure 2.5). This result with coastal Douglas-fir populations shows the same pattern asin Scots pine populations studied by Mejnartowitcz and Bergmann (1984).In the Douglas-fir populations assayed here with RADF markers, there is nosignificant difference in variation among populations and the correlation of meanheterozygosity and altitude positions is very low (R2 = 0.021). In the study of geneticdifferentiation in ponderosa pine along a steep elevation, Mitton et al., (1980) found thatthe populations from the highest and the lowest elevations showed a significant excess ofheterozygotes whereas the middle population was in Hardy-Weinberg expectation. This44Chapter 2 Random Amplified DNA Fingerprintingmay be attributed to an effective gene flow via long-distance pollen dispersal and a longreceptive period that permits trees to receive pollen from other trees over a considerableelevation range in mountainous topography (Silen, 1978). It might also point to arelatively low response of variation at these loci to selection caused by micro-climatevariation.Table 2.7 Effect of elevation on genetic diversity of coastal Douglas-firpopulationsPopulationElevation(m)Percentage of Mean heterozygosityPolymorphiclociObserved'(±SE)Expected2(±SE)Coastal, Al 450-600 79.4 0.459 0.305(0.036) (0.022)Coastal, A2 > 600 84.5 0.575 0.414(0.034) (0.020)Coastal, B1 0-150 78.4 0.461 0.365(0.034) (0.022)Coastal, B2 150-300 77.3 0.485 0.350(0.036) (0.022)Coastal, Cl 0-150 80.4 0.539 0.386(0.036) (0.022)Coastal, C2 150-300 78.4 0.523 0.374(0.037) (0.022)I Based on a direct count of heterozygotes2 An unbiased estimate based on Hardy-Weinberg expectation (Nei, 1978).4528032364044Y = 35.845 + 0.0018 X- R2= 0.021A2 M• C2B1■ B2AlI^ I^I200 400 600 800 1,000Chapter 2 Random Amplified DNA FingerprintingMean heterozygosity (%)Altitude (m)Figure 2.3 Effect of altitude on genetic diversity of coastal Douglas-fir.46Chapter 2 Random Amplified DNA Fingerprinting2.4 CONCLUSIONSUsing RADF markers with five primers, 97 loci were detected amongamplification products separated in high resolution polyacrylamide gels stained withsilver. This DNA staining procedure is highly sensitive which makes it easy to detectsegregating loci. A practically unlimited number of loci can be assessed and the raw dataeasily preserved. Therefore, the number of potential genetic markers is greatly increasedrelative to isozymes and to other molecular markers and is thus very useful for evaluatinggenetic diversity.Genetic variation of Douglas-fir in two geographic regions, coastal and interior, inB.C. were estimated using RADF markers. Very high levels of genetic variation wereobserved from the RADF data. Among geographic regions, the same level of meanexpected heterozygosity was observed in both interior (0.440) and coastal region (0.432),the mean expected heterozygosity for all populations (both regions) was 0.380. Withinpopulations genetic diversity and degree of population subdivision estimates were foundto be similar for both regions. The pattern of genetic diversity of Douglas-fir in bothregions (higher diversity within populations than among populations) was the same asother conifers, suggesting widespread and heterogeneity habitat over the entire range orsome particular portions of natural range of the species. Random genetic drift is probablyan important evolutionary force in nuclear genome of Douglas-fir in B.C.Effect of altitude position of Douglas-fir populations was studied in coastalpopulations located on Vancouver island. The elevation ranged from sea level to over600 m m maximum), as classified from stratification zones. There was no generaltrend observed to indicate that genetic variation of Douglas-fir populations varyaccording to altitude positions (R2 = 0.021 ). This observation shows the same result asobserved in Scots pine populations and some other species. This may be the result of47Chapter 2 Random Amplified DNA Fingerprintinglong distance pollen dispersal and long receptive periods providing an opportunity fortrees to receive pollen from other trees over considerable elevational range in mountaintopography.48Chapter 2 Random Amplified DNA Fingerprinting2.5 LITERATURE CITEDAdams, R., and D. Simmons. 1987. A Chemosystematic Study of Callitris(Cupressaceae) in South-eastern Australia Using Volatile Oils. Aust. For. Res.17:113-125.Bassam, B.J., G. Caetano-Anolles, and P.M. Gresshoff. 1991. Fast and Sensitive SilverStaining of DNA in Polyacrylamide Gel. Anal. Biochem. 196:80-83.Bergmann, F., and H.-R. Gregorius. 1979. Comparison of the Genetic Diversities ofVarious Populations of Norway Spruce (Picea abies). In Rudin, D. (Ed.), Proc. Conf. Biochem. Genet Forest Trees.  Ulna Sweden, pp. 99-107.Berntzky, R., and S.D. Tanksley. 1989. Restriction Fragments as Molecular Markers forGermplasm evaluation and Utilization. In Brown, A.H.D., D.R. Marshall, O.H.Frankel, and J.T. Williams. (Eds), The Use of Plant Genetic Resources.Cambridge University Press, Cambridge, pp. 353-362.Bowie, J.U., J.F. Reidhaar-Olson, W.A. Lim, and R.T. Sauer. 1990. Deciphering theMessage in Protein Sequences: Tolerance to Amino Acid Substitutions. Science247:1306-1310.Caetano-Aimoles, G., B. J. Bassam, and P.M. Gresshoff. 1991. DNA AmplificationFingerprinting: A Strategy for Genome Analysis. Plant Mol. Biol. Rep. 9(4):294-307.Carlson, J.E., L.K. Tulsieram, J.C. Glaubitz, V.W.K. Luk, C. Kauffeldt, and R. Rutledge.1991. Segregation of Random Amplified DNA Markers in F 1 Progeny ofConifers. Theor. Appl. Genet. 83:194-200.Chalmers, K.J., R. Waugh, J.I. Sprent, A.J. Simons and W. Powell. 1992. Detection ofVariation between and within Populations of Gliricidia sepium and G. maculataUsing RAPD Markers. Heredity. 69:465-472.DeVerno, L.L., J.R. Byrne, J.A. Pitel, and W.M. Cheliak. 1989. Constructing ConiferGenomic Libraries: A Basic Guide.  Information Report PI-X-88. PetawawaNational Forestry Institute, Forestry Canada.El-Kassaby, Y.A., and 0. Sziklai. 1982. Genetic Variation of Allozyme and QuantitativeTraits in a Selected Douglas-fir [Pseudotsuga menziesii var. menzisii (Mirb)Franco] Population. Forest Ecol. Manage. 4:115-126.El-Kassaby, Y.A., M.D. Meagher, J. Parkinson, and F.T. Portlock. 1987. AllozymeInheritance, Heterozygosity and Outcrossing Rates among Pinus monticola NearLadysmith, British Columbia. Heredity. 58:173-181.Endler, J.A. 1977. GeographicvaiLadm,spggutio„_Id Clines.. Princeton UniversityPress, Princeton. 246 p.49Chapter 2 Random Amplified DNA FingerprintingForde, M.B., and M.M. Blight. 1964. Geographical Variation in the Turpentine ofBishop Pine. New Zeal. J. Bot. 2:44-52.Gilbert, D.A., C. Packer, A.E. Pusey, J.C. Stephens, and S.J. O'Brien. 1991. AnalyticalDNA Fingerprinting in Lions: Parentage, Genetic Diversity, and Kinship. J.Hered. 82:378-386.Guries, R., and F.T. Ledig. 1982. Genetic Diversity and Population Structure in PitchPine (Pinus rigida Mill.). Evolution 36:378-402.Hamrick, J.L., and M.J.W. Godt. 1990. Allozyme Diversity in Plant Species. In Brown,A.H., M.T. Clegg, A.L. Kahler, and B.S. Wier. (Eds.), Plant PopulationBr e^ar1gjRegiefic Resources. Sinauer. Inc. Sunderland, Mass. pp. 43-63.Hamrick, J.L., J.B. Mitton, and Y.B. Linhart. 1981. Level of Genetic variation in ForestTrees . Influence of Life History Characteristics. In Conkle, M.T. (Ed.), Isozymesof North American Forest Trees and Forest Insects. USDA Gen. Tech. Rep.PSW-48. pp35-41.Hamrick, J.L., M.J.W. Godt, and S.L. Sherman-Broyles. 1992. Factors InfluencingLevels of Genetic Diversity in Woody Plant Species. New Forest 6:95-124.Hartl, D.L. 1980. Principles of Population Genetics. ^Sinauer Associates, Inc.Sunderland, Mass.Harry, D.E. 1987. Shoot Elongation and Growth Plasticity in Incense-cedar. Can. J. For. Res. 17:484-489.Jain, S.K., and A.D. Bradshaw. 1966. Evolutionary and Divergence Among AdjacentPlant Populations. I. Evidence and Its Theoretical Analysis. Heredity 21:407-442.Jeffreys, A.J., V. Wilson, and S.L. Thein. 1985. Hypervariable "Minisatellite" Regions inHuman DNA. Nature 314:67-73.Li, P. and W.T. Adams. 1989. Range-wide Patterns of Allozyme Variation in Douglas-fir(Pseudotsuga menziesii ). Can. J. For. Res. 19:149-161.Li, W-H, and D. Graur. 1991. Fundamentalsof Molecular Evolution. Sinauer Associates,Inc. Publishers. Sunderland, Massachusetts.Li, W. -H., C. -I. Wu, and C.-C. Luo. 1985. A New Method for Estimating Synonymousand Nonsynonymous Rates of Nucleotide Substitution Considering the RelativeLikelihood of Nucleotide and Codon Changes. Mol. Biol. Evol. 2:150-174.Mayr, E. 1963. Animal Species and Evolution.  Harvard University Press , Cambridge.797p.Mejnartowitcz, L. 1976. Genetic Investigation on Douglas-fir [Pseudotsuga menziesii(Mirb.) Franco] Populations. Aboretum Kornickie 21:126-187.50Chapter 2 Random Amplified DNA FingerprintingMejnartowitcz, L., and F. Bergmann. 1984. Genetic Differentiation among Scots PinePopulations from the Lowlands and the Mountains in Poland. Lecture Notes inBiometric 60:253-266.Merkle, S., and W.T. Adams. 1987. Patterns of Allozyme Variation within and amongDouglas-fir Breeding Zones in Southwest Oregon. Can. J. For. Res. 18:181-187.Milgroom, M. G., S.E. Lipari, and W.A. Powell. 1992. DNA Fingerprinting and Analysisof Population Structure in the Chestnut Blight Fungus, Cryphonectria parasitica.Genetics. 131:297-306.Millar, C.I., and W.J. Libby. 1991. Strategies for Conserving Clinal, Ecotypic, andDisjunct Population Diversity in Widespread Species. In. Falk, D.A., and K.E.Holsinger (Eds), Genetics and Conservationof Rare Plants. Oxford UniversityPress, New York. pp. 149-170.Mitton, J.B. 1983. Conifers. In Tanksley, S.D., and T.J. Orton. (Eds.), Isozyme in PlantGenetics and Breeding. Part B. Elsevier Science Publishers B.V., Amsterdam. pp.443-472.Mitton , J.B., K.B. Sturgeon, and M.L. Davis. 1980. Genetic Differentiation in PonderosaPine along a Steep Elevational Gradient. Silvae Genet. 29:100-103.Moran, G.F., and W.T. Adams. 1989. Microgeographical Patterns of AllozymeDifferentiation in Douglas-fir from Southwest Oregon. Forest Science 35:2-15.Mouna, 0., and A.E. Szmidt 1984. A Multilocus Study of Natural Populations of Pinussylvestris. In Gregorius, H.-R. (Ed), Population Genetics in Forestry. LectureNotes in Biomethemetics 60. Springer-Verlag. Berlin Heidelberg. pp. 226-240.Mullis, K.B., and F.A. Faloona. 1987. Specific Synthesis of DNA in vitro via aPolymerase Catalyzed Chain Reaction. Methods Enzymol. 155:335-350.National Council on Gene Resources. 1982. Douglas-fir Genetic Resources. AnAssessment and Plan for California. California Gene Resources Program,National Council on Gene Resources, Berkeley, California.Neale, D.B., and W.T. Adams. 1985. The Mating System in Natural and ShelterwoodStands of Douglas-fir. Theor. Appl. Genet. 71:201-207.Nei, M. 1973. Analysis of Gene Diversity in Subdivided Populations. Proc. Natl. Acad. Sci. USA. 70:3321-3323.Nei, M. 1978. Estimation of Average Heterozygosity and Genetic Distance from a SmallNumber of Individuals. Genetics 89:583-590.Nei, M. 1986. Stochastic Errors in DNA Evolution and Molecular Phylogeny. InGershowitz, H., D.L. Rucknagel, and R.E. Tashian, (Eds.), Proceedings ofInternational Symposium Honoring Dr. James V. Neel. Evolutionary Perspectivesand the New Genetics. pp. 133-147.51Chapter 2 Random Amplified DNA FingerprintingNei, M., and R.K. Chesser. 1983. Estimation of Fixation Indices and Gene Diversities.Ann. Hum. Genet. 47:253-259.Plessas, M.E., and S.H. Strauss. 1986. Allozyme Differentiation Among Populations,Stands, and Cohorts in Monterey Pine. Can. J. For. Res. 16:1155-1164.Ryan, A.W., E.J. Duke and J.S. Fairley. 1993. Polymorphism, Localization andGeographical Transfer of Mitochondrial DNA in Mus muculus domesticus (Irishhouse mice) Heredity 10:75-81.Sherman-Broyles, S.L., S.B. Bryles, and J.L. Hamrick. 1992. Geographic Distribution ofAllozyme Variation in Ulmus crassifolia. Syst. Bot. 17:33-41.Silen, R.R. 1978. Genetic of Douglas-fir. USDA For. Serv. Res. Pap. Wo-35. 34 p.Soltis, D.E., M.S. Myer, P.S. Soltis, and M. Edgerton. 1991. Chloroplast DNAGenotypes Variation in Tiarella trifoliata (Saxifragaceae). Am. J. Bot. 78: 1379-1390.Steinhoff, R. J., D.G. Joyce, and L. Fins. 1983. Isozyme Variation in Pinus monticola.Can. J. For. R&. 13:1122-1132.Strauss, S.H., R. Lande, and G. Namkoong. 1992. Limitations of Molecular-marker-aided Selection in Forest Tree Breeding. Can. J. For. Res. 22:1050-1061.Swofford, D.L., and R.B. Selander. 1989. BIOSYS-1 (release 1.7): User's Manual.Department of Genetics and Development, University of Illinois at Urbana-Champaign, Urbana.Triggs, S.J., M.J. Williams, S.J. Marshall, and G.K. Chambers. 1992. Genetic Structureof Blue Duck (Hymenolaimus malacorhynchos) Populations Revealed by DNAFingerprinting. The Auk. 109(1):80-89.Tulsieram, L.K., J.G. Glaubitz, G. Kiss, and J.E. Carlson. 1992. Single Tree GeneticLinkage Mapping in Conifers Using Haploid DNA from Megagametophytes.Bio/Technology 10:686-690.Wagner, D.B., G.R. Furrier, M.A. Saghai-Maroof, S.M. Williams, B.P. Dancik, and R.W.Allard. 1987. Chloroplast DNA Polymorphisms in Lodgepole Pine and JackPines and Their Hybrids. Proc. Natl. Acad. Sci USA. 84:2097-2100.Weir, B.S. 1990. Intraspecific Differentiation. In. Hillis, D.M., and C. Moritz, MolecularSystematics. Sinauer Association, Inc Publisher. Sunderland. pp. 373-410.Welsh, J., and M. McClelland. 1990. Fingerprinting Genomes Using PCR with ArbitraryPrimers. Nucleic Acids Research 18(24):7213-7218.Wheeler, N.C., R.P. Guries, and D.M. O'Malley. 1983. Biosystematic of Genus Pinus,Subsection Contortae. Biochem. Syst. Ecol. 11:333-340.52Chapter 2 Random Amplified DNA FingerprintingWhitkus, R. 1988. Modified Version of GENESTAT: A Program for Computing GeneticStatistics from Allele Frequency Data. Plant Genet. Newslet. 4:10Wilde, J., R. Waugh, and W. Powell. 1992. Genetic Fingerprinting of Theobroma ClonesUsing Randomly Amplified Polymorphic DNA Markers. Theor. Appl. Genet.83:871-877.Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski, and S.V. Tingey. 1990. DNAPolymorphisms Amplified by Arbitrary Primers are Useful as Genetic Markers.Nucleic Acids Res. 18:6531-6535.Woodruff, D.S., and G.A.E. Gall. 1992. Genetic and Conservation. Agric. Ecosystemsand Environ. 42:53-73.Yang, J. -Ch., T.M. Ching, and K.K. Ching. 1977. Isozyme Variation of CoastalDouglas-fir. I. A Study of Geographic Variation in Three Enzymes. SilvaeGenet. 26:10-18.Yeh, F.C. 1981. Analyses of Genetic Diversity in Some Species of Conifers. In Conkle,M.T. (Tech. Coord.), Proc. Symp. Isozymes of North America Forest Trees andForest Insects. USDA For. Serv. Gen. Tech. Rep. PSW-48. pp. 48-52.Yeh, F.Ch.-H., and D. O'Malley. 1980. Enzyme Variations in Natural Populations ofDouglas-Fir, Pseudotsuga menziesii (Mirb.) Franco, from British Columbia. 1.Genetic Variation Patterns in Coastal Populations. Silvae Genet. 29:83-92.Yeh, F.C., and Y.A El-Kassaby. 1980. Enzyme Genetic Variation in Natural Populationsof Sitka Spruce [Picea sitchensis (Bong.) Cam]. I. Genetic Variation Patternsamong Trees from Ten IUFRO Provenances. Can. J. For. Res. 10:415-422.53CHAPTER 3CHLOROPLAST DNA VARIATION OF DOUGLAS-FIR IN B.C.3.1 INTRODUCTIONIn recent years information from chloroplast DNA (cpDNA) analyses, includingsequencing and restriction site mapping, has increased rapidly. Restriction site data haveproven to be a valuable tool for estimating of genetic diversity within populations as wellas for resolving phylogenetic relationships (Nei and Miller, 1990; Saitou and Nei, 1987;McDonald and Mabry, 1992). CpDNA in conifers is paternally inherited (Neale et al.,1986; Neale and Sederoff, 1989; Wagner et al., 1989, and 1992), while it is maternallyinherited in most angiosperms (Palmer, 1987). CpDNA is passed on from one generationto the next with only an occasional mutation altering the molecule (Schaal et al., 1991).Therefore, cpDNA polymorphism within species is often very low (Banks and Birky,1985), but can occasionally be considerable (Soltis et al., 1989) and may extend beyondspecies limits among very closely related species (Doyle et al., 1989; Hong, 1991). It isnow well established that the chloroplast genome evolves at a conservative rate (Palmer,1987) and the evolutionary rates of chloroplast-encoded genes are considerably belowthose observed for animal or plant nuclear genes (Wolfe et al., 1987). However, Wagneret al., (1987) found that cpDNA variations in Pinus banksiana and P. contorta areexceptionally high and sufficient for studies at the population level. Milligan, (1991) alsodemonstrated that plastid genomes in some species may not be as invariant as previouslybelieved, but may instead exhibit high levels of genetic diversity at the population level.In genetic diversity studies, patterns of cpDNA and mtDNA variation are notnecessarily congruent with patterns of morphological or allozyme variation (Moore et al.,1991). The patterns of geographic variation in cpDNA could differ from those ofmtDNA, and also nucleotide substitution in mtDNA is less than one-third that in cpDNA54Chapter 3 Chloroplast DNA Varation of Douglas-fir in B.C.(Wolfe et al., 1987; Li and Graur, 1991). Wolfe et al., (1987) concluded that the fasterrate in cpDNA evolution than in mtDNA is probably due to a faster mutation rate . Thusit is possible that geographic structure has evolved differently in relatively youngpopulations with regard to cpDNA and mtDNA. Many species of conifers are known tobe geographically and morphologically divergent (Conkle, 1981; Hamrick et al., 1981),which lead us to look for differences at the DNA level.A number of traits have been used by taxonomists, evolutionary biologist, andgeneticists to develop a picture of distribution of genetic variationin natural populations,including morphology, development and physiology, and bicchemical traits (NationalCouncil on Gene Resources, 1982). Most morphological characters are known to beeasily susceptible to environmental influences and may fail to provide the fine resolutionfor phylogenetic relationships among species (Hillis, 1987). In recent studies, DNAmarker can provide direct and genetically interpretable measures of individual andpopulation level variation (Woodruff and Gall, 1992). Of the three genomes (nuclear,chloroplast, and mitochondrial) the chloroplast genome has proven to be the most usefulfor genetic diversity and phylogenetic analyses due to its high copy number (Olmstead etal., 1989) and high nucleotide substitution rate. Restriction site analysis is mostappropriate for phylogenetically closely related organisms (Olmstead et al., 1989).However, the relationships of the same genera in different geographic regions may not besimilar. As reviewed in Sytsma (1990), several sections of Oncidium in Brazil, forexample, are phylogenetically more closely related, based on cpDNA, to other genera inthe same region than they are to other sections of Oncidium with similar floramorphologies occurring in other geographical areas. Also, the present geographicdistribution of individual species often shows little correlation with the degree of cpDNAdifferentiation among species (Szmidt et al., 1988).55Chapter 3 Chloroplast DNA Varation of Douglas-fir in B.C.To date, several investigators have used molecular characters derived fromcpDNA to study genetic diversity and to infer phylogenetic relationships among plantgenera. However, there are very few reports on genetic diversity of forest tree speciesestimated with cpDNA markers. In agriculturally important genera, restriction fragmentanalyses of cpDNA have provided phylogenetic information at the interspecific level,such as Lycopersicon-Solanum (tomato, potato), Nicotiana (tobacco), Triticum-Aegilops(wheat), Brassica-Raphanus (cabbage, mustard), Pisum (pea), Cucumis (melon,cucumber), Linum (flax), and some non-crop genera (Palmer, 1987). A phylogeny basedon RFLP of cpDNA has recently been reported in Glycine (Doyle et al., 1990) Hordeum(Baum and Bailey, 1991), Ranunculaceae (Johansson and Jansen, 1991), Ipomoea(Convolvulaceae) (McDonald and Mabry, 1992), and Gossypium (cotton) (Wendel andAlbert, 1992). In forest trees, the chloroplast molecule has been used to studyphylogenetic relationships among pines (Strauss and Doerksen, 1990; Hong, 1991),Pseudotsuga (Strauss et al., 1990) and Salix (Brunsfeld et al., 1992). Results fromphylogenetic studies based on cpDNA polymorphism agreed better with classicaltaxonomy than with such genetic factors as crossability patterns among species (reviewedin Szmidt, 1991).The utilization of molecular genetic markers to clarify relationships that aredifficult to ascertain with morphology will increase, especially in plant groups thatexhibit remarkable divergence related to adaptive radiation (Sytsma, 1990). Although,some cpDNA sequences as well as mtDNA have been found in nuclear DNA (Kemble etal., 1983), these organelle DNAs are largely independent of events dictated by the cellnucleus (Cann, 1989). In addition, cpDNA plays a role in photosynthesis that is essentialfor plant survival. Some chloroplast genes are of particular interest to thebiotechnologist, including valuable characters such as a mutant gene giving enhanced56Chapter 3 Chloroplast DNA Varation of Douglas-fir in B.C.resistance to herbicides (reviewed in Strauss et al., 1989). Therefore, genetic diversityand phylogenetic relationships based on cpDNA in valuable tree species should be carriedout to understand the evolution of the species so that tree breeders can use this knowledgein breeding, testing and selecting trees.The objective of the present study is to utilize RFLP analysis of chloroplast DNAto assess genetic variation of Douglas-fir populations from three geographic regions inB.C., including coastal, transitional, and interior.3.2 MATERIALS AND METHODS3.2.1 Plant materialSix populations of Douglas-fir from three geographic regions in B.C. (coastal,transitional, and interior) were sampled (Figure 3.1). The coastal region samples werefrom Vancouver Island, and the populations were designated based on seed stratificationzones (A, B, and C). In each coastal zone, 2 populations were chosen, based onelevation. Transitional populations were located along the zone of introgression ofcoastal and interior Douglas-fir. Interior populations were chosen from seed orchardplanning zones from the interior of B.C. (Table 3.1). Four individuals were sampled fromeach population. Young needle samples of coastal and transitional Douglas-fir werecollected from a clone bank at Cowichan Lake Research Station and interior Douglas-firsamples were collected from Kalamalka Research Station & Seed Orchard, Vernon, B.C.57Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Figure 3.1 Approximate locations of coastal, transitional, and interior Douglas-firpopulations used in studying cpDNA variation of the species in B.C.(Al, Bl, Cl - coastal lower elevation; A2, B2, C2 - coastal higher elevation;T1 - T6 - transition; Interior -- CP - Central Plateau, CT - Cariboo Transition,MR - Mt Robson, MI - Mica, SA - Shuswaps Adams, WK - West Kootenay).58Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Table 3.1 Sources of needle and seed samples used for studying cpDNA variation inDouglas-firPopulations Locations Sources of materialsLatitude(°N)Longitude(°W)Elevation(m)CDFAl 48°-49° 123°-125° 450-600Cowichan Lake Research Station,Mesachie Lake, B.C. (needle andA2 48°-49° 123°-125° > 600 cone).B1 49°-50° 125° 0-150B2 48°-49° 123°-125° 150-300Cl 49°-50° 125°-127° 0-150C2 49°-50° 125°-127° 150-300TDFTl 49° 121° 610-1158Cowichan Lake Research Station,Mesachie Lake, B.C.T2 50° 122° 762-975T3 50° 123° 366-742T4 50° 124° 128-720T5 510 124° 37-536T6 52° 126° 98-439IDF Kalamalka Research Station &CP 53°-54° 122°-124° 610-900 Seed Orchard, Vernon, B.C.CT 51°-52° 121 0 -1220 760-1000SA 50° 118°-119° 650-1250WK 49° 116°-117° 659-900MR 52°-53° 119°-120° 732-1140MI 51° 118° 585-685CDF - Coastal Douglas-firTDF - Transitional zone Douglas-firIDF^- Interior Douglas-fir59Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.3.2.2 Isolation and restriction of genomic DNATotal genomic DNA was prepared from young needles using a modified CTABmethod (DeVerno et al., 1989; Wagner et al., 1987). Needles were kept in a -70°Cfreezer and were ground using a coffee grinder machine. Twenty grams of groundmaterial from each sample were used to extract DNA.Twenty grams of powder were homogenized again in 150 ml extraction buffer [50mM Tris HC1 (pH 8.0), 5 mM EDTA, 0.35 M Sorbitol, 0.1% Bovine Albumin Serum(BSA), 10% Polyethylene Glycol (PEG4000), 0.1% Spermine, 0.1% Spermidine, and justprior to use, 0.1% B-mercaptoethanol was added] using a Polytron (PT3000, Brinkmann)for 3 x 5 seconds at high speed with 10 seconds between pulses.The homogenate was filtered through 4 layers of cheese cloth and 1 layer ofmiracloth into a 350 ml centrifuge bottle and spun at 9,000 rpm for 10 minutes using aBeckman J2-21 centrifuge with JA-10 fixed angle at 4°C. The pellet was resuspended in15 ml wash buffer (50 mM Tris HC1 (pH 8.0), 25 mM EDTA, 0.35 M Sorbitol, just priorto use, add 0.35 M B-mercaptoethanol) using a paint brush and the suspension transferredto a 50 ml Falcon 2027 screw cap tube. The temperature of the suspension was thenbrought up to room temperature (over 5-10 minutes). Then 1/5 volume (3 ml) of 5%sarkosyl was added to the suspension, mixed by swirling and incubated at roomtemperature for 10 minutes.One seventh volume (2.5 ml) of 5 M NaC1 was added and mixed. After that 1/10volume (2 ml) of 8.6% CTAB in 0.7 M NaC1 was added, mixed and incubated in a 60°Cwater bath for 10 minutes. After incubation, the bottle was cooled at room temperaturefor 10-15 minutes. An equal volume of chloroform : isoamyl alcohol (24:1) was addedand mixed by gently inverting the tube many times until an emulsion was formed (about2 minutes). The tube was spun at 5,000 rpm at room temperature for 10 minutes using a60Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.benchtop centrifuge (Megafuge, Baxter) with a swing-out rotor (C1725-21). The aqueouslayer (upper layer) was transferred to a new 50 ml Falcon screw cap tube. If the aqueouslayer was too green, chloroform extraction would be repeated. DNA was precipitated inan equal volume of ice cold absolute ethanol and mixed gently by inverting the tubeseveral times. The precipitated DNA was hooked out using a bent pasteur capillary pipetand transferred to a 15 ml Corex tube containing 5 ml of 7.6% ethanol, 0.2 M sodiumacetate. DNA was left on the hook in the tube for 20 minutes or longer to remove anyresidual pigments. DNA was then rinsed in 1 ml of 7.6% ethanol, 10 mM ammoniumacetate in an eppendorf tube and air dried for 10 - 20 minutes.DNA was dissolved in 1 ml TE Buffer (10:1) pH 8.0 (10 mM Tris HCI, 1 mMEDTA) (depending on the size of DNA). Quantitation of purified DNA was based onabsorbance of ultraviolet light at wavelength 260 nm and the DNA concentration wasadjusted to be 0.5 pg/gl.Restriction endonuclease digestion was carried out under the conditionsprescribed by the supplier. Ten restriction enzymes (six-base cutter), including Barn HI,Dra I, Eco RI, Eco RV, Hind III, P st I, Pvu II, Sst I, Xba I, and Xho I, were used to digestgenomic DNA. Each reaction mixture usually contained 5-10 mg DNA and 25-50 unitsof the restriction endonuclease, and reactions proceeded for more than 12 hours at 37°C.3.2.3 Electrophoresis and alkaline transfers of genomic DNA0.8% agarose gels were prepared in lx TAE buffer (8 mM Tris-HC1, 1mM sodiumacetate, and 0.4 mM EDTA, pH 8.0) and run in the same buffer. Electrophoresis wasperformed in a horizontal slab gel apparatus at constant voltage (22-40 V) overnight (or15-20 hours) with occasional recirculation of running buffer. A HindIII digest ofbacteriophage X DNA was used to provide molecular size markers. After electrophoresis,61Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.gels were stained in 1 pg/m1 ethidium bromide, visualized on UV light box andphotographed using a polaroid film.DNA was transferred from the gel to a Hybond-N± membrane (Amersham) in0.4M sodium hydroxide (NaOH) according to the alkaline capillary transfer procedure(Molenaar and Wilkins, 1991). To study large fragments (>10 Kb), DNA was firstdepurinated by soaking the gel in several volumes of 0.25 M HCl with gentle shaking forabout 9 minutes, then drained and rinsed once with distilled water. DNA was denaturedby soaking the gel in several volumes of 0.4 M NaOH with gentle shaking for 20-30minutes. The transfer proceeded overnight. The filter was rinsed briefly in 6x SSC (pH7.0) and placed between folded Whatman 3 MM paper and dried for about 1 hour at roomtemperature. Then the filter was baked in a 80 °C vacuum oven for 2 hours. Finally themembrane was stored in sealed plastic bag or plastic wrap (Handi wrap) at 4 °C or roomtemperature until use. Prior to first use, the membrane was prewashed in a solution of0.1x SSC, 0.5% SDS, at 65°C for 1 hour.3.2.4 Labeling cpDNA probe and hybridizationTwenty-three cloned cpDNA fragments from lodgepole pine (Pinus contorta)(Lidholm and Gustafsson, 1991) were used as probes for hybridization. The overlappingclones comprise 96% of the chloroplast genome. Four sets of pooled probes weredesignated, consisting of 4-8 cpDNA clones in each set (Figure 3.2).Probes were labeled with 32P-dCTP via primer extension with random hexamersas primers (Fienberg and Vogelstein, 1983). Hybridization proceeded overnight at 65°Cin a (Robbins Scientific) hybridization incubator.621 2a,^Probe set(.,..) 3 4 HYV--' ^/v\gidA psbB rbcLpsbE-psbF psbAlK178^B28^B206^ K32^K79^K50^- B623^B754X914 E243 E94^ E43 H46^X914H326^H220^H273^11228^H302^H63^H380^H157^H132^X789Size (kb)^0^10^20^30^40^50^60^70^80^90^100^110^120Figure 3.2 Lodgepole pine cpDNA restriction site map (Lidholm and Gustafsson, 1991) and probe sets used in hybridization.Clones not used as hybridization probes, were not mapped. (B-BamHI,E-EcoRI,H-HindIII,K-KpnI, X- XbaI),Not include in probe set.Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.After hybridization, the blots were washed once in 2x SSC in hybridization tubesin the incubator at 65°C for 5 minutes and two more times in 0.2x SSC, 0.5% SDS, and0.1x SSC, 0.1% SDS at 65°C for 30 minutes each. The blots were wrapped with Handiwrap and exposed to Kodak XAR 5 with intensifying screen at -80°C for 3 days to 1week before autoradiographs were developed. The old probes were stripped off bywashing once in 0.4 N NaOH at 42°C for 30 minutes and one more time in washingsolution [0.1x SSC, 0.5% SDS, 0.2 M Tris-HC1 (pH 7.5)] at 42°C for 30 minutes. Theblots were reprobed 10-20 times.Six restriction enzymes, including BamHI, EcoRI, EcoRV , HindIII, SstI and XbaI,were chosen and twenty-four enzyme probe combinations were used to evaluate geneticvariation of cpDNA.3.2.5 Data analysisThe evaluation of DNA variation was considered in two categories, includingallelic frequencies and haplotype frequencies. Because of uncertain identification of themutational basis of RFLPs, restriction enzyme and probe combination was designated asa locus and the observed restriction fragment pattern from each combination wasdesignated as an allele. Then allelic frequency was calculated by counting alleles at eachlocus.Haplotypes were designated by unique restriction fragment patterns from allrestriction enzyme and probe combinations for an individual. The frequency ofhaplotypes in each population was estimated and used to calculate genetic differentiation.The number of observed haplotypes was used to calculate haplotype frequency. Bothallelic and haplotypes frequencies were used in further quantitative analyses.64Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Genetic diversity and the degree of population differentiation for unbiased samplesize (Nei and Chesser, 1983) were estimated using the computer package GENESTAT-PC version 2.1 (Whitkus, 1988).3.3 RESULTS AND DISCUSSION3.3.1 Chloroplast DNA variationA total of 175 bands were observed. Thirty restriction sites were detected to bepolymorphic in all 4 sets of hybridization probes (Table 3.2). However, only a fewindividuals showed variation. Based on the resolution in the autoradiograms, the cause ofthe cpDNA variation observed in this study could not be verified. When a point mutationcauses a restriction site change, three different restriction fragments between the twoindividuals are expected. But two different restriction fragments were always observedwhich suggest length mutations. However, large length mutations in cpDNA are knownto be not tolerated as less than 10% of the genome is non-coding region. Therefore,length mutations in the chloroplast genome may not survive.Length mutation was observed using H302 probe (a probe in set 3) hybridized toblots with Xbal digested DNA (H302/XbaI). Four different fragment sizes were detected(Table 3.2). Based on restriction enzyme and probe combinations, twenty-four loci with1 to 3 alleles in each locus were obtained (Table 3.3). For haplotype estimates, 16haplotypes were characterized among all of the Douglas-fir populations. (Tables 3.4-3.5).From these haplotypes, coastal and interior populations exhibited 5 and 7 haplotypes,respectively. Transitional populations showed 11 haplotypes. Haplotype I (40%) wasmost frequently observed over all populations (29% in coastal, 29% in transitional, and62% in interior regions). However, 42% of the individuals in coastal populations were in65Chapter 3 Chloroplast DNA Variation of Douglas-fir in B. C.haplotype II. On the basis of haplotype analysis, the results indicate that cpDNA inDouglas-fir in the transition zone is more polymorphic than in coastal and interiorregions. The level of polymorphism of cpDNA observed in this study is quite low,however, which is a common observation in many plant and forest tree species (Banksand Birky, 1985; Hong, 1991). Due to the slow rate of evolution of cpDNA, thepopulation diversity is frequently low (Birky et al., 1989). No variation can be observedin the chloroplast genome of some species. In Pinus torreyana, for example, Waters andSchaal, (1991) found no variation in the chloroplast genome in two separate populationsfrom San Diego and Santa Rosa island.66Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Table 3.2 Polymorphisms observed in cpDNA of Douglas-fir after digesting with 6different enzymes and hybridizing with 4 different probes[*-indicates probe set (1- 4), A, B, C, D, E, and G - indicate band sizes in Kbin each enzyme and probe combination]Enzyme^Sizeand^(Kb)ProbePopulationsAl A2 B1 B2 Cl C2 Ti T2 T3 T4 T5 T6 CP CT SA WK MR MIBamHI2*^C 2.75 - 1 - -EcoRI1 E 1.52 - - 12 B 4.29 4 4 4 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4D 1.52 - 1 - -EcoRV1 B10.91 - - - 12 A 1.90 - 1 - 1 13 E 3.06 - 1 -HindIII2 A15.30 4 4 4 3 4 4 4 3 4 4 3 4 4 4 4 4 4 4C 8.35 4 4 4 3 4 4 4 3 4 4 3 4 4 4 4 4 4 4D 4.69 - - - - - - 1 - - -E 3.40 - - - - - - 1 - - - -3 A14.54 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 4 4B11.20 - - - - 1 - - - - -C 9.79 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 4 44 Al2.28 - - 1 - - - - - -Sstl2 C 4.10 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 4 43 A15.85 4 4 4 4 4 4 3 3 4 3 4 4 4 4 4 4 4 4B10.44 - - - I - - 1 - - -C 8.88 - - - - - 1 - - - - -4 A16.44 4 4 4 4 4 4 3 3 4 4 4 4 4 4 4 4 4 4B11.72 4 4 4 4 4 4 3 3 4 4 4 4 4 4 4 4 4 4C11.50 - - 1 - - - -E 8.50 - 1 - -F 8.26 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 4 4Xbal1 E 2.61 - - 1 - - - -2 G 5.48 4 4 3 4 4 3 4 4 4 4 4 4 4 4 4 4 4 43 A 2.94 3 1 1 1 - 3 2 2 2 1 2 2 3 4 1 1 3 4B 2.79 1 2 1 - - 1 - 1 2 -C 2.74 1 2 1 3 3 1 2 2 2 2 2 2 1D 2.56 - - - - 2 2 1Xbal / probe set 3 (H302) showed length mutation.67Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Table 3.3 Haplotype frequency of coastal Douglas-fir populations based oncpDNA probes (* represent probe set and locus)Locus Allele PopulationsAl A2 B1 B2 Cl C2 T1 T2 T3 T4 T5 T6 CP CT SA WK MR MIBamHI1 * 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - - - - - - - - - - 0.25 - - - - - -3 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.004 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00EcoRI1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - 0.25 - - - -2 1 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.002 0.25 - - - -3 - 0.25 - - - - -3 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.004 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00EcoRV1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - 0.25 - -2 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 0.75 0.75 0.752 - - 0.25 - - - - -3 - - - - 0.25 0.25 0.253 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - 0.25 - - - - -4 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00Madill1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 1 1.00 1.00 1.00 0.75 1.00 1.00 1.00 0.75 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.002 0.25 0.25 - -3 - - - - - 0.25 - - - - -3 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - - 0.25 - - - -4 1 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - 0.25 - -SstI1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - - 0.25 - - - - - -3 1 1.00 1.00 1.00 1.00 1.00 1.00 0.75 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - - 0.25 - - -3 - - - 0.25 - - - - -4 1 1.00 1.00 1.00 1.00 1.00 1.00 0.75 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - - 0.25 - - - - - -3 - - - 0.25Xbal1 1 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.002 - 0.25 - - -2 1 1.00 1.00 0.75 1.00 1.00 0.75 0.50 0.50 0.75 0.75 0.75 0.75 1.00 1.00 1.00 1.00 1.00 1.002 0.25 0.25 - 0.25 0.25 - 0.25 - - - -3 - - - - 0.50 0.50 - 0.25 - - - - - -3 1 0.75 0.25 0.25 0.25 - 0.75 0.50 0.50 0.50 0.25 0.50 0.50 0.75 1.00 0.25 0.25 0.75 1.002 - 0.25 0.50 - 0.25 - - 0.25 - 0.25 - 0.50 - -3 0.25 0.50 0.25 0.75 0.75 0.25 0.50 0.50 0.50 0.50 0.50 0.50 - - 0.25 - -4 - - - - - - - - - 0.50 0.25 0.25 -4 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.0068Table 3.4 Haplotype designation for cpDNA analysis. (A, B, C, D, E, F, and G refer to DNA fragment sizes in Table 4.2,+ and - indicate presense and absence of a fragment)HaplotypesB1 EcoRI EcoRV HindIII SstI XbaI1 2 1 2 1 2 3 1 3 4 2 3 4 1 2 3CE B D B A E ACDE ABC AC A BC A BCE F E G A BCD1 - - +^- - - - +^+^-^- +^- - - + +^-^+ +^+ - -^+ - - + -^- -2 - - +^- - - - +^+^-^- +^- - - + +^-^+ +^+ - -^+ - - - -^+ -3 - - +^- - - - +^+^-^- +^- - - + +^-^+ +^+ - -^+ - - - +^- -4 - - +^- - - - +^+^-^- +^- - - + +^-^+ +^+ - -^+ - - - -^- +5 - - +^- - - - +^+^-^- +^- - - + +^-^+ +^+ - -^+ - + - +^- -6 - - -^- - - - -^-^+^- +^- - + + +^-^+ +^+ - -^+ + + - -^+ -7 - - +^- - - - +^+^-^- +^- - - + +^-^+ +^+ - -^+ + + + -^- -8 - - +^- - - - +^+^-^- +^- - - + +^+^+ -^- + -^+ + + - +^- -9 - - +^- - - + +^+^-^- +^- - - + +^-^+ +^+ - -^+ - + + -^- -1011 - - +^- - - - +^+^-^- + -^- - + +^-^+ +^+ - -^+ - + + -^- -12 - - +^- - - - -^-^+^+ + -^+ - + +^-^+ +^+ - -^+ - + - -^+ -13 + + +^- + + - +^+^-^- +^- - - + +^-^+ +^+ - -^+ - + + -^- -14 - - +^- - + - +^+^-^- + -^- - + +^-^+ +^+ - -^+ - + - -^+ -15 - - +^- - + - +^+^-^- + -^- - + +^-^+ +^+ - -^+ - + - -^- +16 - - +^- - + - +^+^-^- + -^- - + +^-^+ +^+ - -^+ - + + -^- -1 - BamHI2 probe set.Table 3.5^Haplotypes observed from the RFLP analysis of cpDNA (as percent of total)Population HaplotypesI II III IV V VI VII VIII IX X XI XII XIII XIV XV XVICoastal Al .75 .25 - - - - -Coastal A2 .25 .50 .25Coastal B1 .25 .50 .25Coastal B2 .25 .50 - .25 - -Coastal Cl .75 .25Coastal C2 .50 .25 .25Transition 1 .50 .25 .25,)c) Transition 2 .50 - - .25 .25Transition 3 .25 .50 .25Transition 4 .25 .25 .25 .25 -Transition 5 .25 .25 .25 .25Transition 6 .50 .25 .25 - - -Interior CP .75 .25 - - -Interior CT 1.0 - -Interior SA .25 .25 - .50Interior WK .25 .25 .25 - .25Interior MR .75 - .25Interior MI .75 - - - .25Total .40 .22 .10 .042 .042 .014 .028 .014 .014 .014 .042 .014 .014 .014 .014 .014Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.3.3.2 Estimates of cpDNA diversityOn the basis of cpDNA allelic frequencies, within population genetic diversity(Hs) estimates in Douglas-fir in all three regions ranged from 0.106 (interior) to 0.189(transition) (Table 3.6). Transition zone populations exhibited the highest geneticdiversity. Genetic diversity among populations (D sT) in all three regions was very low.Population subdivision (GsT) values for Douglas-fir in all three different geographicregions were 0.016, 0.011, and 0.105 for coastal, transition, and interior populations,respectively. The results suggest that about 2% to 10% of the total genetic diversity isdue to interpopulation genetic diversities, whereas on average 90% to 98% of geneticdiversity in Douglas-fir is intrapopulational.Table 3.6 Hierarchical partitioning of genetic diversity estimation based onallelic frequencies of cpDNALevel of Analysis Hs HT DST GSTTotal population 0.140 0.153 0.013 0.081Geographic regions 0.058 0.063 0.004 0.071Coastal 0.127 0.129 0.002 0.016Transition 0.189 0.191 0.002 0.011Interior 0.106 0.118 0.012 0.105Hs = gene diversity within populations; HT= total gene diversity; DsT gene diversity amongpopulations (D sT = HT - Hs), GsT= proportion of total gene diversity resulting from geneticdifferentiation among populations (G sT = DsT/HT)•The estimates of cpDNA diversity based on haplotype frequencies at differenthierarchical levels are given in Table 3.7. In this cpDNA study, high degrees of cpDNAdiversities within populations (Hs) were exhibited in all regions and all values were71Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.higher than genetic diversities among populations (D sT). Although the degrees ofdiversity among all three regions was not significantly different, transition zonepopulations showed higher level of diversity than interior and coastal populations .Table 3.7 Hierarchical partitioning of genetic diversity estimation based onhaplotype frequencies of cpDNALevel of Analysis Hs(±SE)HT(±SE)DsT(±SE)GsT(±SE)Total population 0.619 0.776 0.156 0.201(0.226) (0.286) (0.068) (0.086)Geographic regions 0.724 0.777 0.053 0.071(0.352) (0.364) (0.029) (0.041)Coastal 0.619 0.718 0.099 0.138(0.384) (0.456) (0.080) (0.110)Transition 0.762 0.853 0.091 0.106(0.486) (0.348) (0.070) (0.082)Interior 0.476 0.590 0.114 0.193(0.321) (0.383) (0.082) (0.134)Hs = gene diversity within populations; HT total gene diversity; DsT gene diversity amongpopulations (D sT = HT - Hs), GsT proportion of total gene diversity resulting from geneticdifferentiation among populations (G sT = DsT/HT), SE estimated using Jackknifing overpopulations (Weir, 1990).In woody plants, reported genetic diversities based on allozyme markers withinspecies ranged from 0.0 to 0.40. (Hamrick et al., 1992), and within populations rangedfrom 0.0 to 0.35. Among gymnosperms, genus Pseudotsuga showed higher geneticdiversity than other pine species both within species and within populations (Hamrick etal., 1992). In a previous study in cpDNA variation based on allelic frequencies of RFLPlength variants in several Pinus species (Hong, unpublished data), higher levels of geneticdiversity within populations (H s = 0.556) than diversity among populations (D sT =72Chapter 3 Chloroplast DNA Variation of Douglas-fir in B. C.0.061) were observed in Monterey pine, whereas a lower level of diversity withinpopulations (H s = 0.107) than diversity among populations (DsT = 0.243) was observed inKnobcone pine. However, no variation was observed in Bishop pine populations and thisresult was the same as observed in Pinus torreyana (Waters and Schaal, 1991).Furthermore, Milligan (1991) found a very high degree of genetic diversity withinpopulations (0.617) of cpDNA in Trifolium pratense and this is comparable to thediversity estimates for Douglas-fir observed in this study.Population subdivision (GsT) in all hierarchical estimates in this study rangedfrom 0.106 (transition) to 0.193 (interior), whereas a wide range of degrees of populationsubdivision from 0.00 to 0.694 were observed in pine populations (Hong, unpublisheddata) based on cpDNA marker. The estimates of population subdivision (G sT) suggestthat populations in the interior region of B.C. have been more subdivided than those ofcoastal and transition regions.In pine species, the low level of observed population subdivision based onisozymes is typical of widespread and continuously distributed populations, such as Pinusbanksiana (Dancik and Yeh, 1983) and P. ponderosa (reviewed in Millar and Libby,1991), while P. torreyana, P. halepensis and P. muricata are distributed as scatteredisolated populations have more genetic diversity among populations (reviewed inHamrick et al., 1992). Yeh and O'Malley (1980) used enzyme electrophoresis to detectvariation in natural populations of Douglas-fir and found moderately high diversitywithin populations (H s 0.154) with almost no differences among populations with GsTonly 0.026. The results in cpDNA diversity from this study show the same patterns asfound by El-Kassaby and Sziklai (1982), and Moran and Adams (1989) for isozymes, i.e.a higher level of genetic diversity within populations than among populations. ForcpDNA diversity of Douglas-fir populations studied here, all three regions can be73Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.classified as possessing high genetic variation within population and low variation amongpopulations and this result showed the same pattern as observed in nDNA (Chapter 2.).The absolute level of genetic diversity observed in Douglas-fir using cpDNA markers inthis study was higher than observed with allozyme markers and higher than for nDNA inthis study with RADF markers.The chloroplast genome is known to be structurally stable. Both the rate ofnucleotide substitution and structural evolution are very low and the two processes occurindependently (Li and Graur, 1991). The synonymous substitution rate in plantmitochondrial, chloroplast, and nuclear genes are in the approximate ratio of 1:3:12.Therefore, the level of genetic diversity in cpDNA should be higher than in mtDNAbased on synonymous substitution rate. CpDNA in Douglas-fir is paternally inherited(Neale et al., 1986) and the variation in cpDNA increases only through mutation.Therefore, cpDNA genes effectively disperse in both pollen and seed. Because themovement of cpDNA genes in Douglas-fir is accomplished by the movement of wind-born pollen and seed, the variation of cpDNA within populations remained high in allthree regions.In Douglas-fir, most seed is dispersed within 60 m of the source, withcomparatively little or no distribution beyond 300 m (Isaac, 1930). The isolation ofDouglas-fir in remote interior regions and the very wide climatic range of areas occupiedmight contribute to the evolutionary force through short distance migration and increasethe degree of population subdivision. Isolation of populations and intracellular geneticdrift can change levels of diversity within stands and are probably important in producingcurrent population variation of discontinuous geographical pattern. In nature, gradualchanges in genetic variation occur over space when populations are large. Since interiorpopulations in this study were sampled from a range of distribution, from south to north,74Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.wider than in coastal and transition regions, diversity within different populations in theinterior region may contribute to the higher level of population subdivision than occurs incoastal and transition zone populations.Genetic diversity estimates based on allelic and haplotype frequencies gavedifferent results. Diversity calculated from allelic frequencies usually resulted in lowervalues than from haplotype frequency, except for GsT. The estimation of geneticdiversity using allelic frequency was calculated by assessing enzyme and probecombinations to assign a locus and many loci were obtained. The individuals werecompared over many loci and an average of the diversity for all loci was presented foreach population. Horeover, linkage of loci in an individual may result in incorrectestimates of genetic variation, and all loci are linked in cpDNA.For haplotype frequency, the estimate of genetic diversity was obtained bypooling all the restriction fragment patterns and each haplotype was assigned as an allele.Since organelle DNA exists as one molecule and is small in size compared to nuclearDNA, each individual was compared as a single haplotype for the whole genome. Onlyone restriction site difference can cause the individual to be totally different from theothers. Therefore, the estimate of genetic diversity by means of haplotype frequencygave a higher value than the estimate using allelic frequency. However, the haplotypefrequency estimate may be the better method for evaluating genetic diversity at thepopulation level, when taking into account the structure of the chloroplast genome.75Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.3.4 CONCLUSIONSVery low levels of polymorphism were detected with cpDNA probes in all of thepopulations sampled. However, transition zone populations had cpDNA morepolymorphic than coastal and interior regions. This contributed to higher degrees ofgenetic diversity estimates for cpDNA within transition zone populations than in coastaland interior populations. Based on cpDNA haplotype frequency estimates, levels ofpopulation subdivision ranging from 0.106 to 0.193 were observed to be not significantlydifferent among the three geographic regions, with interior populations the highest. Thepattern of cpDNA diversity observed in Douglas-fir in B.C. was the same as observed innDNA. Different degrees of genetic diversity in the three regions probably resulted fromdiffering degrees of intracellular random drift on chloroplast DNA mutants. Gene flowvia seed and pollen and founder effect during population expansion are evolutionaryforces in cpDNA diversity.76Chapter 3 Chloroplast DNA Variation of Douglas-fir in B. C.3.5 LITERATURE CITEDBanks, J.A., and C.W. Birky. 1985. Chloroplast DNA Diversity is Low in Wild Plant,Lupinus texensis Proc. Natl. Acad. Sci. USA.  82:6950-6954.Baum, B.R. and L.G. Bailey. 1991. Relationships among Native and Introduced NorthAmerican Species of Hordeum, Based on Chloroplast DNA Restriction-SiteVariation. Can. J. Bot. 9:2421-2426.Birky, Jr, C.W., P. Fuerst, and T. Maruyama. 1989. Organelle Gene Diversity underMigration, Mutation, and Drift: Equilibrium Expectations, Approach toEquilibrium, Effect of Heteroplasmic Cells, and Comparison to Nuclear Genes.Genetics 121:613-627.Brunsfeld, S.J., D.E Soltis, and P.S. Soltis. 1992. Evolutionary Patterns and Processes inSalix Sect. Longifoliae: Evidence from Chloroplast DNA. Syst. Bot.17:239-256.Cann, R.L. 1889. Cytoplasmic Inheritance. In Yearbook of Science & Technology.McGraw-Hill. pp. 86-89.Conkle, M.T. 1981. Isozyme Variation and Linkage in Six Conifer Species. In Conkle,M.T. (Ed.), Isozymes of North American Forest Trees and Forest Insects.  USDAGen. Tech. Rep. PSW-48. pp. 11-17.Dancik, B.P., and F.C. Yeh. 1983. Allozyme Variability and Evolutions of LodgepolePine (Pinus contorta var. latifolia) and Jack Pine (Pinus banksiana) in Alberta.Can. J. Genet. Cytol. 25:57-64.DeVerno, L.L., J.R. Byrne, J.A. Pitel, and W.M. Cheliak. 1989. Constructing ConiferGenomic Libraries: A Basic Guide. Information Report PI-X-88. PetawawaNational Forestry Institute, Forestry Canada.Doyle, J.J., J. Doyle, and A.H.D. Brown. 1989. The Limit of Chloroplast DNA inPhylogeny Reconstruction: Polymorphism and Phylogeny in the B Genome ofGlycine. Am. J. Bot. 76:239 (suppl.)Doyle, J.J., J.L. Doyle, and A.H.D. Brown. 1990. A Chloroplast DNA Phylogeny of theWild Perennial Relatives of Soybean (Glycinn Subgenus Glycine: Congruencewith Morphological and Crossing Groups. Evolution 44:371-389.El-Kassaby, Y.A., and O. Sziklai. 1982. Genetic Variation of Allozyme and QuantitativeTraits in a Selected Douglas-Fir [Pseudotsuga menziesii var. menziesii (Mirb.)Franco] Population. Forest Ecol. Manage. 4:115-126.Feinberg, A.P., and B. Vogelstein. 1983. A Technique for Radiolabeling DNARestriction Endonulclease Fragments to High Specific Activity. Anal. Biochem.132:6-13.77Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Hamrick, J.L., J.B. Milton, and Y.B. Linhart. 1981. Level of Genetic variation in ForestTrees: Influence of Life History Characteristics. In Conkle, M.T. (Ed.), Isozymesof North American Forest Trees and Forest Insects. USDA Gen. Tech. Rep.PSW-48. pp. 35-41.Hamrick, J.L., M.J.W. Godt, and S.L. Sherman-Broyles.Levels of Genetic Diversity in Woody Plant Species.Hartl, D.L. 1980. Principles of Population Genetics. Sunderland, Mass.1992. Factors InfluencingNew Forest 6:95-124.Sinauer Associates, Inc.Hillis, D.M. 1987. Molecular Versus Morphological Approaches to Systematics. Ann. Rev. Ecol. Syst. 18:23-43.Hong, Y.-P. 1991. Chloroplast DNA Variability and Phylogeny in the California ClosedCone Pines. Ph.D. Thesis. Oregon State University.Issac, L.A. 1930. Seed Flight in the Douglas-fir Region J. Forestry 28:492-499.Johansson, J.T., and R.H. Jansen. 1991. Chloroplast DNA Variation among Five Speciesof Ranunculaceae: Structure, SequenceDivergence, and PhulogeneticRelationship. Pl. Syst. Evol. 178:9-25.Kemble, R.J., R.J. Mans, S. Gabay-Laughnan, and J.R. Laughnan. 1983. SequencesHomologous to Episomal Mitochondrial DNAs in the Maize Nuclear Genome.Nature 304:744-747.Li, W-H, and D. Graur. 1991. Fundamentalsof Molecular Evolution. Sinauer Associates,Inc. Publishers. Sunderland, Massachusetts.Lidholm, J., and P. Gustafsson. 1991. The Chloroplast Genome of the GymnospermPinus contorta: A Physical Map and a Complete Collection of OverlappingClones. Curr. Genet. 20:161-166.McDonald, J.A., and T.J. Mabry. 1992. Phylogeny Systematics of New World Ipomoea(Convolvulaceae) Based on Chloroplast DNA Restriction Site Variation. Pl. Syst. Evol. 180:243;259.Millar, C.I., and W.J. Libby. 1991. Strategies for Conserving Clinal, Ecotypic, andDisjunct Population Diversity in Widespread Species. In Falk, D.A., and K.E.Holsinger (Eds.), Genetics and Conservation of Rare Plants. Oxford UniversityPress. New York. pp. 149-170.Milligan, B.G. 1991. Chloroplast DNA Diversity Within and Among Populations ofTrifolium pratense. Curr. Genet. 19:411-416.Molenaar, A.J., and R.J. Wilkins. 1991. A Simple and Convenient Way of BlottingNucleic Acids. BioTechniques 10:146-147.78Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Moore, W.S., J.H. Graham, and J.T. Price. 1991. Mitochondrial DNA Variation in theNorthern Flicker (Colaptes auratus, Ayes.). Mol. Biol. Evol. 8:327-334.Moran, G.F., and W.T. Adams. 1989. Microgeographical Patterns of AllozymeDifferentiation in Douglas-fir from Southwest Oregon. Forest Science 35:2-15.National Council on Gene Resources. 1982. Douglas-fir Genetic Resources. AnAssessment and Plan for California. California Gene Resources Program,National Council on Gene Resources, Berkeley, California.Neale, D.B., N.C. Wheeler, and R.W. Allard. 1986. Paternal Inheritance of ChloroplastDNA in Douglas-fir. Can. J. For. Res. 16:1152-1154.Neale, D.B., and R.R. Sederoff. 1989. Paternal Inheritance of Chloroplast DNA andMaternal Inheritance of Mitochondrial DNA in Loblolly Pine. Theor. Appl. Genet. 77:212-216.Nei, M., and R.K. Chesser. 1983. Estimation of Fixation Indices and Gene Diversities.Ann. Hum. Genet. 47:253-259.Nei, M., and J.C. Miller, 1990. A Simple Method for Estimating Average Number ofNucleotide Substitutions within and between Populations from Restriction Data.Genetics 125:873-879.Olmstead, R.G., R.K. Jansen, H.J. Micheals, S.R. Downie, and J.D. Palmer. 1989.Chloroplast DNA Phylogenetic Studies in th Asteridae. In Kawano, S. (Ed),Biological Approaches and Evolutionary Trends in Plants Academic Press,London. pp. 119-134.Palmer, J.D. 1987. Chloroplast DNA Evolution and Biosystematic Used of ChloroplastDNA Variation. Am. Natur. 130, S6 - S29.Saitou, N., and M. Nei. 1987. The Neighbor-Joining Method: A New Method forReconstructing Phylogenetic Trees. Mol. Biol. Evol. 4:406-425.Schaal, B.A., S.L. O'Kane, Jr., and S.H. Rogstad. 1991. DNA Variation in PlantPopulations. TREE. 6:329-333.Soltis, D.E., P.S. Soltis, and B.D. Ness. 1989. Chloroplast DNA Variation and MultipleOrigins of Autopolyploidy in Heuchera micrantha (Saxifragaceae). Evolution43:650-656.Strauss, S.H., D.B. Neale, and D.B. Wagner. 1989. Genetics of the Chloroplast inConifers Biotechnology Research Reveals Some Surprises. J. Forestry 87:11-17.Strauss, S.H., and A.H. Doerksen. 1990. Restriction Fragment Analysis of PinePhylogeny. Evolution 44:1081-1096.79Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Strauss, S.H., and A.H. Doerksen, and J.R. Byrne. 1990. Evolutionary Relationships ofDouglas-fir and Its Relatives (Genus Pseudotsuga) from DNA RestrictionFragment Analysis. Can. J. Bot. 68:1502-1510.Sytsma, K.J. 1990. DNA and Morphology: Inference of Plant Phylogeny. TREE.5:104-110Szmidt, A.E. 1991. Phylogenetic and Applied Studies on Chloroplast Genome in ForestConifers. In Fineschi, S., M.E. Malvolti, F. Cannata, and H.H. Hattemer. (Eds.),Biochemical Markers in The Population Genetics of Forest Trees. SPB AcademicPublishing by. pp. 185-196.Szmidt, A.E., A. Sigurgeirsson, X.-R. Wang, J.-E. Hallgren, and D. Lindgren. 1988.Genetic Relationships among Pinus species Based on Chloroplast DNAPolymorphism. In Hallgren, J.-E. (Ed.), Proc. Frans Kempe Symp.. MolecularGenetic of Forest Trees. limed, Sweden, pp. 33-47.Wagner, D.B., D.R. Govindaraju, C.W. Yeatman, and J.A. Pitel. 1989. PaternalChloroplast DNA Inheritance in a Diallel Cross of Jack Pine (Pinus banksianaLamb.). J. Hered 80:483-485.Wagner, D.B., G.R. Furnier, M.A. Saghai-Maroof, S.M. Williams, B.P. Dancik, and R.W.Allard. 1987. Chloroplast DNA Polymorphisms in Lodgepole Pine and JackPines and Their Hybrids. Proc. Natl. Acad. Sci USA. 84:2097-2100.Wagner, D.B., W.L. Nance, C.D. Nelson, T. Li, R.N. Patel, and D.R. Govindaraju. 1992.Taxonomic Patterns and Inheritance of Chloroplast DNA Variation in a Survey ofPinus echinata, Pinus elliottii, Pinus palustris and Pinus taeda. Can. J. For. Res.22:683-689.Waters, E.R., and B.A. Schall. 1991. No Variation in Detected in the ChloroplastGenome of Pinus torreyana. Can. J. For. Res. 21:1832-1835.Weir, B.S. 1990. Intraspecific Differentiation. In. Hillis, D.M., and C. Moritz, MolecularSystematics. Sinauer Association, Inc Publisher. Sunderland. pp. 373-410.Wendel, J., and V.A. Albert. 1992. Phylogenetics of the Cotton Genus (Gossypium):Character-State Weighted Parsimony Analysis of Chloroplast DNA RestrictionSite Data and Its Systematic and Biogeographic Implications. Syst. Bot. 17: 115-143.Whitkus, R. 1988. Modified Version of GENESTAT: A Program for Computing GeneticStatistics from Allele Frequency Data. Plant Genet. Newslet. 4:10Wolfe, K.H., W.-H. Li, and P.M. Sharp. 1987. Rates of Nucleotide Substitution VaryGreatly among Plant Mitochondrial, Chloroplast, and Nuclear DNAs. Proc. Natl. Acad. Sci. USA. 84:9054-9058.Woodruff, D.S., and G.A.E. Gall. 1992. Genetic and Conservation. Agric. Ecosystemsand Environ. 42:53-73.80Chapter 3 Chloroplast DNA Variation of Douglas-fir in B.C.Yeh, F.Ch.-H., and D. O'Molley. 1980. Enzyme Variations in Natural Populations ofDouglas-Fir, Pseudotsuga menziesii (Mirb.) Franco, from British Columbia. 1.Genetic Variation Patterns in Coastal Populations. Silvae Genet. 29:83-92.81CHAPTER 4MITOCHONDRIAL DNA VARIATION OF DOUGLAS-FIR IN B.C.4.1 INTRODUCTIONMitochondrial genomes in plants are small in size (200-2,500 Kb) (Newton, 1988)relative to the nuclear genome in conifers (10 7 Kb) (Miksche, 1971), but larger thanmitochondrial genomes from animal (15-18 Kb) and fungi (17-78 Kb) (Newton, 1988)and possess a partially independent genetic system (Verma, 1990). In addition, themitochondrial genome contains nucleotide sequences that code for some important traits,such as cytoplasmic male sterility (CMS), apocytochrome b and cytochrome oxidase(reviewed in Levings III and Brown, 1989), and mtDNA is highly conserved in primarysequences. Since mtDNA analyses are useful in assessing patterns of relatedness overrelatively recent time scales, present genetic structure can often provide insight into theecological and demographic influences of the past. Therefore, it has attractedconsiderable attention and may be a useful molecule for resolving forces (gene flow, drift,and selection) and historical events that have determined genetic structure in forest treespecies.A large number of genetic diversities based on mtDNA have been studied inhumans and other animals and genetic diversity estimated based on mtDNA was muchhigher than that with allozymes (Crease et al., 1990). Soodyal and Jenkins (1992) foundvery high degree of mtDNA diversity within human populations (0.768-0.829). Inanimals, relatively high levels of mtDNA intraspecific variation were found in rock hyrax(Prinsloo and Robinson, 1992). But low levels of mtDNA intraspecific variation havebeen reported in many crop species, such as soybean (Sisson et al., 1978), barley(Holwerda et al., 1986), tomato (McClean and Hanson, 1986), oat (Rines et al., 1988),pearl millet (Chowdhury and Smith, 1988), and Malawian bean (Khairallah et al., 1990).82Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.However, less is known about the evolutionary history of plant mitochondrial genomesthan is known of animal mitochondrial genomes (Brown, 1983; Terachi et al., 1990) andplant chloroplast genomes (Palmer, 1987). In addition, the use of mtDNA to investigatemolecular evolution and determine phylogenetic relationships are more commonlyconducted in animals, such as fish (Smouse et al., 1991), bat (van Den Bussche, 1992)and human populations (Horai, 1991). In plants, phylogenetic relationships constructedusing mtDNA are not well documented. Only a few plant species have been studied, suchas Triticum aestivum (bread wheat) (Graur et al., 1989), and Suillus (mushroom) (Brunsand Palmer, 1989). In conifers, Sutton et al., (1991) used mtDNA gene probes todistinguish the white (Picea glauca), Engelmann (P. engelmannii) and Sitka spruce (P.sitchensis) populations that occur in B.C. However, no work on mtDNA diversity withinspecies in Douglas-fir has been reported. Only between species comparisons in thegenera Pseudotsuga (Strauss et a/.,1990) and Pinus (Strauss and Doerksen, 1990) havebeen assessed using mtDNA markers. Information on mtDNA diversity may contributeto the understanding of population genetics and phylogenetic relationships in Douglas-fir.The purpose of this study is to utilize RFLP analysis of mitochondrial DNA toassess genetic variation of Douglas-fir populations from three geographic regions in B.C.,including coastal, transitional, and interior.83Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.4.2 MATERIALS AND METHODS4.2.1 Plant material.Eighteen Douglas-fir populations from three geographic regions (coastal,transition, and interior) in B.C., were selected for this study (Figure 3.1 in Chapter 3).Four individuals were sampled from each population. Coastal region populations werelocated on Vancouver Island and were designated based on seed stratification zones (A,B, and C). In each coastal zone, 2 populations were chosen, which were separated byelevation. Transition populations were located along the introgression zone of coastaland interior Douglas-fir. Interior populations were chosen from seed orchard planningzones for the interior region of B.C. (Table 3.1 in Chapter 3). In addition to theseDouglas-fir populations, eleven individuals of coastal Douglas-fir from the state ofWashington were included and used for phylogenetic analysis as an outgroup to root thephylogenetic tree. Young needle samples of coastal and transitional Douglas-fir werecollected from the clone bank at Cowichan Lake Research Station, and interior Douglas-fir samples were collected from Kalamalka Research Station & Seed Orchard, Vernon,B.0 (British Columbia Forest Service).4.2.2 Isolation, restriction, electrophoresis and transfer of genomic DNA.Total genomic DNA was prepared from young needles using a modified CTABmethod (DeVerno et al., 1989; Wagner et al., 1987). Ten restriction enzymes (six-basecutter), including Barn HI, Dra I, Eco RI, Eco RV, Hind III, Pst I, Pvu II, Sst I, Xba I, andXho I, were used to digest genomic DNA. Restriction endonuclease digestion was carriedout under the conditions prescribed by the supplier. Digested DNAs were separated on0.8% agarose gel and transferred from the gel to a Hybond-N± membrane (Amersham) in84Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.0.4M sodium hydroxide (NaOH) according to the alkaline capillary transfer procedure(Molenaar and Wilkins, 1991). Details of methods are provided in Chapter 4.4.2.3 MtDNA probes and hybridization.Three different cloned mitochondrial genes, including Cytochrome oxidasesubunit I (CoxI), Cytochrome oxidase subunit II (CoxII ), and ATPasea, were used inRFLP analysis. CoxI was a polymerase chain reaction (PCR) product amplified frompartially purified mitochondrial DNA of Douglas-fir using two specific primers toconserved regions: primer 1. TTA TTA TCA CTT CCG GTA CT, and primer 2. AGCATC TGG ATA ATC TGG (Glaubitz and Carlson, 1992). ATPasea and CoxII cloneswere mitochondrial genes of white spruce (Picea glauca) (Sutton et al., 1991) providedby the Forest Biotechnology Centre, British Columbia Research Corporation. ATPaseaand CoxII clones were 3.4 kb and 5.5 kb inserts in 3.96 Kb EMBL vectors, respectively.MtDNA clones were used as hybridization probes for the same blots used for cpDNAhybridization in Chapter 4.Five restriction enzymes, including BamHI, EcoRI, HindIII, SstI and XbaI werechosen and combined with 3 mitochondrial gene probes, CoxI, CoxII, and ATPasea.Twelve enzyme/probe combinations were used to determine mtDNA diversity (Table4.1).4.2.4 Data analysisTwo approaches to formatting the data were evaluated for determining mtDNAvariation, allelic frequencies and haplotype frequencies. For allelic frequency estimates,each restriction enzyme and probe combination was designated as a locus and theobserved restriction fragment pattern from each combination was designated as an allele.85Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.For haplotype frequency estimates, the restriction fragment patterns observed from allenzyme and probe combinations for each individual were pooled to identify haplotypes.Genetic diversity and degree of population differentiation estimates for unbiasedsample size followed Nei and Chesser (1983), using the computer package GENESTAT-PC version 2.1 (Whitkus, 1988).4.3 RESULTS AND DISCUSSION4.3.1 Mitochondrial DNA variationTwenty-eight different restriction fragments were observed from 12 enzyme andprobe combinations and twenty-five polymorphisms were observed (Table 4.1). Novariation was observed in the coastal and transitional B.C. and Washington populationsbased on CoxI and CoxII probes, whereas four polymorphisms were identified by ATPasea probe from BamHI and HindIII digestions. Polymorphisms from restriction sitechanges both within and among populations were observed in interior populations withall five enzymes and three mitochondrial gene probes (10 for CoxI, 8 for CoxII, 7 forATPasea). Even though only three mitochondrial gene probes were used in the study, thelevel of RFLPs observed in mtDNA of Douglas-fir was relatively higher than that incpDNA from the same samples (Chapter 4). The result indicates a higher divergence ofmtDNA genotypes than cpDNA genotypes in Douglas-fir. Terachi et al., (1990) alsoreported that mtDNA genomes in wheat show larger evolutionary divergence, based onrestriction site changes, than do cpDNA genomes from the same strains.86Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Table 4.1 Polymorphisms observed in mtDNA of Douglas-fir after digesting with 5different enzymes and hybridizing with 3 different probes(* 1- coxI,2-coxII, 3- ATPasea, A, B, C, D and E - indicate band size ineach enzyme and probe combination)Enzyme Sizeand (Kb)ProbePopulations (# individuals)Coastal Transition InteriorAl A2 B1 B2 Cl C2 T1 T2 T3 T4 T5 T6 CP CT SA WK MR MIB a mHI 11*^A 16.92B 13.992^A 12.06B^9.383^A 11.98B^6.85D^4.16-44-31--44-31-444444-444-4431444-44444-444-4444-44-4-311313-132231-44-4-4--4-4-13222113113-31EcoRI21 A 11.04 4 4 4 4 4 4 4 4 4 4 4 4 1 3 4 - 3 1B^6.15 - - - - 3 1 - 4 1 32 A 12.06 4 4 4 4 4 4 4 4 4 4 4 4 1 3 4 - 3 1B^9.38 - - - - 3 1 - 4 1 3Hind1H31 A^2.07 - - - 3 1 - 4 1 3B^1.72 4 4 4 4 4 4 4 4 4 4 4 4 1 3 4 - 3 12 A^8.45 4 4 4 4 4 4 4 4 4 4 4 4 4 3 4 - 2 1B^6.64 - - - - 1 - 4 2 33 A 13.68 - - - - 3 1 - 4 1 3B^8.45 - 3 1 4 1 3C^6.81 3 3 4 4 4 2 4 4 3 4 3 3 1 2 1 - 2 -E^5.29 - - 1 1 1 1 - 1 3 1 1SstI1 A 14.05 4 4 4 4 4 4 4 4 4 4 4 4 1 3 4 - 3 12 B 10.39 - - - - - 3 1 - 4 1 3XbaI1 A 23.86 4 4 4 4 4 4 4 4 4 4 4 4 1 3 4 3 1B 22.89 - - - - - 3 1 - 4 1 32 A 15.96 4 4 4 4 4 4 4 4 4 4 4 4 4 3 4 3 1B 12.31 - - - - - 1 - 4 1 31 site C in ATPasea probe (3) is in common in all populations.2 no variation in ATPasea probe (3).3 site D in ATPasea probe (3) is in common in all populations.87Chapter 4 Mitochondrial DNA variation of Douglas-fir in B. C.Twelve loci with 1-3 alleles each were observed (Table 4.2) and used in allelicfrequency estimates. Interior populations showed variation in almost all loci, whereas,only one and two polymorphic loci were observed in transitional and coastal populations,respectively. Nine haplotypes were characterized over all individuals in all populations(Tables 4.3 and 4.4). Haplotype I was most common in coastal (83.3%) and transitional(87.5%) populations. Of these two geographic regions (coastal and transitional), only 3different haplotypes were observed. However, 8 differing haplotypes were identified ininterior populations out of a total 9 haplotypes observed over all geographic regions.88Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Table 4.2 Haplotype frequency of coastal Douglas-fir populations based onmtDNA probes (* 1- coil, 2-coxll, 3- ATPasea)Locus^AllelePopulationsCoastal Transition InteriorAl A2 B1 B2 Cl C2 Tl T2 T3 T4 T5 T6 CP CT SA WK MR MIBamHI1*^122^123^1231.001.000.750.251.001.000.750.251.001.001.001.001.001.001.001.001.00-1.001.000.750.251.001.001.00-1.001.001.001.001.001.001.001.001.001.001.001.001.00-1.00-1.00-0.250.750.250.750.250.750.750.250.500.500.750.251.001.001.00-1.001.001.000.750.250.500.500.500.250.250.250.750.250.750.000.750.25EcoRI1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.25 0.75 1.00 - 0.75 0.252 - - 0.75 0.25 - 1.00 0.25 0.752 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.25 0.75 1.00 - 0.75 0.252 - - 0.75 0.25 - 1.00 0.25 0.753 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00HindIll1^1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.25 0.75 1.00 - 0.75 0.25- 0.75 0.25 - 1.00 0.25 0.752 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.75 1.00 - 0.50 0.252 - - - - - 0.25 - 1.00 0.50 0.753 1 0.75 0.75 1.00 1.00 1.00 0.50 1.00 1.00 0.75 1.00 0.75 0.75 0.25 0.50 0.25 - 0.50 0.002 0.25 0.25 - 0.25 - - - - 0.75 0.25 0.75 1.00 0.25 0.753 - - 0.25 - - 0.25 0.25 0.25 - 0.25 - 0.25 0.25Ssa1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.25 0.75 1.00 - 0.75 0.252 - - - - - - - - - - - - 0.75 0.25 - 1.00 0.25 0.75XbaI1 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.25 0.75 1.00 - 0.75 0.252 - - - - - - - - - - - 0.75 0.25 - 1.00 0.25 0.752 1 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 - 0.25 1.00 - 0.75 0.252 - - - - - - - 1.00 0.75 - 1.00 0.25 0.7589Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Table 4.3 Haplotype designation for mtDNA analysis. (A, B, C, D, and E refer toDNA fragment size in Table 5.1 , + and - indicate presence and absence ofa fragment)HaplotypesBamHI EcoRI Hind III SstI XbaI1' 2 3 1 2 1 2 3 1 1 2AB AB ABD AB A B AB AB A BCE AB AB AB1 -^+ +^- +^-^- +^- +^- -^+ +^- -^-^+^- +^- +^- +^-2 -^+ +^- -^+^- +^- +^- -^+ +^- -^-^+^- +^- -^- +^-3 -^+ +^- +^-^- +^- +^- -^+ +^- -^-^-^+ +^- +^- +^-4 +^- -^+ +^-^- -^+ -^+ +^- -^+ +^+^-^- -^+ -^+ -^+5 -^+ +^- +^-^- +^- +^- -^+ -^+ -^-^+^- +^- +^- -^+6 +^- -^+ -^+^- +^- +^- -^+ +^- -^-^+^- +^- +^- +^-7 -^+ -^+ +^-^- -^+ -^+ +^- -^+ +^+^-^- -^+ -^+ -^+8 -^+ -^+ +^-^- +^- +^- -^+ + + -^-^+^- +^- +^- +^-9 -^+ +^- -^-^+ +^- +^- -^+ +^- -^-^-^+ +^- +^- +^-- coxl, 2-coxll, 3- ATPasea).90Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Table 4.4^Haplotypes observed from the RFLP analysis of mtDNA(as percent of total)Populations HaplotypesI II III IV V VI VII VIII IXCoastal Al 0.75 0.25Coastal A2 0.75 0.25Coastal B1 1.00Coastal B2 1.00Coastal Cl 1.00Coastal C2 0.50 0.25 0.25Transition 1 1.00Transition 2 1.00Transition 3 0.75 - 0.25Transition 4 1.00Transition 5 0.75 0.25Transition 6 0.75 0.25Interior CP - 0.75 0.25Interior CT 0.25 0.25 - 0.25 0.25Interior SA 0.25 0.75Interior WK - - 1.00 -Interior MR 0.25 - 0.25 - 0.25 0.25Interior MI - 0.75 - - 0.25Total 0.61 0.05 0.11 0.16 0.01 0.01 0.01 0.01 0.0391Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.4.3.2 Estimates of mtDNA diversityPopulations in the interior region showed the highest level of genetic diversityamong the three geographic regions (D ST 0.201) (Table 4.5) In addition, the proportion ofdiversity attributable to the differentiation of the regions (G ST) were 0.193, 0.240, and0.424 in coastal, transitional and interior regions, respectively. Interestingly, these resultsshow that genetic subdivision is more pronounced in the interior region than eithercoastal or transitional populations.Table 4.5 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimates based on allelic frequencies of mtDNALevel of Analysis Hs HT DST GSTTotal population 0.121 0.310 0.189 0.610Geographic regions 0.181 0.294 0.113 0.384Coastal 0.056 0.069 0.013 0.193Transition 0.034 0.044 0.010 0.240Interior 0.274 0.475 0.201 0.424Hs = gene diversity within populations; HT= total gene diversity; DST= gene diversity amongpopulations (DsT = HT - Hs), Gs1= proportion of total gene diversity resulting from geneticdifferentiation among populations (G sT = DsT/HT)•The results of diversity estimates based on haplotype frequency yielded the samepatterns--higher genetic diversity within populations than among populations--asestimates using allelic frequency (Table 4.6). However, haplotype frequency may be thebetter method to estimate genetic diversity, as discussed in Chapter 3.92Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Table 4.6 Hierarchical partitioning of genetic diversity (Nei and Chesser, 1983)estimates based on haplotype frequencies of mtDNALevel of Analysis Hs(±SE)HT(±SE)DST(±SE)GsT(±SE)Total population 0.325 0.590 0.264 0.448(0.130) (0.401) (0.111) (0.174)Geographic regions 0.422 0.591 0.169 0.285(0.246) (0.344) (0.090) (0.129)Coastal 0.262 0.294 0.032 0.108(0.189) (0.205) (0.024) (0.078)Transition 0.214 0.223 0.009 0.040(0.157) (0.161) (0.008) (0.043)Interior 0.500 0.743 0.243 0.327(0.333) (0.478) (0.167) (0.222)Hs = gene diversity within populations; HT= total gene diversity; DsT= gene diversity amongpopulations (DST = HT - Hs), GsT= proportion of total gene diversity resulting from geneticdifferentiation among populations (GST = DST/HT), SE estimated using Jackknifing overpopulation (Weir, 1990).MtDNA diversity within (Hs) populations based on haplotype frequencies was notsignificantly different among all three regions. Interior populations showed the highestmtDNA diversity both within (Hs) and among (D ST) populations. Because nucleotidesubstitution rate in mtDNA is thought to be relatively slow, genetic variation withinpopulations that are depleted by any evolutionary forces will recover slowly by mutationalone (Banks and Birky, 1985). However, mtDNA evolves rapidly in structure (Palmerand Herbon, 1988) and the rate of blockwise sequence rearrangement in mitochondrialgenomes is higher than in chloroplast genomes (Coulthart, 1990). The higher mtDNAvariation of interior populations over that observed for the coastal and transitional93Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.populations could be caused by higher rate of sequence rearrangement of mtDNA ininterior populations.Genetic diversity based on mitochondrial haplotype estimates in some other Pinusspecies showed higher levels of genetic diversity among populations than withinpopulations, and these resulted in very high levels of population subdivision (G sT)(Strauss et al., in press), for example Monterey pine (0.755), Knobcone pine (0.856), andBishop pine (0.958). The degree of population subdivision (G sT) of 0.327 in interiorpopulations of Douglas-fir found in this study is not as high as reported for those species,but was much higher than the level of GsT estimated from isozyme markers (0.043) (Yeh,1981; Li and Adams, 1989) and cpDNA haplotypes (Chapter 3).Since mtDNA is haploid, the effective population size is expected to be one-halfof that for nuclear genes. Therefore, diversity in mtDNA is susceptible to genetic driftand population bottlenecks (Wilson et al., 1985). CpDNA in Douglas-fir is paternallyinherited (Neale et al., 1986), whereas mtDNA is maternally inherited (Marshall andNeale, 1991). Paternal gametes effectively disperse twice, as both pollen and seed,whereas maternal gametes disperse only via seed. MtDNA thus acts as a seed-specificgenetic marker. The migration rate of mtDNA (1N) is also expected to be only half ofnuclear DNA (2N). The movement of genes in conifers is accomplished by thedispersion of wind-born pollen and seed. In Douglas-fir, most seed is dispersed within 60m of the source, with comparatively little or no distribution beyond 300 m (Isaac, 1930).This restricted level of migration together with maternal inheritance may result in higherpopulation subdivision estimates for mitochondrial genomes than for nuclear allozymes.In B.C., the distribution of interior Douglas-fir is very wide ranging from south tonorth along the Rocky Mountains. Physiologically, mitochondria are sources of energy,sites of ATP production, which are important in respiration. In addition, isolation and94Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.intracellular genetic drift are probably important roles in producing the current populationvariation in the interior. The higher mtDNA variation in interior Douglas-fir may be theresult of adaptation to a greater variety of specific habitats, including lower temperatureand lower precipitation, than in coastal and transitional regions.4.4 CONCLUSIONSPolymorphisms resulting from restriction site changes were observed both withinand among interior populations with all five enzymes and three mitochondrial geneprobes. However only one probe, ATPasea, detected polymorphism in coastal andtransitional populations. Although genetic diversity estimates for mtDNA withinpopulations were not significantly different among all three regions, level of mtDNAdiversity among populations in interior populations were significantly higher than foreither coastal or transition populations. The different patterns of diversity betweencpDNA and mtDNA in Douglas-fir among the three regions might be caused by thedifferences in mode of inheritance, rates of evolution and also the functions of genes inboth genomes. The expansion of populations followed by intracellular genetic drift mightbe a cause of mtDNA evolution. Because of rapid structural rearrangement in themtDNA genome and the wide distribution of populations in interior regions, populationsubdivision (GsT of 0.327) estimates were relatively high.95Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.4.5 LITERATURE CITEDBanks, J.A., and C.W. Birky. 1985. Chloroplast DNA Diversity is Low in Wild Plant,Lupinus texensis Proc. Natl. Acad. Sci. USA 82:6950-6954.Brown, W.M. 1983. Evolution of Animal Mitochondrial DNA. In Nei. M., and K.Koehn (Eds.), Evolution of Genes and Protein. Sinauer, Sunderland, USA. pp.62-88.Bruns, T., and J.D. Palmer. 1989. Evolution of Mushroom Mitochondrial DNA: Suillusand Related Genera. J. Mol. Evol. 28:349-362.Chowdhury, M.K.U., and R.L. Smith. 1988. Mitochondrial DNA Variation in PearlMillet and Related Species. Theor. Appl. Genet. 76:25-32.Coulthart, M. 1990. Pattern and Process in Plant Mitochondrial Genome Evolution. InDudley, E.C. (Ed.), The Unity of Evolutionary Biology. Proceedings of the FourthInternational Congress of Systematic and Evolutionary Biology Volume 2. pp.828-837.Crease, T.J., M. Lynch, and K. Spitze. 1990. Hierarchical Analysis of Population GeneticVariation in Mitochondrial and Nuclear genes of Daphnia pulex. Mol. Biol. Evol.7:444-458.DeVerno, L.L., J.R. Byrne, J.A. Pitel, and W.M. Cheliak. 1989. Constructing ConiferGenomic Libraries: A Basic Guide. Information Report PI-X-88. PetawawaNational Forestry Institute, Forestry Canada.Farris, J.S. 1972. Estimating Phylogenetic Trees from Distance Matrices. Am Natur.106:645-668.Glaubitz, J.C., and J.E. Carlson. 1992. RNA Editing in the Mitochondria of a Conifer.Cum Genet. 22:163-165.Graur, D., M. Bogher, and A. Breiman. 1989. Restriction Endonulclease Profiles ofMitochondrial DNA and the Origin of the B Genome of Bread Wheat, Tritiumaestivum. Heredity 62:335-342.Horai, S. 1991. Molecular Phylogeny and Evolution of Human Mitochondrial DNA. InKimura, M., and N. Takahata (Eds.), New Aspects of Genetics of MolecularEvolution. Tokyo: Japan Sci. Soc. Press. pp. 135-152.Holwerda, B.C., S. Jana, W.L. Crosby. 1986. Chloroplast and Mitochondrial DNAVariation in Hordeum vulgare and Hordeum spontaneum. Genetics 114:1271-1291.Issac, L.A. 1930. Seed Flight in the Douglas-fir Region J. Forestry 28:492-499.96Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Khairallah, M.M., M.W. Adams, and B.B. Sears. 1990. Mitochondrial DNAPolymorphisms of Malawian Bean Lines: Further evidence for Two Major GenePools. Theor. Appl. Genet. 80:753-761.Levings III, C.S.,and G.G. Brown. 1989. Molecular Biology of Plant Mitochondria. Cell56:171-179.Li, P. and W.T. Adams. 1989. Range-wide Patterns of Allozyme Variation in Douglas-fir(Pseudotsuga menziesii ). Can. J. For. Res. 19:149-161.Marshall, K.A., and D.B. Neale. 1991. The Inheritance of Mitochondrial DNA inDouglas-fir (Pseudotsuga menziesii (Mirb.) Franco). Can. J. For. Res. 22:73-75.McClean, P.E., and M.R. Hanson. 1986. Mitochondrial DNA Sequence Divergenceamong Lycopersicon and Related Solanum Species. Genetics 112:649-667.Miksche, J. 1971. Intraspecific Variation of DNA per Cell Between Picea sitchensisProvenances. Chromosoma (Berl) 32:343-352.Molenaar, A.J., and R.J. Wilkins. 1991. A Simple and Convenient Way of BlottingNucleic Acids. BioTechniques 10:146-147.Neale, D.B., N.C. Wheeler, and R.W. Allard. 1986. Paternal Inheritance of ChloroplastDNA in Douglas-fir. Can. J. For. Res. 16:1152-1154.Nei, M.. and R.K. Chesser. 1983. Estimation of Fixation Indices and Gene Diversities.Ann. Hum. Genet. 47:253-259.Newton, K. 1988. Plant Mitochondrial Genomes: Organization, Expression andVariation. Annu. Rev. Plant Physiol. Plant Mol. Biol. 39:503-532.Palmer, J.D. 1987. Chloroplast DNA Evolution and Biosystematic Used of ChloroplastDNA Variation. Am. Natur. 130: S6 - S29.Palmer, J.D., and L.A. Herbon. 1988. Plant Mitochondrial DNA Evolves Rapidly inStructure, but Slowly in Sequence. J. Mol. Evol. 28: 87-97.Prinsloo, P., and T.J. Robinson. 1992 Geographic Mitochondrial DNA Variation in theRock Hyrax, Procavia capensis. Mol. Biol. Evol. 9:447-456.Rines, H.W., B.G. Gengenbach, K.L. Boylan, K.K. Storey. 1988. Mitochondrial DNAVariation in Oat Cultivars and Species. Crop Sci. 28:171-176.Sisson, V.A., C.A. Brim, and C.S. Levings III. 1978. Characterization of CytoplasmicDensity in Soybeans by Restriction Endonuclease Activity. Crop Sci. 18:991-996.Smouse, P.E., T.E. Dowling, J.A. Tworek, W.R. Hoeh, and W.M. Brown. 1991. Effectsof Intraspecific Variation in Phylogenetic Inference: A Likelihood Analysis ofmtDNA Restriction site Data in Cyprinid Fishes. Syst. Zool. 40:393-409.97Chapter 4 Mitochondrial DNA variation of Douglas-fir in B.C.Soodyal, H., and T. Jenkins. 1992. Mitochondrial DNA Polymorphisms in KhoisanPopulations from Southern Africa. Ann. Hum. Genet. 56:315-324.Strauss, S.H., and A.H. Doerksen. 1990. Restriction Fragment Analysis of PinePhylogeny. Evolution 44:1081-1096.Strauss, S.H., and A.H. Doerksen, and J.R. Byrne. 1990. Evolutionary Relationships ofDouglas-fir and Its Relatives (Genus Pseudotsuga) from DNA RestrictionFragment Analysis. Can. J. Bot. 68:1502-1510.Strauss, S.H., Y-P, Hong, and V.D. Hipkin. High Level of Population Differentiation forCoxI- Associated Mitochondrial DNA Haplotypes in Pinus radiata, muricata, andattenuata. Theor. Appl. Genet. (in press).Sutton, B.C.S., D.J. Flanagan, J.R. Gawley, C.H. Newton, D.T. Lester, and Y.A. El-Kassaby. 1991 Inheritance of Chloroplast and Mitochondrial DNA in Picea andComposition of Hybrids from Introgression Zones. Theor. Appl. Genet. 82:242-248.Terachi, T., Y. Ogihara, and K. Tsunewaki. 1990. The Molecular Basis of GeneticDiversity Among Cytoplasms of Triticum and Aegilops. 7. RestrictionEndonuclease Analysis of Mitochondrial DNAs from Polyploid Wheats and TheirAncestral Species. Theor. Appl. Genet. 80:366-373.Van Den Bussche, R.A. 1992. Phylogenetic Analysis of Restriction Site Variation in theRibosomal DNA Complex of New World Leaf-Nosed Bat Genera. Syst. Zool.40:420-432.Verma, R.S. 1990. The Genome. Frontiers in Molecular and Cellular Biology. VCH.Wagner, D.B., G.R. Furnier, M.A. Saghai-Maroof, S.M. Williams, B.P. Dancik, and R.W.Allard. 1987. Chloroplast DNA Polymorphisms in Lodgepole Pine and JackPines and Their Hybrids. Proc. Natl. Acad. Sci USA. 84:2097-2100.Weir, B.S. 1990. Intraspecific Differentiation. In. Hillis, D.M., and C. Moritz, MolecularSystematics. Sinauer Association, Inc Publisher. Sunderland. pp. 373-410.Whitkus, R. 1988. Modified Version of GENESTAT: A Program for Computing GeneticStatistics from Allele Frequency Data. Plant Genet. Newslet. 4:10Wilson, A.C., R.L. Cann, S.M. Carr, M. George, U.B. Gyllensten, K.M. Helm-Bychowski, R.G. Higuchi, S.R. Palumbi, E.M. Prager, R.D. Sage, and M.Stoneking. 1985. Mitochondrial DNA and Two Perspectives on EvolutionaryGenetics. Biological Journal of the Linnean Society 26:375-400.Yeh, F.C. 1981. Analyses of Genetic Diversity in Some Species of Conifers. In Conkle,M.T. (Tech. Coord.), Proc. Symp. Isozymes of North America Forest Trees andForest Insects. pp. 48-52. USDA For. Serv. Gen. Tech. Rep. PSW-48.98CHAPTER 5GENETIC PARAMETERS OF 100 FAMILIES OF 17-YEAR OLDCOASTAL DOUGLAS-FIR PROGENIES5.1 INTRODUCTIONThe primary effort in forest tree improvement is to maximize the productivity ofland committed to plantations. Productivity generally refers to the yield of high qualityproducts from well adapted, faster growing, better form, pest resistant trees (Zobel, et al.,1972; Kriebel, 1988; Zobel and van Buijtenen, 1989). Most tree breeders are concernedwith improving several traits simultaneously. Conventional methods for treeimprovement and propagation in coastal Douglas-fir in B.C. began in 1957 withphenotypic selection in natural stands and use of these materials for seed production inclonal or seedling seed orchards (Orr-Ewing, 1969). To evaluate the genetic worth ofthese phenotypic selections, forest tree breeding programs usually rely on genetic orprogeny testing (Place, 1969; Bridgwater and Franklin, 1985). Such field progeny testsare used to rank parents and to choose candidates with sufficiently broad genetic basis toprevent inbreeding in advanced generation orchards (Zobel, et al., 1972). In general,progeny testing combined with selection provides the primary means of obtaining treeimprovement goals.Most traits of silvicultural importance are controlled by many genes with smalleffects, rather than one or two major genes (Olsen, 1988; Zobel and van Buijtenen, 1989).Phenotypic variation in such polygenic characters as stem height, diameter, and wooddensity provide the raw material for a forest tree breeding program. The choice ofbreeding strategies is determined by the relative proportions of additive and non additivegene actions in the source, or base, population available to the breeder. With anappropriate mating design, the genetic parameters are determined through testing99Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeniesprograms. By means of progeny testing, plus-trees that provide poor progenies can beidentified for removal from seed orchards, so that only the best genotypes will be kept asparents for producing genetically superior seeds.Time in the developmental phases of tree improvement is important. Therotations of forest tree crops and the period before most tree species reach reproductivematurity are quite long. The generation intervals in forest trees are usually measured indecades rather than in growing seasons. Douglas-fir, for example, begins to produce seedcones at about 20 to 25 years of age (Eis and Craigdallie, 1983); however, this periodcould be manipulated and substantially reduced. Not only time, but the cost of setting upa tree improvement program to produce large quantities of genetically improved seed is amajor factor. As a result, searching for ways to reduce the time required for one cycle ofselection and breeding is often attempted. Alternatively, indirect selection is possible forsome characteristics rather than selecting directly for the specific trait (Zobel and Talbert,1984). The development of age-age correlations for selecting at young ages would resultin a shorter generation interval (Lambeth, 1980). Moreover, greater genetic gain per unittime will be attained through such efforts which will help to maximize the treeimprovement effort.The objectives of the study were to estimate genetic statistics in the 17-year-oldcoastal Douglas-fir progeny test and they were as follow:1. to evaluate phenotypic performances of the plus-trees' progenies by comparingtheir growth rates (height, diameter at breast height (dbh), and volume), andwood density by means of pilodyne measurement to the growth of controlprogenies; individuals were evaluated for the purpose of continuousselections,2. to estimate the amount of genetic and phenotypic variations,100Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies3. to estimate heritabilities for different family types, and4. to estimate age-age correlations.5.2 MATERMLS AND METHODS5.2.1 BackgroundIn 1968, the four coastal forest companies, British Columbia Forest Products Ltd.(BCFP), Crown Zellerbach Canada Ltd. (CZ Canada Ltd.), Rayonier Canada (B.C.) Ltd.,and Tahsis Co. Ltd.(presently Canadian Pacific Forest Products Ltd.), initiated anintraspecific hybridization program among several selected Douglas-fir plus-trees. Anincomplete diallel mating design was used and a total of 244 crosses were completed bythe end of 1968 (Sziklai, 1971). Of these crosses, 55 were ultimately included in theprogeny test consisting of 54 full-sib families (Figure 5.1) and one polymix family. Inaddition to these full-sib families, 21 half-sib families from original ortets (i.e., wind-pollinated) and 15 half-sib families from grafted clones of plus-trees in the companies'clone banks were included. Therefore, a total of 60 parent trees (Appendix 1) wasincluded in the test. Moreover, 9 control families, 7 from randomly selected seed lots(ROM) and 2 from bare root planting stocks from each company, were added into theproject. The designation of family numbers is shown in appendix 2. Finally, the progenytest consisted of a total of 100 families (Sziklai, 1971). In 1969, seeds were sown at theDuncan nursery and the seedlings were transplanted into Jiffy pots in spring 1970. Threelocations on Vancouver island, Caycuse, Courtenay, and Gold River, were selected to betest sites ( Figure 5.2). Information on the three test sites is presented in Table 5.1. The(1+1) seedlings were outplanted at the test sites in the spring of 1971.101Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesPM28 35 36 45 60 63 70 92 118 133 134 157 175 177 213 220 33525 3148^14*4915^ 19^2133 *^20^22 *1 10 362^ 44505311*42^41^ 405434*^8*^9^12^1353437^16^ 2351^17 24262718^ 2529394328^30^32384635526745472835364555606162638287114118125145153160162167172175176181196215218223224232233281356Total 4 5 2 12Figure 5.1 Mating design for controlled pollination families of coastal Douglas-fir,including 33 maternal parents, 17 paternal parents, and 54 crosses.(P - Paternal tree, M - Maternal tree, Numbers in the left column indicatematernal tree codes, Numbers in the top row indicate paternal tree codes,Numbers in the table indicate family codes, *- Reciprocal cross. Paternal andmaternal trees and family codes are presented in appendix 2)102DISTRIBUTION OF DOUGLAS-FIR PLUS TREESChapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesFigure 5.2 Location of coastal Douglas-fir plus-trees (Heaman, 1967) andthree progeny test plantations. (@ Caycuse, 0 Courtenay,0 Gold River, • Location of one or more plus trees ).103Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesTable 5.1 Site information for three Coastal Douglas-fir progeny test plantationsLocations(Company)Elevation Latitude Longitude(m)^(°N)^(°W)Slope and aspectCaycuse(BC Forest Product Ltd)425 48° 48' 124° 33' Moderate slope with southaspectCourtenay 455 49° 41' 125° 10' Relatively gentle south facing(CZ Canada Ltd.)Gold River 335 49° 57' 126° 07'(Tahsis Co. Ltd.)Source: Sziklai, 1971.The progeny tests used a systematic single-tree plot design with 16 (Caycuse andGold River) to 18 (Courtenay) replications of each family and a 3.05 x 3.05 m. (10 X 10ft.) spacing (Figure 5.3). Surrounding the test sites, buffer strips of 20 feet wide wereplanted.Initially, a single row of 100 trees was planted. However, the shape of landallocated to the project forced the partition of each replication into two sections. Theheight growth of each plus-tree progeny was measured in nursery phase in 1969 and 1970(Sigurdson, 1971) and at the test sites in 1973, 1974, 1975 (Bartram, 1977) and 1977(Labrie, 1978).104Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesReplication 1^ Replication 2Block Row (family #) Block Row (family #)1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15^161 1 7 14 20 27 34 40 47 54 60 67 73 80 86 93 99 1 51 57 64 70 77 84 90 97 4 10 17 23 30 36 43^492 2 8 15 21 28 35 41 48 55 61 68 74 81 87 94 100 2 52 58 65 71 78 85 91 98 5 II 18 24 31 37 44^503 3 9 16 22 29 36 42 49 56 62 69 75 82 88 95 1 3 53 59 66 72 79 86 92 99 6 12 19 25 32 38 45^514 4 10 17 23 30 37 43 50 57 63 70 76 83 89 96 2 4 54 60 67 73 80 87 93 100 7 13 20 26 33 39 46^525 5 11 18 24 31 38 44 51 58 64 71 77 84 90 97 3 5 55 61 68 74 81 88 94 1 8 14 21 27 34 40 47^536 6 12 19 25 32 39 45 52 59 65 72 78 85 91 98 4 6 56 62 69 75 82 89 95 2 9 15 22 28 35 41 48^547 7 13 20 26 33 40 46 53 60 66 73 79 86 92 99 5 7 57 63 70 76 83 90 96 3 10 16 23 29 36 42 49^558 8 14 21 27 34 41 47 54 61 67 74 80 87 93 100 6 8 58 64 71 77 84 91 97 4 II 17 24 30 37 43 50^569 9 15 22 28 35 42 48 55 62 68 75 81 88 94 1 7 9 59 65 72 78 85 92 98 5 12 18 25 31 38 44 51^5710 10 16 23 29 36 43 49 56 63 69 76 82 89 95 2 8 10 60 66 73 79 86 93 99 6 13 19 26 32 39 45 52^5811 11 17 24 30 37 44 50 57 64 70 77 83 90 96 3 9 11 61 67 74 80 87 94 100 7 14 20 27 33 40 46 53^5912 12 18 25 31 38 45 51 58 65 71 78 84 91 97 4 10 12 62 68 75 81 88 95 1 8 15 21 28 34 41 47 54^6013 13 19 26 32 39 46 52 59 66 72 79 85 92 98 5 11 13 63 69 76 82 89 96 2 9 16 22 29 35 42 48 55^6114 14 20 27 33 40 47 53 60 67 73 80 86 93 99 6 12 14 64 70 77 83 90 97 3 10 17 23 30 36 43 49 56^6215 15 21 28 34 41 48 54 61 68 74 81 87 94 100 7 13 15 65 71 78 84 91 98 4 11 18 24 31 37 44 50 57^6316 16 22 29 35 42 49 55 62 69 75 82 88 95 1 8 14 16 66 72 79 85 92 99 5 12 19 25 32 38 45 51 58^6417 17 23 30 36 43 50 56 63 70 76 83 89 96 2 9 15 17 67 73 80 86 93 100 6 13 20 26 33 39 46 52 59^6518 18 24 31 37 44 51 57 64 71 77 84 90 97 3 10 16 18 68 74 81 87 94 1 7 14 21 27 34 40 47 53 60^6619 19 25 32 38 45 52 58 65 72 78 85 91 98 4 II 17 19 69 75 82 88 95 2 8 15 22 28 35 41 48 54 61^6720 20 26 33 39 46 53 59 66 73 79 86 92 99 5 12 18 20 70 76 83 89 96 3 9 16 23 29 36 42 49 55 62^6821 21 27 34 40 47 54 60 67 74 80 87 93 100 6 13 19 21 71 77 84 90 97 4 10 17 24 30 37 43 50 56 63^6922 22 28 35 41 48 55 61 68 75 81 88 94 1 7 14 20 22 72 78 85 91 98 5 11 18 25 31 38 44 51 57 64^7023 23 29 36 42 49 56 62 69 76 82 89 95 2 8 15 21 23 73 79 86 92 99 6 12 19 26 32 39 45 52 58 65^7124 24 30 37 43 50 57 63 70 77 83 90 96 3 9 16 22 24 74 80 87 93 100 7 13 20 27 33 40 46 53 59 66^7225 25 31 38 44 51 58 64 71 78 84 91 97 4 10 17 23 25 75 81 88 94 1 8 14 21 28 34 41 47 54 60 67^7326 26 32 39 45 52 59 65 72 79 85 92 98 5 11 18 24 26 76 82 89 95 2 9 15 22 29 35 42 48 55 61 68^7427 27 33 40 46 53 60 66 73 80 86 93 99 6 12 19 25 27 77 83 90 96 3 10 16 23 30 36 43 49 56 62 69^7528 28 34 41 47 54 61 67 74 81 87 94 100 7 13 20 26 28 78 84 91 97 4 11 17 24 31 37 44 50 57 63 70^7629 29 35 42 48 55 62 68 75 82 88 95 1 8 14 21 27 29 79 85 92 98 5 12 18 25 32 38 45 51 58 64 71^7730 30 36 43 49 56 63 69 76 83 89 96 2 9 15 22 28 30 80 86 93 99 6 13 19 26 33 39 46 52 59 65 72^7831 31 37 44 50 57 64 70 77 84 90 97 3 10 16 23 29 31 81 87 94 100 7 14 20 27 34 40 47 53 60 66 73^7932 32 38 45 51 58 65 71 78 85 91 98 4 11 17 24 30 32 82 88 95 1 8 15 21 28 35 41 48 54 61 67 74^8033 33 39 46 52 59 66 72 79 86 92 99 5 12 18 25 31 33 83 89 96 2 9 16 22 29 36 42 49 55 62 68 75^8134 34 40 47 53 60 67 73 80 87 93 100 6 13 19 26 32 34 84 90 97 3 10 17 23 30 37 43 50 56 63 69 76^8235 35 41 48 54 61 68 74 81 88 94 1 7 10 20 27 33 35 85 91 98 4 11 18 24 31 38 44 51 57 64 70 77^8336 36 42 49 55 62 69 75 82 89 95 2 8 15 21 28 34 36 86 92 99 5 12 19 25 32 39 45 52 58 65 71 78^8437 37 43 50 56 63 70 76 83 90 96 3 9 16 22 29 35 37 87 93 100 6 13 20 26 33 40 46 53 59 66 72 79^8538 38 44 51 57 64 71 77 84 91 97 4 10 17 23 30 36 38 88 94 1 7 14 21 27 34 41 47 54 60 67 73 80^8639 39 45 52 58 65 72 78 85 92 98 5 II 18 24 31 37 39 89 95 2 8 15 22 28 35 42 48 55 61 68 74 81^8740 40 46 53 59 66 73 79 86 93 99 6 12 19 25 32 38 40 90 96 3 9 16 23 29 36 43 49 56 62 69 75 82^8841 41 47 54 60 67 74 80 87 94 100 7 13 20 26 33 39 41 91 97 4 10 17 24 30 37 44 50 57 63 70 76 83^8942 42 48 55 61 68 75 81 88 95 1 8 14 21 27 34 40 42 92 98 5 II 18 25 31 38 45 51 58 64 71 77 84^9043 43 49 56 62 69 76 82 89 96 2 9 15 22 28 35 41 43 93 99 6 12 19 26 32 39 46 52 59 65 72 78 85^9144 44 50 57 63 70 77 83 90 97 3 10 16 23 29 36 42 44 94 100 7 13 20 27 33 40 47 53 60 66 73 79 86^9245 45 51 58 64 71 78 84 91 98 4 11 17 24 30 37 43 45 95 1 8 14 21 28 34 41 48 54 61 67 74 80 87^9346 46 52 59 65 72 79 85 92 99 5 12 18 25 31 38 44 46 96 2 9 15 22 29 35 42 49 55 62 68 75 81 88^9447 47 53 60 66 73 80 86 93 100 6 13 19 26 32 39 45 47 97 3 10 16 23 30 36 43 50 56 63 69 76 82 89^9548 48 54 61 67 74 81 87 94 1 7 14 20 27 33 40 46 48 98 4 11 17 24 31 37 44 51 57 64 70 77 83 90^9649 49 55 62 68 75 82 88 95 2 8 15 21 28 34 41 47 49 99 5 12 18 25 32 38 45 52 58 65 71 78 84 91^9750 50 56 63 69 76 83 89 96 3 9 16 22 29 35 42 48 50 100 6 13 19 26 33 39 46 53 59 66 72 79 85 92^98Figure 5.3 Systematic single-tree plot planting design of coastal Douglas-firprogeny test [16 rows (1,600 trees) for Caycuse and Gold River and 18rows (1,800 trees) for Courtenay were designed]. Family numbers andparent trees designation are presented in appendix 2.105Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.2.2 Data collection and analysesHeight and diameter at breast height measurements on the progenies in the threetest sites were made in winter 1988-1989 (i.e., 17 years after sowing). Two readings ofpilodyne measurement were taken at breast height on the east and west sides of the treesand the mean of these measurements was then calculated. The height growth of 1970,1975 and 1977 were also used in statistical analysis. From the data collected, single treevolume was calculated using the method of Kovats (1977).5.2.2.1 Analysis of tree growth characteristics and condition of family typesThe 100 families were grouped into 4 family types, including full-sib, half-sibfrom grafted clones (C), half-sib from original ortets or plus-trees (P), and control. Thegrowth rates (height, dbh, and volume) of the trees and wood density (pilodyne) in thethree test sites were estimated from the 1989 data, using the following linear model.Yijkm^=^[1, + Fi + Si + FSii + Bka) + FBika) 4- Eijklwhere Yijkl^=^the mean growth of the tree lth in the ith familygrown in the kth block in the jth site (1=16-18),the overall mean,the effect of ith family type (i=4),the effect of jth site (j=3),interaction effect of the ith family type and the jthsite,the effect of block within site (k=2),interaction effect of the ith family type in the kthblock and the jth site,Eijkl^=^random tree error of lth tree in ijkth plot1 =Fi =Si =FSii =Bka ) =FBika) =106Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.2.2.2 Analysis of parent performanceHalf - sib familiesGrowth rate in each half-sib family in the progeny test was evaluated using thesame linear model in (5.2.2.1) and the structure of analysis of variance was determined asgiven in Table 5.2. An equal pollen contribution was assumed in both half-sib (C) andhalf-sib (P) crosses, but the average breeding value of pollen from natural stands shoulddiffer from that of pollen from trees in the clone bank. Therefore, the parents in eachgroup were compared to unknown pollen parents originating from a natural stand and aclone bank. Ranking of maternal performances was based on mean performance acrossall three test sites.Full - sib familiesDue to unbalance between male and female parents and scattered mating design(Figure 5.1), pairs of reciprocal crosses were pooled and treated as single families. Fifty-one families were analyzed using the same linear model and structure of analysis ofvarince (Table 5.2) as used for half-sib families. Variance components were used toestimate heritabilities as explained in 5.2.2.3.The evaluation of parents from full-sib crosses were separated into paternal andmaternal effects and the parents were ranked according to mean performance across allthree test sites.107Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesTable 5.2 Structure of the analysis of variance and variance componentsestimates for half-sib and full-sib familiesSource of variance^d.f.^Variance componentsSite (S)^(s-1) 6E  k a2 + k a2 + k a 2 + k 0 2 + k a 2E^1 FR(S)^2 FS^3 F^4 R(S)^5 SReplications withinsite (R/S)^s(r - 1)Family (F) (f-1)F x S^(f-1)(s - 1)F x R/S s(f - 1)(r - 1)Error (E)^N-srf6E^2 0„2 kE▪ erFR(S) '7' FS '8 F^96 R(S),2  k^,2 kE^+FR(S) '11" FS^12' F2_^2^21(13aFR(S) + k 14aFSer2 kE^FR(S)„„2Ef = number of familes, (half-sib (C) = 16; half-sib (P) = 21; full-sib = 51),s = number of sites (3),r = number of replications within site (2),N = total number of trees (half-sib (C) = 648; half-sib (P) = 720; full-sib = 2,168),r‘.2 ^2E FR(S) FS F "rr R(S) rr S are variance components due to random error, family x replication withinsite, family x site, family, replication within site, and site.Table 5.3 Coefficients of variance components for variance estimatesFamily k1 k2 k3 k4 k5 k6 kg k9 k10 k11 k12 k13 k14 k15Half-sib (C) 7.145 13.675 0.244 109.770 214.490 6.877 0.295 0.295 105.52 7.199 13.809 40.387 6.938 13.289 6.437Half-sib (P) 6.236 11.859 0.378 120.870 235.83 6.025 0.305 0.305 117.05 6.334 12.051 34.146 5.845 11.047 5.465Full-sib 7.993 15.208 0.184 371.880 718.710 7.547 0.289 0.289 349.050 7.586 14.425 42.426 7.398 14.000 6.681108Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.2. 2.3 Estimation of heritability.The estimates of narrow sense heritabilities (h 2) were calculated from componentsof variance for height, dbh, volume, and pilodyne measurement. Since the half-siboffspring had only their female parentage in common, the component of family varianceamong all parents in these experiments was used as an estimator of one forth of additivevariance. Therefore, an estimator of the additive genetic variance was 46 2F . In full-sibfamily, twice of maternal component of variance (2a 2F ) was used to estimate of additivegenetic variance.The formulas for estimating heritabilities and standard error (Namkoong, 1979)were as follow:Heritability estimated from half-sib analysish2 4a 2F2^2^2GF GFS GFR(S) + 2Heritability estimated from full -sib analysis2a2F2^2^2GF + GF + GF(S)+ 6E2=T,2^,T2^(x)2 h2 [  VCYF2^VCY2F "" FRO) V/7- V"' FS^E 2^2 )2^( GF2 ) 2^(GF2 GFR2 (S) + CT FS ± G Ewhere^h2S.E.Vxnarrow sense heritability,an approximation for estimated h2 assuming abalanced design,variance of the component of variance (Andersonand Bancroft, 1952),coefficient of h2 (half-sib = 4, and full-sib = 2).h2S.E.109Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.2.2.4 Genetic gainEstimates of genetic gain expected from selecting across all plantations werecalculated for best families and for best parents in full-sib families using the followingmodel (Falconer, 1981):G^=^japhh2where^G^=^the increment in growth rate, or genetic gain,i^=^standardized selection intensity,aPh^=^the phenotypic standard deviation.In this study the selection intensity of the best 50%, 40%, 30%, 20% and 10% ofparent trees of each family types were evaluated for the combined sites.5.2.2.5 Age-age correlationAge-age correlation was only conducted on height. The purpose of the analysiswas to investigate if juvenile-mature (age-age) correlation is a good predictor. If so, itmight be possible to estimate efficiency of early selection for all ages in genetic testswithout needing to measure the plants through the complete rotation. The calculation ofheight correlation was related to the natural logarithm of the younger age to the older age(LAR) (Lambeth, 1980). The parameters of the model were estimated by simple linearregression as follows:rage,age^=^B0 + B i (LAR)where^rage,age^=^correlation of height measured at two different ages,LAR^=^loge of age ratio (younger age/older age).110Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesThe correlations of height growth at different ages were calculated from meanheight of each family using the following formula:rxy I(x—R)(y — y)Vx(x - To 2 E(y - y) 2where^rxy^=^coefficient of correlation of family mean height atage x and age y,x^=^mean height of age x (younger age),y^=^mean height of age y (older age).111Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeny5.3 RESULTS AND DISCUSSION5.3.1 Survival and growth performanceSurvival of Douglas-fir progenies was estimated for all families and for differentfamily types across the three test sites, Caycuse, Courtenay, and Gold River. The averagesurvival percentage over the three plantations was 77%. Analysis of variance for survivalacross three test sites showed highly significance (Table 5.4). Duncan multiple range testfor survival percentages suggested that Courtenay showed significantly difference fromCaycuse and Gold River with highest survival (85%) whereas Caycuse and Gold Riverwere not significantly different from each other and had survivals of 74% and 71%,respectively .Table 5.4 Analysis of variance for survival of 100 families 17-year old coastalDouglas-fir progenies across three test sitesSource of^d.f.varianceMSSurvivalSite (S)^2^5139.703**Family (F) 99 370.687**F x S (Error)^198^136.929 Location Survival percentageCaycuse^74.38 bCourtenay 84.89 aGold River^71.19 bAverage 76.82** Significant at p < 0.01, Letters (a,b,c) indicate significant at P<0.05.112Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyBased on family types, full-sib, half-sib from clone banks (C), half-sib from plustrees (P), and control, the analysis of variance across three test sites indicatedsignificantly differences amomg family types (Table 5.5). The average survival of full-sib and half-sib (C) did not differ from each other but were significantly different fromhalf-sib (P) and control families. The families derived from the ortets or clone banks,full-sib and half-sib (C), always gave the highest survival percentage among all threeplantations.Table 5.5 Analysis of variance for survival of four different family types of 17-yearold coastal Douglas-fir progeniesSource of^d.f.varianceMSSurvivalSite (S)^2Family type (F)^3F x S (Error)^6Family type5139.703 **2822.105 *482.767Survival percentageFull-sib^80.09 aHalf-sib (C) 80.86 aHalf-sib (P)^68.00 bControl 70.58 b* * Significant at p < 0.01, * Significant at p < 0.05,Letters (a,b,c) indicate significant at P<0.05.Growth values, including height, DBH, volume, and pilodyne measurement forwood density of Douglas-fir progenies for all families, were measured over the three testsites. Analysis of variance for growth characteristics is shown in Table 5.6. Growthperformance of progenies per test site showed a significant difference (p<0.05) among thethree test sites (Table 5.7).113Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.6 Analysis of variance of growth variables of 100 families of coastalDouglas-fir progeniesSource ofvariationd.f. Mean squareHeight (m)Diameter (cm Volume (m3) Pilodyne (mm)Site (S) 2 2274.765 ** 15826.768 ** 2.187** 10941.312**Family (F) 99 12.735** 57.288** 0.008** 34.176**F x S (Error) 198 5.045 19.011 0.003 6.423Minimum 9.214 11.794 0.051 23.209(Fam./ Maternal) (86/96) (64/207) (64/207) (15/36)Maximum 11.729 18.221 0.126 18.659(Fam./ Maternal) (42/82) (55/36) (2/60) (36/55)Mean 10.642 15.053 0.0887 20.590** Significant at p < 0.01,NS Non-significant,The combined test sites analysis of variance for site and family showed highlysignificant differences for all traits but site x family variance was not significant. Meansfor height, diameter, and volume of all 100 families over three plantations were 10.642 m,15.053 cm, and 0.087 m3 , respectively. Mean pilodyne measurement for wood densitywas 20.59 mm (the higher the pilodyne value the lower the wood density).Among the three test sites, Gold River supported fastest growth in all three traits,whereas Courtenay showed the slowest growth rates (Table 5.7). However, wood densityperformed in just the opposite way. Many studies showing negative correlation of wooddensity and diameter and volume growth have been reported in coastal Douglas-fir (Kinget at., 1988b; El-Kassaby, and Park, 1990; PNWTIRC, 1990) and other conifers.Moreover, high wood specific gravity, low growth rate, and good form are usuallycorrelated with each other (van Buijtenen, 1984). In this study, the ranks of familiestended to be variable among plantations but the groups of higher ranked families andlower ranked families were consistent.114Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.7 Estimation of growth performance of 100 families of Coastal Douglas-firprogenies at three test sitesTest siteGrowth performanceHeight (m)(SD)Diameter (cm)(SD)Volume (m3)(SD)Pilodyne (mm)(SD)Caycuse 11.381 b 16.767 b 0.1087 b 19.859 b(2.005) (4.656) (0.0613) (2.521)Courtenay 9.308 c 11.561 c 0.0477 c 18.502 a(2.097) (3.289) (0.0306) (2.550)Gold River 11.660 a 17.945 a 0.1229 a 24.166 c(1.943) (4.427) (0.0625) (2.564)3 Test sites 10.642 15.053 0.0887 20.590(2.297) (4.996) (0.0617) (3.496)Different letters (a,b,c) in the same column indicate significantly at P<0.05.Height growths were observed and evaluated in 1973, 1974, 1975 (Bartram, 1977)and 1977 (Labrie, 1978). The comparison of height growth of younger ages and at age 17is given in Table 5.8. The Caycuse plantation was the fastest growing in 1973, 1974, and1975, while Courtenay was the slowest. However, the Gold River plantation became thefastest growing in the next two years, while Courtenay maintained the slowest growthrate. At age 17, the pattern of growth rate was the same as in 1977.Table 5.8 Comparison of height growth of coastal Douglas-fir progenies overthree test sites from 1973 to 1989Test site Height growth1973 1(cm) 1974 1 (cm) 1975 1(cm) 19772(cm) 1989 (cm)Caycuse 46.00 a 71.93 a 112.70 a 186.30 b 1138.13 bCourtenay 41.12 b 57.61 c 82.29 c 167.43 c 930.81 cGold River 46.00 a 68.59 b 92.20 b 200.53 a 1166.01 a3 Test sites 44.37 66.04 95.73 184.75 1064.20Different letters (a,b,c) in the same column indicates significant at p<0.05,1 Data from Bartram (1977), 2 Data from Labrie (1978).115Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeny5.3.2 Growth performance and condition of family typesAnalysis of variance for height, dbh, volume, and pilodyne measurement of full-sib, half-sib (C), half-sib (P) and control families of Douglas-fir is presented in Table 5.9.The result gives further insight into the magnitude of family type variation. Family typemean squares were significant at P=0.05 for height, dbh, and volume, and was significantat 1)=0.01 for wood density, indicating that much of the variation among full-sib and half-sib families was due to genetic causes.Table 5.9 Analysis of variance and growth characteristics of 4 family types ofcoastal Douglas-fir progeniesSource ofvarianced.f. MSHeight(m)Diameter(cm)Volume(m3)Pilodyne(mm)Site (S) 2 2274.765** 15826.767 ** 2.187 ** 10941.312 **Family type (F) 3 109.217* 342.952 * 0.050 * 149.928**F x S (Error) 6 14.261 38.108 0.008 10.143Family type Growth performanceFull-sib 10.802 a 15.305 a 0.0915 a 20.896 c(SD) (2.207) (4.910) (0.0621) (3.516)Half-sib (C) 10.954 a 15.764 a 0.0980 a 20.573 c(SD) (2.250) (5.069) (0.0650) (3.341)Half-sib (P) 10.164 b 14.160 b 0.0778 b 20.122 b(SD) (2.394) (4.969) (0.0575) (3.444)Control 10.045 b 13.914 b 0.0758 b 19.609 a(SD) (2.487) (5.059) (0.0571) (3.376)* 4, Significant at p < 0.01, * Significant at p < 0.05,Letters (a,b,c) in the same column indicate significant at P<0.05.116Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyThe comparison of four family types of 17-year old coastal Douglas-fir progeniesaccording to means of growth values, height, dbh, and volume, and pilodynemeasurement over three plantations is presented in the Table 5.9, and growth patterns ofthe four family types are presented in Figure 5.4. The Duncan Multiple Range test(demonstrated by letters a,b,c) shows significance at the 0.05 level of probability. Intheory, families with two plus-tree parents should rank first. In this experiment, twotypes of families from two plus-tree parents, including full-sibs and half-sibs from clonebanks (C) show better performances over half-sibs from plus trees (P) and controlfamilies. At this age, there is no difference in height growth of progenies from differenttypes of families; therefore, volume growth of the progenies is dependent on the growthof dbh. On the other hand, the growth rates of half-sib (C) families and full-sib familiesare not statistically different, but half-sib (C) families usually show a better growth rate.Hybrid vigor offers a possible explanation of the growth performance difference. Themeans of full-sib and half-sib (C) families have resulted from crossing among plus treesselected from different populations, whereas, the half-sib (P) families were derived fromthe selected maternal parents and surrounding trees in the same populations. From thisanalysis, it would appear that hybridization among plus trees in a clone bank or a seedorchard is the better method for breeding coastal Douglas-fir for maximizing growth rate.117Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeny2 42 01 612840Full—sib^Half—sib (C)^Half—sib (P)^ControlVolume (x .01) M height (m)^Diameter (cm)    Pilodyne (aim)Figure 5.4 Growth characteristics of four different family types of coastalDouglas-fir progenies.118Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyThe comparison of height growth of 4 different family types in 1989 with youngerages (Table 5.10) showed that half-sib (C) and full-sib families consistently performedbetter in height growth than half-sib (P) and control families in all ages, whereas half-sib(C) families showed better growth rate than full-sib families at younger ages. However,at age 6 and 17 there was no statistically significant difference between full-sib and half-sib (C) families.Table 5.10 Mean height of 4 family types in 1973, 1974, 1975, 1977 and 1989Family type Height growth1973 1 (cm) 1974 1 (cm) 1975 1 (cm) 19772(cm) 1989 (cm)Full-sib 48.16 b 70.39 b 101.30 b 190.4 a 1080.20 aHalf-sib (C) 52.58 a 76.46 a 109.40 a 190.0 a 1095.40 aHalf-sib (P) 41.34 c 61.04 c 88.14 c 161.6 b 1016.40 bControl 35.40 d 59.29 d 84.12 c 162.1 b 1004.50 bDifferent letters (a,b,c) in the same column indicates significance at p<0.05,I Data from Bartram (1977),2 Data from Labrie (1978).Analysis of variances for family types based on growth rates and pilodynemeasurement at each plantation and ranking of family types for each trait are given inTables 5.11-5.13. At the Caycuse plantation (Table 5.11), only diameter showedsignificantly difference among family types. Half-sibs (C) grew faster, whereas theother three family types had similar growth rates.119Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.11^Estimation of growth performance of 4 family types of coastal Douglas-fir progenies for Caycuse plantationSource of^d.f.varianceMSHeight(m)Diameter(cm)Volume(m3)Pilodyne(mm)Replication (R)^1 202.934** 370.617 ** 0.126 ** 219.172 **Family type (F)^3 17.235NS 105.016* 0.019NS 27.019NsF x R (Error)^3 4.139 11.577 0.004 13.216Family type Growth performanceFull-sib 11.36 a 16.67 b 0.107 a 20.06 a(SD) (2.02) (4.64) (0.061) (2.50)Half-sib (C) 11.80 a 17.84 a 0.123 a 19.84 a(SD) (1.86) (4.46) (0.062) (2.56)Half-sib (P) 11.12 a 16.24 b 0.103 a 19.46 a(SD) (1.86) (4.88) (0.063) (2.37)Control 11.22 a 16.37 b 0.101 a 19.24 a(SD) (1.80) (4.36) (0.054) (2.75)** Significant at p < 0.01, * Significant at p < 0.05, NS Non-significant,Letters (a,b,c) in the same column indicate significant at P<0.05.Analysis of variances for family types at Courtenay (Table 5.12) and Gold River(Table 5.13) showed the same pattern with significantly differences of family types for allgrowth variables except for pilodyne measurement. Duncan multiple range test (P=0.05)suggested that growth rates in full-sib and half-sibs (C) were not different from each otherand grew faster than half-sibs (P) and control families whereas half-sibs (P) and controlfamilies were not significantly different from each other. No differences in wood densitywas observed among these family types in both plantations.120Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.12 Estimation of growth performance of 4 family types of coastal Douglas-fir progenies for Courtenay plantationSource ofvarianced.f. MSHeight(m)Diameter(cm)Volume(m3)Pilodyne(mm)Replication (R) 1 224.010** 659.400 ** 0.073 ** 349.249 *Family type (F) 3 86.179* 160.394* 0.012* 53.663NsF x R (Error) 3 3.313 8.685 0.001 15.257Family type Growth performanceFull-sib 9.52 a 11.80 a 0.050 ab 18.68 a(SD) (2.01) (3.12) (0.030) (2.54)Half-sib (C) 9.60 a 12.10 a 0.052 a 18.63 a(SD) (1.99) (3.25) (0.031) (2.42)Half-sib (P) 8.82 b 10.89 b 0.042 be 18.20 a(SD) (2.21) (3.43) (0.030) (2.60)Control 8.59 b 10.66 b 0.040 c 17.91 a(SD) (2.22) (3.63) (0.031) (2.60)* * Significant at p < 0.01, * Significant at p < 0.05, NS Non-significant,Letters (a,b,c) in the same column indicate significant at P<0.05.121Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.13^Estimation of growth performance of 4 family types of coastal Douglas-fir progenies for Gold River plantationSource ofvarianced.f. MSHeight(m)Diameter(cm)Volume(m3)Pilodyne(mm)Replication (R) 1 70.666 ** 102.995NS 0.038 * 70.501NSFamily type (F) 3 34.324 * 153.759* 0.036* 98.790NsF x R (Error) 3 2.364 14.630 0.002 16.297Family type Growth performanceFull-sib 11.83 a 18.30 a 0.128 a 24.55 a(SD) (1.89) (4.30) (0.063) (2.41)Half-sib (C) 11.83 a 18.35 a 0.131 a 23.84 a(SD) (2.07) (4.79) (0.067) (2.58)Half-sib (P) 11.08 b 16.83 b 0.105 b 23.78 a(SD) (1.96) (4.15) (0.054) (2.67)Control 11.27 b 16.54 b 0.109 b 22.85 a(SD) (2.22) (4.59) (0.060) (2.81)*4, Significant at p < 0.01, * Significant at p < 0.05, Ns Non-significant,Letters (a,b,c) in the same column indicate significant at P<0.05.122Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyThe summary for family types ranking (Table 5.14) clearly showed that half-sib(C) usually ranked first, although it was not significantly different from full-sib, in everyplantation and for every trait except wood density. Half-sib (C) showed the fastest involume growth when averaged across three plantations and was significantly differentfrom the other family types. Growth rates of progenies at Courtenay and Gold Riverplantations displayed the same pattern. Full-sibs and half-sibs (C) always showed bettergrowth rate than half-sibs (P) and control families. Wood densities in all threeplantations showed negative correlation to growth rates and were highest for the controlfamilies. Interestingly, wood density in half-sibs (C) was higher than for full-sib familiesin all three plantations and higher than for the over all average (Table 5.13), althoughthey were not significantly different in all three plantations. The result suggests thatselection in half-sib (C) families may improve both growth rate and wood density andmay be a good practice for coastal Douglas-fir improvement program.123Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.14^Summary of family types ranking of 17-year old coastal Douglas-firand average growth characteristics for all three plantationsFamilytypesRank 3 sites average1989# of treesCaycuse Courtenay^Gold RiverHeight (m)Full-sib 2 a 2 a^1 a 2 10.802 a 2,168Half-sib (C) 1 a 1 a 2 a 1 10.954 a 648Half-sib (P) 4 a 3 b^4 b 3 10.146 b 720Control 3 a 4 b 3 b 4 10.045 b 321Diameter (cm)Full-sib 2 b 2 a^2 a 2 15.305 b 2,168Half-sib (C) 1 a 1 a 1 a 1 15.746 a 648Half-sib (P) 4 b 3 b^4 b 3 14.160 c 720Control 3 b 4 b 3 b 4 13.914 c 321Volume (m3)Full-sib 2 a 1 ab^2 a 2 0.0915 b 2,168Half-sib (C) 1 a 2 a 1 a 1 0.0980 a 648Half-sib (P) 3 a 4 be^4 b 3 0.0778 c 720Control 4 a 3 c 3 b 4 0.0758 c 321Pilodyne (mm)Full-sib 4 a 4 a^4 a 4 20.896 d 2,168Half-sib (C) 3 a 3 a 3 a 3 20.573 c 648Half-sib (P) 2 a 2 a^2 a 2 20.122 b 720Control 1 a 1 a I a 1 19.609 a 321Different letters (a,b,c) in the same column in each trait indicate significance at p<0.05.124Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeny5.3.3 Parent performance5.3.3.1 Half-sib (C) familiesAnalysis of variances for combined test sites of half-sib (C) families (Table 5.15)showed highly significance for both and site and family variances for all four traits. Theresults suggested that high variation in growth rate and wood density existed in thisfamily type.Table 5.15 Analysis of variance of growth variables of half-sib (C) familiesSource of^d.f.variationMean squareHeight^Diameter^Volume PilodyneSite (S) 2 376.151 ** 2773.589** 0.4324** 1573.735 **Family (F) 15 10.096** 68.284** 0.0086 ** 35.722 **F x S (Error) 30 4.676 27.494 0.0051 5.682** Significant at p < 0.01,Both genetic and phenotypic variances were estimated (Table 5.16) and used toestimate individual heritabilities for total height, diameter, volume and wood density.The heritability estimates range from 0.116 for volume to 0.494 for wood density.Diameter heritability (0.234) was highest in growth traits. The results indicate thatgrowth traits are under low to moderate genetic control and wood density is under highergenetic control. However, recurrent selection on the basis of family comparisons ispossible to improve these traits.125Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.16 Genetic and phenotypic variances, phenotypic means, andheritabilities of growth variables of half-sib (C) familiesComponents Height(m)Diameter(cm)Volume(m3)Pilodyne(mm)Family (c4) 0.133 0.998 0.000083 0.743F x S (6 2Fs ) 0.152 0.827 0.000186 -0.038*F x R/S (c:TR(s)) -0.146* 0.213 0.000006 0.156Error (6 2E ) 3.670 15.029 0.002587 5.113Phenotypic (a;) 3.955 17.067 0.002862 6.012h2 0.135 0.234 0.116 0.494(±SE) (0.509) (0.597) (0.665) (0.592)Minimum 9.332 11.794 0.0506 22.120(Fam./ Maternal) (64/207) (64/207) (64/207) (55/36)Maximum 11.462 18.221 0.1239 19.221(Fam./ Maternal) (57/110) (55/36) (55/36) (66/215)Mean 10.954 15.764 0.0980 20.572* negative variance components were considered as 0.0Mean height for half-sib (C) family was 10.954 m, with maternal tree #110 thebest (Table 5.16). Maternal tree #36 performed the best for diameter (18.221 cm) andvolume (0.1239 m 3), whereas the mean for diameter was 15.764 cm. and the mean forvolume was 0.098 m3 . Ranking of maternal parents for height, diameter, volume, andpilodyne measurement of half-sib (C) families are given in Appendices 3-6, respectively.According to Duncan multiple range test (P=0.05), the families were distributedinto 3, 4, 2 and 5 undifferentiated groups for height, diameter, volume and wood density,respectively. Diameter and pilodyne measurement characters displayed moreheterogeneity than height and volume. Thirteen families were ranked as the fastestgrowing in height (Appendix 3), while 11 families were ranked the best in diameter126Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenyperformance (Appendix 4). In volume (Appendix 5) and Pilodyne assessment (Appendix6), 9 families were ranked in first group for both traits. Those maternal parents suggestedfor selection are also listed in the appendices.5.3.3.2 Half-sib (P) familiesAnalysis of variance for combined test sites of half-sib (P) families (Table 5.17)showed the same pattern as for half-sib (C), highly significance for family variances forall traits.Table 5.17 Analysis of variance of growth variables of half-sib (P) familiesSource of^d.f.variationMean squareHeight^Diameter^Volume PilodyneSite (S) 2 456.303 ** 2797.664 ** 0.3306 ** 1931.097**Family (F) 20 9.744** 38.650** 0.0051 ** 16.991 **F x S (Error) 40 4.194 15.524 0.0027 6.262** Significant at p < 0.01,Both genetic and phenotypic variance estimates are given in Table 5.18.Individual heritability estimates range from 0.116 for volume to 0.189 for wood density.The heritabilities for growth traits in half-sib (P) families were comparable to half-sib (C)families. However, wood density heritability in half-sib (P) family was much less than inhalf-sib (C) family (h2=0.494). In addition, the average of growth rates in half-sib (P)families were less than in half-sib (C) families.127Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.18 Genetic and phenotypic variances, phenotypic means, andheritabilities of growth variables of half-sib (P) familiesComponents Height(m)Diameter(cm)Volume(m3)Pilodyne(mm)Family (6) 0.163 0.679 0.000069 0.314F x S (6 2Fs ) 0.037 -0.145* M.000010* 0.108F x R/S (cTFR2 (s)) -0.086* 0.156 0.000105 -0.201*Error (6) 4.290 16.218 0.002206 6.238Phenotypic (a;) 4.490 17.053 0.002380 6.660h2 0.145 0.159 0.116 0.189(±SE) (0.436) (0.435) (0.502) (0.441)Minimum 9.214 12.129 0.0566 21.456(Fam./ Maternal) (86/96) (86/96) (86/96) (87/114)Maximum 11.303 16.049 0.1064 18.867(Fam./ Maternal) (87/114) (93/223) (87/114) (81/49)Mean 10.146 14.160 0.0780 20.122* negative variance components were considered as 0.0Mean height for half-sib (P) family was 10.146 m (9.214-11.303) with maternaltree #114 at the highest rank (Table 5.17). For diameter and volume, maternal parents#223 and #114 performed the best, while parent #96 showed the lowest rank for all threetraits. Wood density is a trait that is usually negatively correlated to growth rate. Thus asexpected, parent #114 performed the best in height and volume but had the lowest rank inwood density. Ranking of parents for height, diameter, volume, and pilodynemeasurement for half-sib (P) are shown in Appendices 7-10, respectively. The Duncantest (P=0.05) rendered family distributions into 5, 6, 5 and 5 undifferentiated groups forheight, diameter, volume, and wood density, respectively. All traits were highlyheterogeneous. Twelve families consistently ranked the best in both height and diametergrowth, while the first 9 and 13 families were the best in volume and wood density.128Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeny5.3.3.3 Full-sib familiesMean square values and their significance levels are presented for combined sitesand genotypes analysis in Table 5.19 The variance in full-sib families used in estimatingadditive genetic variance was highly significant (P<0.01) for all traits.Table 5.19 Analysis of variance of growth variables for full-sib familiesSource ofvariationdf Mean squareHeight Diameter Volume PilodyneSite (S)Family (F)F x S (Error)2501001138.384**10.389**4.9628699.076 **51.869**17.8591.2471 **0.0077 **0.00316681.212**39.460**6.144** Significant at p < 0.01.Genetic and phenotypic variances were estimated for all characters and are shownin Table 5.20. The individual tree heritabilities range from 0.068 (height) to 0.256(pilodyne) (Table 5.20). The results suggest that these traits are under moderately lowgenetic control. King et al., (1988a) estimated heritabilities of total height, incrementheight, diameter, and volume of full-sib progeny tests of 12-year old coastal Douglas-firand found that individual tree heritabilities were between 0.08 and 0.16. El-Kassaby andPark, (1990) also found that diameter heritabilities in young full-sib coastal Douglas-firwere between 0.09-0.16. Although, a heritability estimate applies only to the experimentfrom which it was obtained, their estimates of family heritabilities were comparable tothis study. Similarly, heritability estimates for wood quality characteristics coincide withstudies (0.2 to 0.9) reported by Okwuagwu and Guries, (1981).129Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.20 Genetic and phenotypic variances, phenotypic means, and heritabilitiesof growth variables of full-sib familiesComponents Height (m) Diameter (cm) Volume (m3) Pilodyne (mm)Family (a;) 0.126 0.799 0.000108 0.785F x S (c7;s ) 0.131 0.159 0.000039 0.110F x R/S (a2F-it(s)) -0.043* 0.138 0.000018 -0.086*Error (aE) 3.443 14.614 0.002459 5.237Phenotypic (a,) 3.700 15.710 0.002624 6.132h2 0.068 0.102 0.082 0.256(±SE) (0.151) (0.143) (0.142) (0.171)* negative variance components were considered as 0.0Analysis of variances for both female and male parents showed highlysignificance in all traits (Table 5.21). Mean growth for height was 11 m., with aminimum of 9.80 m. in maternal tree #61 and 10 m in paternal trees #133, and maximumsof 11.7 m. in maternal tree #125 and in paternal tree #60. Mean diameter was 15 cm,with a minimum of 12 cm in maternal tree #175 and 13 cm in paternal tree #157 andmaximum diameters in paternal tree #36 and paternal tree #60 of 17 cm and 16 cm,respectively. The average volume growth was 0.1113 m3 with a minimum of 0.0651 m3and 0.672 m3 observed in maternal tree #25 and paternal tree #157, respectively. Parenttree #60 showed the best volume growth when used as either maternal (0.1113 m 3) orpaternal (0.1155 m3) parent. Wood density was usually negatively associated withgrowth rate. The better progenies derived from parents that produced fastest growingprogenies usually had the lowest wood density. In addition, maternal tree #36 that servedas polymix cross in half-sib (C) families ranked highest in diameter and volume and alsoyielded the biggest diameters among full-sib families.130Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.21 Analysis of variance of growth variable of maternal and paternal generalcombining abilities for full-sib familiesSource ofvariationdf Mean squareHeight (m) Diameter (cm) Volume (m3) Pilodyne (mm)Site (S) 2 1138.384 ** 8699.075** 1.2471** 6681.212 **Maternal (M) 32 13.914** 62.347** 0.0091** 48.996**Paternal (P) 16 18.815** 88.823** 0.0128** 76.407**M x S 1 64 4.788 13.079 0.0025 5.519P x S2 32 6.683 10.970 0.0022 7.039Growth performanceMaternal Paternal Maternal Paternal Maternal Paternal Maternal Paternal(it) (it) (#) (#) (#) (#) (#) (#)Minimum 9.804^10.007 12.803^13.107 0.0651^0.0672 22.314^21.893(61)^(133) (175)^(157) (25)^(157) (125)^(60)Maximum 11.728^11.729 17.353^16.319 0.1113^0.1155 18.839^18.659(125)^(60) (36)^(60) (60)^(60) (215)^(157)Mean 11.082 15.305 0.0915 20.896Significant at p < 0.01,1 Error term for maternal , 2 Error tern for paternal .The Duncan multiple range tests (P=0.05) for ranking of maternal and paternalparents for all four traits in full-sib families are demonstrated in Appendices 11-18. Atotal of 33 maternal parents and 17 paternal parents are identified as the best parents ineach trait. Twelve homogeneous groups of maternal parents were classified for height(Appendix 11) and pilodyne measurement (Appendix 14) and 11 homogeneous groupsfor diameter (Appendix 12) and volume (Appendix 13). In paternal parent ranking, 5groups were homogeneous in height (Appendix 15) and diameter (Appendix 16), while 7and 8 groups were undifferentiated in volume (Appendix 17) and Pilodyne measurement(Appendix 18). The results indicate that all of the traits in full-sib families expressed ashighly heterogeneous characters. In maternal parents, the first 15, 11, 21 and 12 families131Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenyperformed best for height, diameter, volume and wood density, respectively. Forselection of paternal parents, the first 13, 12, 10 and 6 paternal parents should beconsidered for height (Appendix 11), diameter (Appendix 12), volume (Appendix 13) andwood density (Appendix 14), respectively.General combining ability (GCA) is defined as the average performance of theprogeny of an individual crossed to several other individuals (Falconer, 1981). GCA canbe expressed as absolute units; however, it is meaningful and convenient to express themas deviations from the overall mean (Zobel and Talbert, 1984). A positive GCA indicatesa parent that produces above-average progeny, while negative GCA suggests a parent thatproduces progeny with below average performance for the population.GCAs for growth performance and wood density were estimated for full-sibfamilies and are presented in Figures 5.5-5.8. In a tree improvement program, thedesirable parents would be ones that had both high GCA and that combined with otherparents to consistently produce families with high specific combining ability (SCA).From these results, the most promising parents can be selected and kept as goodcombiners, whereas some of the worst parents would be rejected. Table 5.22 summarizesmaternal and paternal parent ranking based on combining ability. Interestingly, parent#60 always performs at the top ranks for growth rate when used either as male or female,whereas parent #63 is very good only as male parent and parents #114 and #125 performvery well as female parents. These parents should be kept and included in seed orchardestablishment. However, some parents consistently ranked the lowest in all growth traits,such as male parent #133, and female parents #61, #160, and #175, and these parents maybe rejected.132Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyM28 35 36 45 60 63 70 92P Mean118 133 134 157 175 177 213 220 335 GCA height(m)Rank25 9.90 -0.90 9.90 3128 10.02 10.96 -0.28 10.52 2435 10.86 +0.06 10.86 1436 10.92 10.75^11.30 +0.20 11.00 1145 10.96 10.40^10.80 -0.07 10.73 1855 11.54 10.80^10.58 +0.17 10.97 1260 11.71 11.11 +0.62 11.42 461 9.80 -1.00 9.80 3362 11.08 +0.28 11.08 963 10.58 -0.22 10.58 2382 11.73 11.34^ 9.92 +0.28 11.08 887 10.78 -0.02 10.78 17114 11.50 +0.70 11.50 2118 11.66^11.14 11.16^10.94^11.28 +0.45 11.25 6125 11.73 +0.93 11.73 1145 10.81 +0.01 10.81 16153 10.85 10.83 11.10 +0.14 10.94 13160 9.50 10.18 10.31 -0.81 9.99 29162 10.86 +0.06 10.86 15167 11.34 +0.54 11.34 5172 10.50 1072 -0.19 10.61 21175 9.92 -0.88 9.92 30176 9.81 -0.99 9.81 32181 10.31 -0.49 10.31 27196 10.66^10.06^ 10.50 -0.38 10.42 26215 10.25 -0.55 10.25 28218 11.02 +0.22 11.02 10223 10.71 -0.09 10.71 20224 10.60 -0.20 10.60 22232 10.73 -0.07 10.73 19233 11.44 +0.64 11.44 3281 10.53 -0.27 10.53 17356 11.18 +0.38 11.18 7GCA +0.03 -0.60^-0.75 +0.35^+0.93^+0.34 +0.20^-0.03^+0.03^-0.79^+0.14^-0.22^-0.23^-0.22^+0.38^-0.02^-0.55Meanheight 10.83 10.20^10.05 11.15^11.73^11.14 11.00^10.77^10.83^10.01^10.94^10.58^10.57^10.58^11.18^10.78^10.25 10.80(cm)Rank 8 15^16 3^1^4 5^10^7^17^6^11^13^12^2^9^14 Test MeanFigure 5.5 Combining ability and ranking of maternal and paternal parents basedon height. (GCA = parent mean - test mean, M = maternal parent, P =paternal parent)133Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyM28 35 36 45 60 63 70 92 118P Mean133 134 157 175 177 213 220 335 GCA diameter(cm)Rank25 13.21 -2.09 13.21 3228 13.81^15.74 -0.47 14.83 2735 14.91 -0.39 14.91 2236 16.96^ 16.81^18.21 +2.05 17.35 145 16.31 14.28^15.59 +0.13 15.43 1555 17.22^ 14.49^13.11 -0.37 14.93 2160 17.80 15.52 +1.39 16.69 361 13.41 -1.89 13,41 3062 16.11 +0.81 16.11 563 14.58 -0.72 14.58 2682 16.32^16.06^ 13.28 +0.07 15.37 1687 14.75 -0.55 14.75 25114 16.28 +0.98 16.28 4118 16.54^15.92^15.03^15,09^15.09 +0.27 15.57 11125 17.09 +1.79 17.09 2145 14.90 -0.40 14.90 23153 15.04^15.56^ 15.12 -0.06 15.24 18160 12.97^14.79 14.86 -1.11 14.19 28162 15.52 +0.22 15.52 13167 15.67 +0.37 15.67 10172 15.13^ 15.93 +0.22 15.52 12175 12.81 -2.49 12.81 33176 15.68 +0.38 15.68 9181 15.68 +0.84 15.46 14196 14.33^13.87^13.99 -1.21 14.09 29215 13,36 -1.94 13.36 31218 14.46 -0.84 14.46 27223 15.85 +0.55 15.85 8224 16.08 +0.78 16.08 6232 15.00 -0.30 15.00 20233 16.07 +0.77 16.07 7281 15.25 -0.05 15.25 17356 15.19 -0.11 15.19 19GCA +0,24^-0.96^+0.27^+0.90^+1.02^+0.62^+0.11^+0.10^+0.59^-1.86^-0.21^-2.19^-1.18^-0.79^-0.11 -0.45^-1.94Meandiameter 15.54^14.34^15.57^1620^16.32^15.92^15.41^15.40^15.89^13.44^1509^13.11^14.12^14.51^15.19 14.85^13.36 15.30(m)Rank 6^13^5^2^1^3^7^8^4^15^10^17^14^12 9^11^16 Test MeanFigure 5.6 Combining ability and ranking of maternal and paternal parents basedon diameter. (GCA = parent mean - test mean, M = maternal parent,P = paternal parent)134Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyM28 35 36 45 60 63 70 92 118P Mean133 134 157 175 177 213 220 335 GCA volume(m3)Rank25 0.065 -0.027 0.065 3228 0.077^0.096 -0.005 0.087 2135 0.092 0.000 0.092 1436 0.103^ 0.100^0.125 +0.018 0.110 345 0.098 0.079^0.090 -0.003 0.089 1755 0.117^ 0.081 0.067 -0.004 0.088 1860 0.126 0.096 +0.019 0.111 261 0.070 -0.022 0.070 3162 0.102 +0.010 0.102 663 0.083 -0.009 0.083 2682 0.116^0.105 0.071 +0.007 0.099 787 0.083 -0.009 0.083 17114 0.105 0.013 0.105 27118 0.109^0.101^0.090 0.089^0.093 0.005 0.097 8125 0.124 +0.032 0.124 1145 0.085 -0.007 0.085 24153 0.089^0.092^ 0.092 -0.001 0.091 15160 0.062^0.083 0.084 -0.016 0.076 29162 0.092 0.000 0.092 13167 0.096 +0.004 0.096 10172 0.084^ 0.091 -0.005 0.087 20175 0.064 -0.028 0.064 33176 0.086 -0.006 0.086 23181 0.091 -0.001 0.091 16196 0.079 0.073^0.081 -0.013 0.079 28215 0.071 -0.021 0.071 30218 0.084 -0.008 0.084 25223 0.093 +0.001 0.093 12224 0.096 +0.004 0.096 9232 0.087 -0.005 0.087 22233 0.093 +0.001 0.093 12281 0.088 -0.004 0.088 19356 0.093 +0.001 0.093 11GCA 0.000^-0.012^-0.004 +0.011^+0.024 +0.009^+0.002^-0.002 +0.001 -0.023^-0.003^-0.025^-0.012^-0.007 +0.001 -0.006 -0.021Meanvolume 0.092^0.080^0.088^0.103^0.116^0.101^0.094^0.090^0.093 0.069^0.089^0.067^0.080^0.085^0.093 0.086 0.071 0.092(m3 )Rank 7^13^10^2^I^3^4^8^6 16^9^17^14^12^5 11 15 Test MeanFigure 5.7 Combining ability and ranking of maternal and paternal parents basedon volume. (GCA = parent mean - test mean, M = maternal parent, P= paternal parent)135Chapter 5 Genetic Parameters of Coastal Douglas-fir Progeny28 35 36 45 60 63 70 92 118 133 134 157 175 177 213 220 335 GCAMeanpilodyne(mm)Rank25 19.71 +1.19^19.71 528 20.54 21.68 -0.25^21.15 2235 19.79 +1.11^19.79 836 23.21^ 22.35^22.40 -1.76^22.66 3345 22.24 20.55^22.28 -0.82^21.72 2955 21.67^ 20.29^18.66 +0,69^20.21 1260 21.80 20.88 -0.46^21.36 2561 20.20 +0.70^20.20 1162 19.46 +1.44^19.46 463 20.91 -0.01^20.91 1982 21.89^21.77^ 20.30 -0.50^21.40 2787 20.56 +0.34^20.56 13114 22.01 -1.11^22.01 31118 21.78^20.77^21.08^21.12^20.53 -0.16^21.06 20125 22.31 -1.41^22.31 32145 20.60 +0.30^20.60 15153 19.82 21.75^ 20.27 +0.30^20.60 16160 21.37 21.41 21.28 -0.46^21.36 26162 20.76 +0,14^20.76 17167 19.73 +1.17^19.73 6172 21.28^ 21,19 -0.34^21.24 23175 19.37 +1.53^19.37 2176 21.09 -0.19^21.09 21181 20.81 +0,09^20.81 18196 19.91^20.57^19.53 +0.88^20.02 10215 18.84 +2.06^18.84 1218 19.98 +0.92^19.98 9223 21.29 -0.39^21.29 24224 21.51 -0.61^21.51 28232 20.56 +0.34^20.56 14233 21.74 -0.81^21.74 30281 19.76 +1.14^19.76 7356 19.43 +1.47^19.43 3GCA -0.11^+0.22^-0.05 -0.88^-0.99^+0.13^+0.57^-0.44^-0.32^+0.82^-0.22^+2.24^+0.94^+0.28^+1.47 +1,03^+2.06Meanheight 21.01^20.68^20.95 21.78^21.89^20.77^20.33^21.34^21.22^20.08^21.12^18.66^19.96^20.62^19.43 19.87^18.84 20.90(cm)Rank 12^9 11^16^17^10^7^15^14^6^13^1^5^8^3 4^2 Test MeanFigure 5.8^Combining ability and ranking of maternal and paternal parentsbased on pilodyne measurement. (GCA = test mean - parent mean,M = maternal parent, P = paternal parent)136Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgenyTable 5.22 Summary of paternal and maternal parents ranking of coastalDouglas-fir based on general combining abilityRank Height Diameter Volume PilodyneMale Female Male Female Male Female Male Female1 60 125 60 36 60 125 157 2152 213 114 45 125 45 60 335 1753 45 233 63 60 63 36 213 3564 63 60 118 114 70 114 220 625 70 167 36 62 213 233 175 256 134 118 28 224 118 62 133 1677 118 356 70 233 28 82 70 2818 28 82 92 223 92 118 177 359 220 62 213 176 134 224 35 21810 92 218 134 167 36 167 63 19611 157 36 220 118 220 356 36 6112 177 55 177 172 177 223 28 5513 175 153 35 162 35 162 134 8714 335 35 175 181 175 35 118 23215 35 162 133 45 335 153 92 14516 36 145 335 82 133 181 45 15317 133 87 157 281 157 45 60 16218 45 153 55 18119 232 356 281 6320 223 232 172 11821 172 55 28 17622 224 35 232 2823 63 145 176 17224 281 28 145 22325 28 87 218 6026 196 63 63 16027 181 218 87 8228 215 160 196 22429 160 196 160 4530 175 61 215 23331 25 215 61 11432 176 25 25 12533 61 175 175 36137Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.3.4 Genetic gainThe term genetic gain is a synonym for the term improvement in the breeder'slanguage (Young and Giese, 1990) and usually refers to economic benefit. Heritability isan index of reliability of the selected phenotypes, and one of several alternative ways toincrease the expected gain is to increase the heritability estimate. The heritabilityestimates from this experiment were low in growth traits and high in wood density,however, these results were similar to other studies.Genetic gain was estimated at five different intensities from 10% to 50% ofretained families or parents. In half-sib (C) families, gain for height was from 0.214 at50% selection intensity to 0.471 at 10% selection intensity (Table 5.23). Percent gainincreased when selection intensity increased, and selection based on height (1) gave thebest result. Moreover, selection based on volume (3) showed a better percent gain forheight than selection based on diameter (2). Gain estimated for diameter ranged from0.771 to 1.696 when selection intensity changed from 50% to 10%, and selection basedon diameter (2) reflected a better percent gain for diameter than height and volumeselection. The same result was presented in volume gain which ranged from 0.005 to0.011, and the highest percent gain was obtained after selecting the highest volume (3)growth families.Genetic gains for height in half-sib (P) families (Tables 3.24) were 0.245 to 0.539when selection intensities were made at 50 to 10%, respectively, while genetic gains fordiameter were 0.524 to 1.152. Genetic gains for volume growth (0.005-0.010) in half-sib(P) were comparable to gains for half-sib (C), whereas percent gains (11.219-30.521%) inhalf-sib (P) were larger than percent gains (10.524-20.467) in half-sib (C). However,volume in half-sib (C) was higher than in half-sib (P) families; therefore, more gain forvolume might be obtained from half-sib (C) than from half-sib (P) families.138Table 5.23 Genetic gain and percent gain of half-sib(C) family after selecting based on best families at different intensitiesand traitsTrait aph h2% Retained50% (i=0.798) 40% (i=0.966) 30% (i=1.159) 20% (i=1.4) 10% (i=1.755)% GainG % Gain G % Gain G % Gain G % Gain G1 2 3 1 2 3 1 2 3 1 2 3 1 2 3HeightDiameterVolume1.9894.1310.05350.1350.2340.1160.2140.7710.0052.9584.7149.1662.6145.7619.7942.4875.74110.5240.2590.9340.0063.5386.15611.7522.8987.49813.5122.8987.49813.5120.3111.1200.0073.7525.38011.3252.9368.50814.2332.5668.21014.6220.3761.3530.0094.3878.02816.2601.87910.28615.1623.66810.06818.2450.4711.6960.0114.6284.55711.6173.17412.48219.6314.26810.11020.4671 - % Gain based on height selection,2 - % Gain based on diameter selection,3 - % Gain based on volume selection.Table 5.24 Genetic gain and percent gain of half-sib (P) family after selecting based on best families at different intensities and traitsTrait aph h2% Retained50% (i=0.798) 40% (i=0.966) 30% (i=1.159) 20% (i=1.4) 10% (i=1.755)G % Gain G % Gain G % Gain G % Gain G % Gain1 2 3 1 2 3 1 2 3 1 2 3 1 2 3HeightDiameterVolume2.1194.1300.0490.1450.1590.1160.2450.5240.0053.6485.19510.1963.5815.79111.2433.5315.41111.2190.2970.6340.0054.8157.71314.8104.8157.71314.8104.8157.71314.8100.3560.7610.0075.6818.61617.4935.1269.13518.1275.1269.13518.1270.4300.9190.0087.0369.32619.7536.60111.20623.8386.60111.20623.8380.5391.1520.0108.95212.97730.5218.95212.97730.5218.95212.97730.5211 - % Gain based on height selection,2 - % Gain based on diameter selection,3 - % Gain based on volume selection.Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesGenetic gain estimates in full-sib families were considered from selection inmaternal parents (Table 5.25). Selection intensities were varied from 10% to 50% ofretained maternal parents. Gain for height ranged from 0.104 to 0.230 when the retainedmaternal parents were reduced from 50% to 10%. For this trait, selection of the parentswith better height in offspring (1) gave the highest percent gain (2.946-6.565), butselection based on diameter (2) provided the lowest percent gain (1.738-4.062). Geneticgain in diameter changed from 0.323 to 0.710 when the retained parents changed from50% to 10% and selection based on diameter was the largest (4.107-10.476). Selectionbased on height gave the lowest percent gain (2.730-7.220) for diameter. Genetic gainestimates from volume selection ranged from 0.003 to 0.007, and selection based onheight gave the lowest percent gain (6.958-17.942) for volume.In general, volume is usually the most valuable trait in tree improvement. Fromthis study, the maximum volume gain can be obtained through selecting the parents thatproduce offsprings with the largest diameter as well as largest volume. Since selection ofdiameter is more convenient, diameter selection to maximize gain is recommended.141Table 5.25 Genetic gain and percent gain of full-sib families after maternal parent selecting at different intensities and traitsTrait crph h2% Retained50% (i=0.798) 40% (i=0.966) 30% (i=1.159) 20% (i=1.4) 10% (i=1.755)G % Gain G % Gain G % Gain G % Gain G % Gain1 2 3 1 2 3 1 2 3 1 2 3 1 2 3HeightDiameterVolume1.9243.9640.05120.0680.1020.0820.1040.3230.0032.9462.7306.9581.7384.1077.4192.6113.8348.3840.1260.3910.0043.2433.1937.9102.2515.1278.8713.2174.80010.2070.1520.4690.0054.1663.0689.3782.3487.09812.2773.5825.33511.5110.1830.5660.0064.8834.15110.8323.5808.33915.4043.8557.00414.7590.2300.7100.0076.5657.22017.9424.06210.47618.8324.06210.49418.8321 - % Gain based on height selection,2 - % Gain based on diameter selection,3 - % Gain based on volume selection.Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.3.5 Age-age correlationCorrelation coefficients were estimated from mean family height of the sameprogenies from the nursery phase (1979) to 17-years of age over the three plantations ofcoastal Douglas-fir (Table 5.26). There were low correlations between nursery phases(1969 and 1970) vs. 1989 while the correlations of height growth in 1975 and 1977 to1989 were moderately high (0.746 and 0.782). These coefficients thus substantiate thehypothesis that major changes in family ranking can occur over time due to differentresponses of each family to the environment in each year. From these results it can beconcluded that 1975 and 1977 heights may provide good estimates for 1989 height. Kinget al., (1988a) had a similar result in their study of a coastal Douglas-fir progeny test,which concluded that age six height was an effective selection trait, and correlated wellwith age 12 volume. Therefore, at least 5 year old growth (1975) of progeny as evaluatedfrom this study can be used to predict the older age (i.e. 17 year old) growth in coastalDouglas-fir.Table 3.26 Correlation coefficients of height growth at different ages.(standard error of correlation coefficient in parentheses)Year 1969 1970 1975 19771970 0.436(0.099)1975 0.438 0.651(0.099) (0.084)1977 0.350 0.619 0.888(0.103) (0.087) (0.047)1988 0.274 0.542 0.746 0.782(0.106) (0.093) (0.068) (0.063)143Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesRegression equations and relationships of log e of age ratio (LAR) and r age ,age aregraphically presented in Figure 5.9. Age ratio and height correlation for each age ratio ofcoastal Douglas-fir progenies from four types of families were estimated to construct aregression equation. The equation was rage,age = 0.793 + 0.245 LAR, and thecorrelation for the regression was r = 0.385 (r 2=0.148). The correlation of regression inthis study is low; therefore, the equation may be of little use for prediction at older agesor rotation age. However, Lambeth (1980) failed to use the correlation of very young age(i.e. age 1 with all subsequent ages) for constructing the regression equation. Heindicated that selection in Douglas-fir at about one-quarter of rotation age is appropriate.In this study, the data were collected from young progenies (4, 6, and 17 year of ages),but the data set was quite small. Therefore, more data at different ages are required for anoptimal estimate of an equation.144r (age,age)1.0 r^ (age,age) =0.793+0.245LAR- R2 = 0.1480.80.6(4:6)(6:17)(4:17)^i ^I^I^,^I^I^(1.4)^(1.2)^(1)^(0.8)LAR (- Value)0.40.20(1.6) (0.6) (0.4) (0.2)Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesFull-sib^Half-sib(C)^Half-sib(P)^Control0^—*--^--R—^--.—Figure 5.9^Regression of age-age correlation for height growth of coastalDouglas-fir progenies to predict age ratio for selection.(LAR - Loge of age ratio)145Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.4 CONCLUSIONSFrom the evaluation of the plus-tree progeny tests, an existing variation among allfamilies was observed and the progenies derived from crossing among plus trees, such asfull-sib and half-sib (C) always show better performances in both survival and growthrate. The results of the study can be concluded as follow:1. Mean survival percentage of coastal Douglas-fir progenies over three test siteswas 77%. Courtenay plantation showed highest survival (85%), whereas Caycuse andGold River were 74% and 71%. On the basis of family types, full-sib and half-sib (C)families had higher survival than half-sib (P) and control families in all three plantations.2. Means for height, diameter, and volume of progenies at Gold River plantationwere highest, while Courtenay showed lowest growth rates. However wood density, wasnegatively correlated with growth rate. Growth performance of family types showedsignificant differences. Half-sib (C) usually performed the best in all three plantations,although not significantly different fom full-sib, whereas half-sib (P) and control familieswere not different from each other. Wood density usually ranked lowest in the familytype with fastest growing rate. However, half-sib (C) families with fastest growing ratehad better wood density than full-sib families.3. Individual heritability estimates for height, diameter, volume, and wooddensity (pilodyne) in three family types, including full-sib, half-sib (C), and half-sib (P),were low in growth traits and moderately high in wood density. In half-sib (C) families,individual heritabilities ranged from 0.116 (volume) to 0.494 (pilodyne). Half-sib (P)families were ranged from 0.116 (volume) to 0.189 (pilodyne). Heritability estimates forfull-sib families were low in growth traits (0.068-0.102) and moderately high in wooddensity (0.260). The results of heritability estimates were comparable to other studies ofthe same species.146Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies4. Ranking of parents was done in all three family types. In full-sib families,parents #60 performed in the top ranks when used as either as maternal or paternalparents. Some parent trees showed good performances when used only as male or femaleparents. These parent trees should be included in Douglas-fir improvement programs.However, some parents consistently performed at the lowest rank and these parentsshould be rejected.5. Prediction of genetic gain by initial mass selection at different intensitiesranging from 50 to 10% retained families were obtained for three traits--height, diameter,and volume. The study suggests that gains of 24 to 30% for volume can be expected atage 17 and that diameter growth is an effective trait for selection in coastal Douglas-fir.6. Age-age correlation estimates based on height growth of progenies fromnursery phase to year 17 indicated that field performance at year 5-6 can be used topredict the performance at 17 years of age. However, regression of age-age correlation topredict age ratio for height selection was not optimal due to small sample size and the lowcorrelation equation (R = 0.385).7. The parameters estimated can be used in the selection process. Crossingamong plus trees, such as in the seed orchard, should be a preferred practice forproducing genetically improved seedlings for plantation establishment.147Chapter 5 Genetic Parameters of Coastal Douglas-fir Progenies5.5 LITERATURE CITEDAnderson, R.L., and T.A. Bancroft. 1952. Statistical Theory in Research. McGraw Hill,New York.Bartram, V.C., 1977. Early Results of the Douglas-fir Cooperative Progeny Test.MaLhssf e i . The Faculty of Graduate Study. The University of British Columbia.Bridgwater, F.E., and E.C. Franklin, 1985.^Forest Tree Breeding: Strategies,Achievement and Constrains. In Cannell, M.G.R.,and J.E. Jackson (Eds.),Attributes of Trees As Crop Plants. Institute of Terrestrial Ecology NaturalEnvironment. pp. 36-48.Eis, S., and D. Craigdallie. 1983. Douglas-fir. Reproduction of Conifers. A Handbookfor Cone crop. Forestry Technical Report 31. Canadian Forestry Service.El-Kassaby, Y.A., and Y.S. Park. 1990. Harvest Index and Wood Density in a Douglas-fir Early Progeny Test. In Joint Meeting of Western Forest Genetics Association.4 IUF • • Worki • Parti- D u • 1 -Fr Co tort. • i, e itk ru e a dAbies Breeding and Genetic Resources. Olympia, Washington, USA. pp. 4.45-4.55.Falconer, D.S. 1981. Introduction to Quantitative Genetics. Second Edition. New York:John Wiley & Sons, Inc.Heaman, J.C. 1967. AReview of t h ePlus Tree Selection^ g^LProgrammefor Dou  s - f r inCoastal British Columbia. British Columbia Forest Service Research Notes No.44.King, J.N., F.C. Yeh, and C.H. Heaman. 1988a. Selection of Growth and Yield Traits inControlled Crosses. Silvae Genet. 37:158-164.King, J.N., F.C. Yeh, C.H. Heaman, and B.P. Dancik. 1988b. Selection of Wood Densityand Diameter in Controlled Crosses of Coastal Douglas-fir. Silvae Genet 37:152-157.Kovats, M. 1977. Estimating Juvenile Tree Volumes for Provenance and ProgenyTesting. Can. J. For. Res. 7:335-342.Kriebel, H.B. 1988. Molecular Biology in Forestry Research: When is It Relevant andHow Can We Use It? In Hallgren, J-E. (Ed.), MolecularProceedings of the Frans Kempe Symposium in timed. June 14-16, 1988. pp. 5-18Labrie, M.G. 1978. Early Results of the Douglas-fir Co-Operative Progeny Test. B.Sc. Thesis. Faculty of Forestry, The University of British Columbia.Lambeth, C.C. 1980. Juvenile-Mature Correlation in Pinaceae and Implications for EarlySelections. For. Sci. 26:571-580.148Chapter 5 Genetic Parameters of Coastal Douglas-fir ProgeniesNamkoong, G. 1979. Introduction to Quantitative Genetics in Forestry. TechnicalBulletin No. 1588. Forest Service, USDA.Okwuagwu, C.O., and R.P. Guries. 1981 Estimates of General and Specific CombiningAbility and Heritability for Juvenile Wood Specific Gravity and Tracheid Lengthin Jack Pine. In DeHayes, D.H., (Ed.), Proceedings of The Twenty-SeventhNortheastern Forest Tree Improvement Conference. Burlington, Vermont. July29-31, 1980. pp. 128-137.Olsen, W.C. 1988. Molecular Biology in Forestry Research: A Review. In Valentine,F.A. (Ed.), Forest and Crop Biotechnology. Progress and Prospect. Springer-Verlag. pp. 315-334.Orr-Ewing, A.L. 1969. The Development of a Programme for the Genetic Improvementof Douglas-fir in British Columbia. The Forestry Chronicle 45:395-399.Pacific Northwest Tree Improvement Research Cooperative 1990. Annual Report 1989-90. Forest Research Laboratory, Oregon State University.Place, I.C.M. 1969. Tree Breeding in Canada. The Forestry Chronicle 45:375-377.Sigurdson, L.C. 1971. Early Progeny Test Results of Douglas-fir. B.Sc. Thesis. Facultyof Forestry, The University of British Columbia.Sziklai, 0. 1971. Working Plan of Cooperative Progeny Test of Douglas-fir.Unpublished paper. April, 1971.van Buijtenen, J.P. 1984. Increasing Forest Productivity and Value by Breeding forOutstanding Combinations of Desirable Characteristics. In Forest Potentials.,Productivity and Value. Weyerhaeuser Science Symposium Volume 4.Proceedings of a Symposium held at Tacoma, Washington. August 20-24, 1984.pp.233-251.Young, R.A., and R.L. Giese. 1990. Introduction to Forest Science. Second Edition.New York: John Wiley & SonsZobel, B., and J. Talbert. 1984. Applied Forest Tree Improvement. New York: JohnWiley & Sons.Zobel, B.J., and J.P. van Buijtenen. 1989. Wood Variation: Its Causes and Control. Springer-Verlag. Berlin Heidelberg.Zobel, B.J., R.J. Weir, and J.B. Jett. 1972. Breeding Methods to Produce Progeny forAdvanced-Generation Selection and to Evaluate Parent Trees. Can. J. For. Res.2:339-345.149CHAPTER 6CONCLUSION AND RECOMMENDATIONIn this research thesis, genetic studies on Douglas-fir in B.C. were conductedusing both molecular genetic markers and morphological traits. The major results can besummarized as follows:6.1 GENETIC DIVERSITY OF DOUGLAS-FIR BASED ON MOLECULAR GENETICMARKERS6.1.1 Nuclear DNAGenetic variation in Douglas-fir populations in two geographic regions, coastaland interior, in B.C. were estimated using RADF markers. Five primers were employedand 97 loci were detected among amplification products separated in high resolutionpolyacrylamide gels stained with silver. This DNA staining procedure is highly sensitivewhich makes it easy to detect segregating loci. A practically unlimited number of locican be assessed and the raw data easily preserved. Therefore, the number of potentialgenetic markers is greatly increased relative to isozymes and to other molecular markersand is thus very useful for evaluating genetic diversity.Very high levels of genetic variation were observed from the RADF data. Amonggeographic regions, the same level of mean expected heterozygosity was observed in bothinterior (0.440) and coastal region (0.432), while the mean expected heterozygosity for allpopulations (both regions) was 0.380. Within populations genetic diversity and degree ofpopulation subdivision estimates were found to be similar for both regions.Effect of altitude position of Douglas-fir populations was studied in coastalpopulations located on Vancouver island. The elevation ranged from sea level to over600 m m maximum), as classified from stratification zones. There was no general150Chapter 6 Conclusion and Recommendationtrend observed to indicate that genetic variation of Douglas-fir populations varyaccording to altitude positions (R2 = 0.021 ). This observation shows the same result asobserved in Scots pine populations and some other species. This may be the result oflong distance pollen dispersal and long receptive periods providing an opportunity fortrees to receive pollen from other trees over considerable elevational range in mountaintopography.6.1.2 Chloroplast DNA variationVery low levels of polymorphism were detected with cpDNA probes in all of thepopulations sampled. However, transition zone populations had cpDNA morepolymorphic than coastal and interior regions. This contributed to higher degrees ofgenetic diversity estimates for cpDNA within transition zone populations than in coastaland interior populations. Based on cpDNA haplotype frequency estimates, levels ofpopulation subdivision ranging from 0.106 to 0.193 were observed to be not significantlydifferent among the three geographic regions, with interior populations the highest.6.1.3 Mitochondrial DNA variationPolymorphisms resulting from restriction site changes were observed both withinand among interior populations with all five enzymes and three mitochondrial geneprobes. However only one probe, ATPasea, detected polymorphism in coastal andtransitional populations. Therefore, genetic diversity estimates for mtDNA of interiorpopulations were much higher than for either coastal or transition populations. Thedifferent patterns of diversity between cpDNA and mtDNA in Douglas-fir might becaused by the differences in mode of inheritance, rates of evolution and also the functionsof genes in both genomes. The expansion of populations followed by intracellular151Chapter 6 Conclusion and Recommendationgenetic drift might be a cause of mtDNA evolution in interior populations. Because ofrapid structural rearrangement in the mtDNA genome and the wide distribution ofpopulations in interior regions, population subdivision (G sT of 0.327) estimates wererelatively high. Population subdivision in interior region was not significantly differentfrom coastal region but was significantly different for transitional region.6.1.4 Evaluation and application of the resultsAmong the three genomes, the pattern of genetic diversity is consistent withhigher gentic diversity within populations than among populations. This type of patternagrees with previous studies based on isozyme marker. However, much higher geneticdiversity could be observed using DNA markers. With haploid genome, mtDNA andcpDNA are powerful genetic markers for population studies. This interpretation dependson complete homoplasmy (no intraindividual variation). Theoretically heteroplamy couldarise either though mutation or by paternal contribution. However, it is important to notethat interpretations based on measuring variation at each component have some inherentlimitations. Furthermore, non-Mendelian DNA turnover mechanisms can alter patterns ofchange over time. For these reasons it is important to compare populations at a numberof genetic components, and contrast these results with the behavior of DNA at themolecular level, and the behavior of the organism at the population level.In evolutionary of Douglas-fir in B.C., the species is widely distributed and isusually broken up into many sub-populations being adapted to local environmentconditions. It is indicated that the diversity of genotype in a population is reallyenormous and the variation in such cases is ecotypical. Very small population generallytends to loss its genetic diversity. When the sub-populations are small and distributeddiscontinuously, the variation among populations is more likely to be caused by randon152Chapter 6 Conclusion and Recommendationdrift. Mutation itself does not shift gene frequency very much but it provides new alleleto be selected through evolution. However, possibility of a new mutant allele to beestablished in the population is very small, being almost zero ultimately, while the mostof them are generally elliminated during the first several generations. Therefore, Randomdrift is a powerful factor of differentiation of subpopulation, especially in the case ofsmall population size. Since natural forests are the most important source of geneticmaterials, efforts are needed for conserving them.6.2 GENETIC PARAMETERS OF 100 FAMILIES OF 17-YEAR OLD COASTALDOUGLAS-FIR PROGENIESGenetic parameters of 17-year old coastal Douglas-fir progenies were carried outthrough progeny testing. Three plantations, Caycuse, Courtenay, and Gold River, wereselected to be test sites. A total of 100 families derived from 60 parent trees, includingfull-sib, half-sib from clone bank (C), half-sib from original plus tree (P) and controlfamilies, were evaluated using height, diameter, volume, and Pilodyne measurement forwood density as variables.Mean survival percentage of coastal Douglas-fir progenies over three test siteswas 77%. Courtenay plantation showed highest survival (85%), whereas Caycuse andGold River were 74% and 71%. Based on family types, full-sib and half-sib (C) familieshad significantly higher survival than half-sib (P) and control families in all threeplantations. Means for height, diameter, and volume of progenies at Gold Riverplantation were highest, while Courtenay showed lowest growth rates. Wood density,was negatively correlated with growth rate. Growth performance of family types showedsignificant differences. Half-sib (C) usually performed the best in all three plantations,whereas half-sib (P) and control families were not different from each other. Wood153Chapter 6 Conclusion and Recommendationdensity usually ranked lowest in the family type with fastest growing rate . However,half-sib (C) families with fastest growing rate had better wood density than full-sibfamilies. Therefore, crossing among plus trees, such as in the seed orchard, should be apreferred practice for producing genetically improved seedlings for plantationestablishment.Individual heritability estimates for height, diameter, volume, and wood density(pilodyne) were carried out in three family types, including half-sib (C), half-sib (P), andfull-sib. In half-sib (C) families, individual heritabilities ranged from 0.116 (volume) to0.494 (Pilodyne). Half-sib (P) families were ranged from 0.116 (volume) to 0.189(pilodyne). Heritability estimates for full-sib families were low in growth traits (0.068-0.102) and moderately high in wood density (0.260).Ranking of parents was done in all three family types (appendices 3-18). In full-sib families, parents #60 performed in the top ranks when used as either as maternal orpaternal parents and some parents showed good performances when used only asmaternal or paternal parents. Since unbalance of male and female parents were used inthe test, some parents showed good general combing ability. These parent trees should beincluded in Douglas-fir improvement programs. However, some parents consistentlyperformed at the lowest rank and these parents should be rejected.Prediction of genetic gain by initial mass selection were obtained for three traits --height, diameter, and volume. Selections were made at different intensities ranging from50 to 10% retained families. Selection based on diameter and volume gave similarresults, with 24% gain obtained when selection was made at 10%. When selection wasfor only the top ten parents, 20% and 30% gain for volume were predicted in half-sib (C)and half-sib (P), respectively, whereas 18% gain was expected from full-sib family.154Chapter 6 Conclusion and RecommendationAge-age correlations of progenies from nursery phase to year 17 were carried outfor height growth. The result indicated that field performance at year 5-6 can be used topredict the performance at 17 years of age. However, regression of age-age correlation topredict age ratio for height selection was not optimal due to small sample size and the lowcorrelation equation (R = 0.385). For this test, more data at different ages are required foran accurated estimation.From the evaluation of the plus-tree progeny tests, an existing variation among allfamilies was observed and the parameters estimated can be used in the selection process.The study showed that gains of 24 to 30% for volume can be expected at age 17 and thatdiameter growth is an effective trait for selection in Douglas-fir. The results suggestedthat crossing among plus trees, such as in the seed orchard, should be a preferred practicefor producing genetically improved seedlings for plantation establishment.155Chapter 6 Conclusion and Recommendation6.3 RECOMMENDATIONS1. Since the three plant genomes do not evolve at the same rate and are notaffected by the same evolutionary forces, evolutionary history and genetic architecture ofall three genomes should be studied.2. In addition to this study, phylogenetic relationships among the three regions ofDouglas-fir in B.C. should be studied using a larger sample size. Moreover, samples ofcoastal Douglas-fir from Washington and interior Douglas-fir from Northern Idaho orMontana should be analyzed for all three genomes and used as outgroups for determiningphylogenetic relationships. Due to the continuous geographic ranges, the informationfrom these samples could help to test the path of migration of Douglas-fir in B.C.3. In genetic diversity calculations based on RFLP analyses observed inorganellar DNA in this study, the estimates based on allelic frequencies usually resultedin lower levels of diversity than were obtained with haplotype frequencies. Sinceorganelle DNAs are small in size, haplotype frequencies designated from all enzyme andprobe combinations for an individual should be the preferred method for evaluatinggenetic diversity at the population level over allelic frequencies.156APPENDICES157AppendicesAppendix 1 Registration of coastal Douglas-fir plus trees(* OB - Over bark, IB - Inner bark)Plus treeNo.Year Age Height(m)DBH (cm) S.G. Location Elev(m)OB* IB*Latitude(°N)Longitude(ow)25 1957 54 6.75 41.40 38.35 0.429 48° 52' 124° 04' 69128 1957 42 29.53 36.58 33.27 0.504 48° 56' 123° 53' 67634 1960 57 48.88 56.64 51.56 - 48° 52' 124° 08' 39035 1958 57 30.18 36.83 33.27 0.364 48° 52' 124° 03' 84136 1959 42 32.15 43.18 38.86 - 48° 51' 123° 47' 48043 1959 37 32.15 39.37 35.56 0.451 49° 52' 125° 38' 51145 1959 72 43.31 56.90 51.82 0.345 49° 33' 125° 03' 42049 1959 63 32.15 47.24 43.18 - 49° 17' 124° 33' 46555 1959 128 73.49 116.33 103.63 0.428 49° 53' 126° 05' 27060 1960 59 46.59 66.80 60.71 0.502 50° 15' 125° 24' 3061 1960 47 39.70 53.34 48.26 0.464 49° 25' 124° 58' 54162 1960 54 45.60 54.86 49.28 0.497 49° 18' 125° 13' 6063 1959 61 37.07 45.21 40.89 0.514 49° 43' 125° 10' 30069 1960 139 61.02 92.20 82.80 0.479 48° 39' 124° 10' 36070 1960 57 45.28 55.37 49.53 0.449 48° 35' 123° 58' 39076 1960 57 43.64 54.61 50.04 0.517 48° 53' 124° 23' 18082 1959 79 44.29 56.90 50.55 - 49° 18' 122° 34' 15087 1961 63 33.79 45.97 40.89 0.427 48° 52' 124° 03' 85692 1961 76 43.96 51.31 46.23 - 48° 47' 123° 58' 54193 1961 83 39.70 49.02 43.69 - 48° 47' 123° 58' 55695 1961 86 54.13 63.50 56.90 - 48° 47' 123° 58' 39096 1961 80 48.56 53.34 48.26 - 48° 47' 123° 58' 420110 1961 105 49.21 77.72 69.09 0.427 49° 52' 124° 19' 285114 1961 74 50.20 62.74 56.39 0.466 49° 52' 124° 19' 368118 1961 104 62.66 86.61 76.20 0.443 49° 52' 124° 19' 450125 1961 102 63.98 81.28 74.93 0.478 49° 54' 124° 19' 372133 1961 250 56.43 110.24 98.04 0.484 48° 37' 123° 53' 721134 1961 60 47.24 56.13 51.05 0.508 50° 15' 125° 34' 90145 1961 102 63.65 95.25 84.58 0.457 49° 52' 124° 19' 423153 1961 63 51.84 66.04 57.91 0.456 48° 37' 123° 57' 474157 1961 144 44.29 71.37 60.20 0.454 49° 54' 126° 10' 480158 1961 56 35.76 56.13 50.04 0.417 49° 15' 124° 30' 976160 1961 59 35.10 55.12 49.02 0.491 49° 15' 124° 30' 991162 1961 135 52.17 90.42 80.26 - 48° 36' 124° 10' 300165 1961 56 45.28 57.15 50.04 0.451 48° 36 123° 59' 255166 1959 74 44.62 65.02 58.42 - 49° 33' 125° 03' 586167 1959 73 41.99 50.04 45.21 - 49° 33' 125° 03' 553172 1960 49 41.99 50.80 46.74 - 49° 25' 124° 58' 480175 1961 59 44.62 56.90 50.29 0.506 49° 05' 124° 13' 270176 1961 148 46.92 74.42 65.53 0.456 49° 52' 126° 09' 270177 1961 150 63.98 91.44 80.52 0.409 49° 52' 126° 09' 240181 1961 91 29.86 54.10 46.48 0.485 49° 53' 126° 12' 210196 1962 80 51.51 74.68 68.83 - 49° 08' 121°34' 781207 1962 88 46.26 72.39 64.52 0.448 50° 07' 127° 19' 60208 1962 84 50.85 78.99 69.09 0.456 50° 02' 127° 18' 30213 1962 85 48.56 73.66 65.53 0.523 50° 02' 127° 17' 60215 1962 142 60.04 96.27 84.07 0.507 50° 04' 127° 05' 240218 1962 130 54.79 88.90 80.77 0.480 50° 04' 127° 04' 450220 1962 48 41.99 54.10 48.51 0.496 49° 54' 126° 49' 60223 1962 80 49.54 67.31 59.18 - 49° 00' 121° 45' 796224 1962 79 41.34 53.59 0.00 - 49° 00' 121° 45' 958226 1962 62 44.62 65.53 55.37 - 50° 03' 125° 00' 30232 1962 52 44.29 57.66 0.00 - 50° 13' 125° 14' 128233 1962 49 44.29 61.21 0.00 - 50° 13' 125° 15' 135235 1962 140 58.07 94.74 85.09 0.425 50° 04' 127° 04' 315281 1962 132 65.94 99.06 90.42 - 49° 06' 125° 34' 240335 1963 41 33.14 43.94 39.88 - 49° 56' 125° 36' 51351 1963 54 34.12 45.97 42.42 - 49° 34' 125° 05' 480356 1964 81 52.49 70.61 63.50 0.511 49° 40' 125° 52' 8444 1964 81 50.85 75.44 68.58 - 48° 52' 124° 16' 210158AppendicesAppendix 2 Family number, parent trees, and average survivalpercentage of coastal Douglas-fir progenies fromthree progeny test sitesFamilyNo.Parents FamilyNo.ParentsMaternal Paternal Maternal Paternal1 55 45 51 160 352 60 45 52 224 353 114 45 53 62 704 118 45 54 87 705 125 45 55 36 Polymix6 232 45 56 55 Open (CB)7 233 45 57 1108 118 63 58 1349 118 92 59 36 II10 55 118 60 45 n11 63 118 61 160 n12 118 134 62 166 II13 118 175 63 93 n14 28 45 64 207 II15 36 45 65 208 n16 153 45 66 215 n17 160 45 67 22018 172 45 68 226 n19 36 92 69 23520 45 92 70 356 n21 36 118 71 ROM 103922 45 118 72 ROM 134223 153 118 73 ROM 43224 160 118 74 ROM 104725 172 118 75 ROM 103326 162 70 76 ROM 98627 167 70 77 ROM 103228 196 70 78 28 Open (P)29 175 133 79 3430 196 133 80 43 n31 25 175 81 49 n32 196 175 82 63 n33 45 28 83 69 n34 145 28 84 70 n35 223 28 85 76 "36 55 175 86 96 "37 153 28 87 114 II38 215 335 88 153 II39 176 36 89 162 n40 82 177 90 165 n41 82 70 91 172 n42 82 60 92 177 "43 181 36 93 223 "44 60 177 94 224 n45 281 220 95 95 n46 218 220 96 158 n47 356 213 97 351 n48 28 35 98 444 n49 35 35 99 Bare root (2+2)50 61 35 100 Bare root (2+2)159AppendicesAppendix 3 Ranking of height growth of half-sib (C) families of coastal Douglas-firprogeniesRank Family # Average1989 Height (m)Duncan Parenta=0.05^M# ofTreesCaycuse Courtenay Gold River1 57 57 68 57 11.461 a 110 442 63 60 60 69 11.461 a 235 413 69 66 63 55 11.377 a 36* 394 58 65 55 56 11.272 ab 55 405 56 67 69 63 11.251 ab 93 436 62 56 59 60 11.226 ab 45 427 67 69 62 58 11.156 abc 134 418 55 59 58 67 11.032 abc 220 449 68 58 56 68 11.013 abc 226 3910 61 55 66 66 10.956 abc 215 3411 65 68 61 65 10.921 abc 208 4212 60 63 67 62 10.910 abc 166 4213 59 62 65 59 10.884 abc 36 4514 66 70 70 61 10.372 be 160 4315 70 64 57 70 10.159 c 356 3816 64 61 64 64 9.332 c 207 31160AppendicesAppendix 4 Ranking of diameter growth of half-sib (C) families of coastalDouglas-fir progeniesRank Family # Average^Duncan Parent # ofDiameter(cm) a=0.05^M^TreesCaycuse Courtenay Gold River 19891 57 60 68 55 18.221 a 36 392 59 55 55 60 17.338 ab 45 423 63 59 60 59 17.098 abc 36 454 67 57 69 57 16.593 abcd 110 445 55 65 63 69 16.363 abcd 235 416 58 66 59 63 16.165 bcde 93 437 69 69 58 67 15.884 bcde 220 448 56 67 61 56 15.800 bcde 55 409 60 56 62 58 15.795 bcde 134 4110 61 63 66 65 15.590 bcde 208 4211 65 68 70 68 15.449 bcde 226 3912 62 58 56 61 15.088 cde 160 4313 68 70 65 66 14.808 de 215 3414 70 61 67 62 14.640 de 166 4215 66 64 57 70 14.271 e 356 3816 64 62 64 64 11.794 f 207 31161AppendicesAppendix 5 Ranking of volume growth of half-sib (C) families of coastal Douglas-firprogeniesRank Family # Average1989 Volume (m3)Duncan Parenta=0.05^M# ofTreesCaycuse Courtenay Gold River1 57 60 68 55 0.1239 a 36 392 63 65 55 57 0.1130 ab 110 443 58 57 60 60 0.1116 ab 45 424 55 55 63 63 0.1078 ab 93 435 67 59 69 59 0.1068 ab 36 456 59 66 59 69 0.1056 ab 235 417 69 67 58 58 0.1018 ab 134 418 56 69 70 67 0.0976 ab 220 449 60 58 62 68 0.0975 ab 226 3910 62 68 66 56 0.0956 b 55 4011 65 70 61 65 0.0942 b 208 4212 61 56 65 66 0.0893 b 215 3413 68 63 67 62 0.0867 b 166 4214 70 61 56 61 0.0861 b 160 4315 66 62 57 70 0.0859 b 356 3816 64 64 64 64 0.0506 c 207 31162AppendicesAppendix 6 Ranking of diameter growth of half-sib (C) families of coastalDouglas-fir progeniesRank Family # AveragePilodyne(mm)Duncan Parenta=0.05^M# ofTreesCaycuse Courtenay Gold River 19891 66 68 66 66 19.221 a 215 342 64 63 64 70 19.276 a 356 383 70 69 63 64 19.532 oh 207 314 62 56 70 62 19.857 abc 166 425 65 62 65 63 19.860 abc 93 436 68 66 69 69 20.159 abc 235 417 56 64 67 68 20.167 abc 226 398 61 70 58 65 20.310 abc 208 429 58 58 56 56 20.350 abc 55 4010 67 67 61 67 20.557 bcd 220 4411 69 61 57 58 20.720 cd 134 4112 63 65 62 61 20.733 cd 160 4313 55 57 59 57 21.477 de 110 4414 57 55 60 59 22.044 e 36 4515 59 60 68 60 22.048 e 45 4216 60 59 55 55 22.128 e 36 39163AppendicesAppendix 7 Ranking of height growth of half-sib (P) families of coastal Douglas-firprogeniesRank Family # Average1989^Height (m)Duncanoc=0.05Parent^# ofM^TreesCaycuse Courtenay Gold River1 93 87 84 87 11.303 a 114 342 79 94 98 93 10.858 ab 223 433 87 81 96 89 10.674 abc 162 384 91 89 92 79 10.653 abc 34 385 89 82 90 91 10.469 abcd 172 356 83 93 85 84 10.427 abcde 70 417 85 79 87 94 10.418 abcde 224 398 95 91 88 82 10.291 abcde 63 349 84 78 93 85 10.203 abcde 76 3010 92 85 94 81 10.157 abcde 49 3011 78 84 89 92 10.128 abcde 177 4012 82 95 91 95 10.121 abcde 95 3413 96 88 82 90 9.997 bcde 165 3014 86 80 79 78 9.939 bcde 28 3815 81 90 86 88 9.848 bcde 153 3316 94 83 80 80 9.712 bcde 43 4117 90 92 95 83 9.574 cde 69 3518 80 96 97 96 9.544 cde 158 3219 88 98 81 98 9.250 de 444 2620 97 97 78 97 9.223 e 351 3521 86 83 86 9.214 e 96 14164AppendicesAppendix 8 Ranking of diameter growth of half-sib (P) families of coastal Douglas-firprogeniesRank Family # Average1989 Diameter (cm)Duncan Parenta=0.05^M# ofTreesCaycuse Courtenay Gold River1 91 87 84 93 16.049 a 223 432 93 82 98 87 15.932 ab 114 343 87 91 82 91 15.777 abc 172 354 79 94 93 79 15.211 abcd 34 385 81 93 88 82 14.965 abcde 63 346 89 81 92 94 14.751 abcde 224 397 82 95 79 89 14.703 abcde 162 388 94 89 90 84 14.666 abcde 70 419 95 78 94 95 14.356 abcdef 95 3410 84 79 89 85 13.963 abcdef 76 3011 85 85 91 90 13.963 abcdef 165 3012 96 84 85 92 13.910 abcdef 177 4013 90 80 87 81 13.553 bcdef 49 3014 83 83 95 78 13.432 cdef 28 3815 80 90 97 80 13.200 def 43 4116 78 88 96 97 13.097 def 351 3517 92 96 78 88 13.042 def 153 3318 97 92 80 96 12.919 def 158 3219 86 98 81 83 12.797 def 69 3520 88 97 86 98 12.638 of 444 2621 98 86 83 86 12.129 f 96 14165AppendicesAppendix 9 Ranking of volume growth of half-sib (P) families of coastal Douglas-firprogeniesRank Family # Average1989 Volume (m3)Duncan Parenta=0.05^M# ofTreesCaycuse Courtenay Gold River1 91 87 90 87 0.1064 a 114 342 87 82 98 93 0.0978 ab 223 433 93 81 84 91 0.0938 abc 172 354 79 94 82 79 0.0882 abcd 34 385 82 91 85 82 0.0843 abcde 63 346 89 93 92 84 0.0828 abcde 70 417 81 78 79 94 0.0815 abcde 224 398 85 85 93 89 0.0812 abcde 162 389 96 80 94 90 0.0811 abcde 165 3010 84 95 87 85 0.0782 bcde 76 3011 95 84 88 92 0.0758 bcde 177 4012 94 89 89 95 0.0741 bcde 95 3413 90 79 91 81 0.0702 bcde 49 3014 80 88 96 80 0.0687 cde 43 4115 83 83 95 98 0.0679 cde 444 2616 92 92 97 78 0.0677 cde 28 3817 78 90 78 96 0.0672 cde 158 3218 97 96 80 88 0.0660 cde 153 3319 86 98 81 97 0.0640 de 351 3520 88 97 86 83 0.0600 de 69 3521 86 83 86 0.0566 e 96 14166AppendicesAppendix 10 Ranking of Pilodyne measurement of half-sib (P) families of coastalDouglas-fir progeniesRank Family # Average1989 Pilodyne (mm)Duncan Parenta=0.05^M# ofTreesCaycuse Courtenay Gold River1 97 86 83 81 18.867 a 49 302 88 97 81 97 19.086 ab 351 353 96 81 78 83 19.143 abc 69 354 78 92 85 78 19.184 abc 28 385 92 83 79 92 19.612 abcd 177 406 85 98 92 85 19.633 abed 76 307 83 80 82 80 19.683 abcd 43 418 81 84 97 88 19.773 abcd 153 339 80 85 91 86 19.821 abcd 96 1410 90 90 90 96 20.094 abcde 158 3211 91 78 80 90 20.100 abcde 165 3012 89 88 93 84 20.171 abcde 70 4113 82 79 84 79 20.289 abcde 34 3814 86 96 96 89 20.447 bcde 162 3815 95 89 86 98 20.577 cde 444 2616 84 95 95 82 20.588 cde 63 3417 94 94 88 95 20.721 de 95 3418 87 91 89 91 20.871 de 172 3519 79 93 87 93 20.977 de 223 4320 93 82 94 94 21.077 de 224 3921 - 87 98 87 21.456 e 114 34167AppendicesAppendix 11 Ranking of maternal parents in full-sib families based on generalcombining ability for heightRank Maternal^DuncanTree#^a=0.05# ofCrosses# ofProgeniesHeight (m)Mean^SD1 125 a 1 43 11.728 2.9412 114 ab 1 42 11.500 1.9533 233 abc 1 38 11.439 2.1264 60 abcd 2 80 11.420 1.9145 167 abcde 1 39 11.341 2.0576 118 abcdef 5 202 11.246 2.0127 356 abcdefg 1 41 11.180 2.2698 82 abcdefgh 3 113 11.081 2.7339 62 abcdefgh 1 48 11.077 2.13210 218 abcdefgh 1 41 11.020 2.30711 36 abcdefgh 3 128 11.002 1.83012 55 abcdefgh 3 127 10.972 2.04513 153 abcdefgh 3 125 10.935 2.07214 35 abcdefghi 1 44 10.864 2.48915 162 abcdefghi 1 44 10.861 2.06216 145 bcdefghi 1 41 10.812 1.82517 87 bcdefghij 1 40 10.780 1.94618 45 bcdefghijk 3 121 10.731 1.92319 232 bcdefghijk 1 39 10.728 2.13320 223 bcdefghijk 1 46 10.711 1.96921 172 bcdefghijkl 2 84 10.608 1.59322 224 cdefghijkl 1 49 10.600 2.36523 63 cdefghijkl 1 39 10.585 2.39524 281 defghijkl 1 40 10.535 2.12825 28 efghijkl 2 77 10.519 2.45126 196 fghijkl 3 108 10.422 2.39227 181 ghijkl 1 43 10.307 2.46628 215 hijkl 1 31 10.252 2.35629 160 ijkl 3 106 9.988 2.23530 175 jkl 1 24 9.925 2.54631 25 kl 1 36 9.897 2.17232 176 1 1 44 9.807 1.90833 61 1 1 45 9.804 2.464Total 54 2,168168AppendicesAppendix 12 Ranking of maternal parents in full-sib families based on generalcombining ability for diameterRank Maternal^DuncanTree#^a=0.05# ofCrosses# of Diameter (cm)Progenies Mean SD1 36 a 3 128 17.353 4.5732 125 ab 1 43 17.086 5.9643 60 abc 2 80 16.690 5.3574 114 abcd 1 42 16.276 4.5195 62 abcd 1 48 16.110 5.0956 224 abcd 1 49 16.078 4.8987 233 abcde 1 38 16.071 5.3918 223 abcdef 1 46 15.846 4.5699 176 abcdef 1 44 15.677 4.93910 167 abcdef 1 39 15.667 4.13811 118 abcdef 5 202 15.575 4.91012 172 bcdef 2 84 15.525 4.31113 162 bcdef 1 44 15.516 4.85714 181 bcdef 1 43 15.465 5.25215 45 bcdef 3 121 15.429 4.45616 82 bcdef 3 113 15.371 5.33317 281 bcdef 1 40 15.252 5.04718 153 bcdef 3 125 15.239 4.86319 356 cdefg 1 41 15.195 4.70020 232 cdefgh 1 39 14.956 4.65121 55 cdefgh 3 127 14.926 4.74922 35 cdefgh 1 44 14.907 5.05723 145 cdefgh 1 41 14.900 4.71124 28 cdefgh 2 77 14.835 5.14125 87 defgh 1 40 14.748 4.09226 63 defgh 1 39 14.579 4.73427 218 defghi 1 41 14.456 4.69728 160 efghi 3 106 14.191 4.96029 196 fghi 3 108 14.087 4.84430 61 ghi 1 45 13.407 5.02031 215 hi 1 31 13.358 4.48632 25 hi 1 36 13.206 4.34733 175 i 1 24 12.808 4.193Total 54 2,168169AppendicesAppendix 13 Ranking of maternal parents in full-sib families based on generalcombining ability for volumeRank Maternal^DuncanTree#^a=0.05# ofCrosses# of Volume (m3)Progenies Mean SD1 125 a 1 43 0.124 0.0792 60 ab 2 80 0.111 0.0763 36 abc 3 128 0.110 0.0644 114 abcd 1 42 0.105 0.0635 233 abcd 1 38 0.105 0.0686 62 abcde 1 48 0.102 0.0667 82 bcdef 3 113 0.100 0.0758 118 bcdef 1 202 0.097 0.0629 224 bcdef 1 49 0.096 0.06010 167 bcdef 1 39 0.096 0.05511 356 bcdefg 1 41 0.093 0.06112 223 bcdefg 1 46 0.093 0.05713 162 bcdefg 1 44 0.092 0.06014 35 bcdefg 1 44 0.092 0.06415 153 bcdefg 3 125 0.091 0.06116 181 bcdefg 1 43 0.091 0.06517 45 bcdefg 3 121 0.089 0.05618 55 bcdefgh 3 127 0.088 0.06119 281 bcdefgh 1 40 0.088 0.05620 172 bcdefghi 2 84 0.087 0.05421 28 bcdefghi 2 77 0.087 0.06222 232 cdefghi 1 39 0.087 0.05823 176 cdefghi 1 44 0.086 0.06224 145 defghi 1 41 0.085 0.05625 218 defghi 1 41 0.084 0.05626 63 defghi 1 39 0.083 0.05327 87 defghi 1 40 0.083 0.05028 196 efghi 3 108 0.079 0.05829 160 fghi 3 106 0.076 0.05930 215 ghi 1 31 0.071 0.05431 61 ghi 1 45 0.070 0.05932 25 hi 1 36 0.065 0.04233 175 i 1 24 0.064 0.046Total 54 2,168170AppendicesAppendix 14 Ranking of maternal parents in full-sib families based on generalcombining ability for pilodyne measurementRank MaternalTree#Duncana=0.05# ofCrosses# ofProgeniesPilodyne (mm)Mean SD1 215 a 1 31 18.839 3.2002 175 ab 1 24 19.375 3.5273 356 ab 1 41 19.427 3.2864 62 ab 1 48 19.458 3.2485 25 abc 1 36 19.708 3.5446 167 abc 1 39 19.731 3.1747 281 abcd 1 40 19.762 3.4578 35 abcd 1 44 19.795 3.6759 218 bcde 1 41 19.976 3.06610 196 bcdef 3 108 20.019 3.53911 61 bcdefg 1 45 20.200 3.16812 55 bcdefg 3 127 20.209 3.69913 87 cdefgh 1 40 20.563 2.81314 232 cdefgh 1 39 20.564 3.69615 145 cdefgh 1 41 20.598 3.43916 153 cdefgh 3 125 20.600 3.34917 162 cdefghi 1 44 20.761 3.39518 181 defghi 1 43 20.814 3.64219 63 efghi 1 39 20.910 3.38320 118 fghij 5 202 21.062 3.48921 176 fghij 1 44 21.091 3.53322 28 ghij 2 77 21.149 3.05223 172 ghij 2 84 21.238 3.32124 223 hijk 1 46 21.293 2.99925 60 hijk 2 80 21.356 3.59126 160 hijk 3 106 21.358 3.36127 82 hijk 3 113 21.398 3.63028 224 hijk 1 49 21.510 3.54529 45 ijk 3 121 21.723 3.43730 233 jk 1 38 21.737 3.57331 114 jk 1 42 22.012 3.56932 125 k 1 43 22.314 3.99333 36 1 3 128 22.656 3.305Total 54 2,168171AppendicesAppendix 15 Ranking of paternal parents in full-sib families based on generalcombining ability for heightRank PaternalTree#Duncanot:).05# ofCrosses# ofProgeniesHeight (m)Mean SD1 60 a 1 42 11.729 2.7612 213 ab 1 41 11.180 2.2693 45 ab 12 492 11.149 2.1114 63 ab 1 45 11.144 2.2065 70 b 6 253 11.003 2.1806 134 be 1 38 10.945 2.0297 118 bcd 7 288 10.832 1.9698 28 bcd 4 169 10.831 1.9639 220 bcd 2 81 10.780 2.22010 92 bcd 3 116 10.772 1.94811 157 bcde 1 41 10.583 2.04612 177 bcde 2 71 10.576 2.41413 175 bcde 3 104 10.575 2.30014 335 cde 1 31 10.252 2.35615 35 de 5 210 10.198 2.56616 36 e 2 87 10.054 2.20317 133 e 2 59 10.007 2.324Total 54 2,168172AppendicesAppendix 16 Ranking of paternal parents in full-sib families based on generalcombining ability for diameterRank PaternalTree#Duncana=0.05# ofCrosses# ofProgeniesDiameter (cm)Mean SD1 60 a 3 42 16.319 5.4602 45 a 2 492 16.197 4.9823 63 ab 2 45 15.924 5.3854 118 abc 5 288 15.586 4.7895 36 abc 1 87 15.572 5.0676 28 abc 1 169 15.543 4.5807 70 abc 6 253 15.413 4.7108 92 abc 2 116 15.399 4.7029 213 abc 3 41 15.195 4.70010 134 abc 2 38 15.095 4.76411 220 abcd 12 81 14.849 4.85912 177 bcde 7 71 14.515 5.20313 35 cde 1 210 14.337 5.06014 175 cde 1 104 14.123 4.80015 133 de 1 59 13.437 4.59916 335 e 1 31 13.358 4.48617 157 e 4 41 13.107 3.964Total 54 2,168173AppendicesAppendix 17 Ranking of paternal parents in full-sib families based on generalcombining ability for volumeRank PaternalTree#Duncana=0.05# ofCrosses# of Volume (m3)Progenies Mean SD1 60 a 1 42 0.116 0.0852 45 ab 12 492 0.103 0.0673 63 abc 1 45 0.101 0.0644 70 bcd 6 253 0.094 0.0605 213 bcd 1 41 0.093 0.0616 118 bcd 7 288 0.093 0.0617 28 bcd 4 169 0.092 0.0568 92 bcd 3 116 0.090 0.0599 134 bcd 1 38 0.089 0.05610 36 bcd 2 87 0.088 0.06311 220 cde 2 81 0.086 0.05512 177 cdef 2 71 0.085 0.06613 35 defg 5 210 0.080 0.06014 175 defg 3 104 0.080 0.05715 335 efg 1 31 0.071 0.05416 133 fg 2 59 0.069 0.05017 157 g 1 41 0.067 0.045Total 54 2,168174AppendicesAppendix 18 Ranking of paternal parents in full-sib families based on generalcombining ability for pilodyne measurementRank PaternalTree#Duncana=0.05# ofCrosses# of Pilodyne (mm)Progenies Mean SD1 157 a 1 41 18.659 3.2062 335 a 1 31 18.839 3.2003 213 ab 1 41 19.427 3.2864 220 be 2 81 19.870 3.2465 175 bcd 3 104 19.957 3.7706 133 bcd 2 59 20.085 3.4897 70 cde 6 253 20.334 3.2318 177 cdef 2 71 20.620 4.0549 35 cdef 5 210 20.681 3.45210 63 def 1 45 20.767 3.70311 36 efg 2 87 20.954 3.56912 28 efg 4 169 21.012 3.33713 134 efgh 1 38 21.118 3.74214 118 efgh 7 288 21.222 3.48015 92 fgh 3 116 21.345 3.21016 45 gh 12 492 21.782 3.48617 60 h 1 42 21.893 3.167Total 54 2,168175

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.831.1-0075159/manifest

Comment

Related Items