UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Western gall rust (Endocronartium harknessii (Moore) Hirat.) on lodgepole pine (Pinus contorta dougl.)… Kojwang, Harrison Ochieng 1989

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

Item Metadata

Download

Media
831-UBC_1989_A1 K63.pdf [ 11.8MB ]
Metadata
JSON: 831-1.0098173.json
JSON-LD: 831-1.0098173-ld.json
RDF/XML (Pretty): 831-1.0098173-rdf.xml
RDF/JSON: 831-1.0098173-rdf.json
Turtle: 831-1.0098173-turtle.txt
N-Triples: 831-1.0098173-rdf-ntriples.txt
Original Record: 831-1.0098173-source.json
Full Text
831-1.0098173-fulltext.txt
Citation
831-1.0098173.ris

Full Text

W E S T E R N G A L L R U S T (END0CR0NARTIUM HARKNESSII (MOORE) HIRAT.) O N L O D G E P O L E PINE (PINUS CONTORTA DOUGL.) IN BRITISH C O L U M B I A - A S T U D Y OF V A R I A T I O N A N D I N H E R I T A N C E O F R E S I S T A N C E IN A N A T U R A L P A T H O S Y S T E M . By Harrison Ochieng Kojwang B. Sc. (Forestry) University of Nairobi, 1980 M. Sc. (Forestry) University of Helsinki, 1983 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR T H E DEGREE OF D O C T O R OF PHILOSOPHY in T H E FACULTY OF GRADUATE STUDIES FOREST SCIENCES We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA 1989 © Harrison Ochieng Kojwang, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Forest Sciences The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: 7a/,, 21, I9tq ABSTRACT Clones, open-pollinated, and full-sib families of lodgepole pine were inoculated with various spore collections of western gall rust to assess and describe variation in early symptom development, host resistance, and rust virulence and to determine the mode of inheritance of resistance. In addition, studies of the cytology of immature, mature and germinating aeciospores and one-dimensional SDS-PAGE (with silver stain) of total spore protein were undertaken. The frequency of some early symptoms varied significantly between open-pollinated families under some inoculation conditions, but was not related to the susceptibility of these families. In addition, the proportion of symptomatic seedlings that became galled was only slightly greater than that of asymptomatic seedlings. Early symptoms were not reliable indicators of successful infection. The frequency of uninucleate cells (58%) did not vary between the youngest and the oldest cells in immature spore chains. In mature spores, 57.5, 41.0 and 1.5 percent were uni-, bi- and trinucleate respectively. The number of nuclei in spores and germtubes increased gradually following germination up to an average of 5.6 (range of 2-9) at 34 hours. At no stage during the development and germination of aeciospores was there evidence of karyogamy in the form of a reduction in the number of nuclei per spore. Karyogamy and meiosis do not occur at spore germination in the coastal rust population sampled. Silver stained SDS-PAGE gels showed some variation among single gall spore sources. The approach has potential as a technique for distinguishing among spore sources. Sixteen grafted clones inoculated with four single-gall spore sources showed a sig-nificant interaction between clone and spore source. There were also large differences in relative susceptibility among pine clones and smaller differences among spore sources with respect to the average infection levels of pine clones. The infection levels of clones was considered to provide a better measure of the genetically determined resistance of parent trees than the degree of infection of those trees in the field. Forty open-pollinated pine families inoculated with coastal and interior spore collec-tions showed significant spore-family interactions attributable to six pine families that showed equal susceptibility to both spore sources. The coastal spore source caused much higher infection than the interior source on the other families. Estimates of narrow sense heritability h2 were as follows; h2Indiv = 0.21 ± 0.10, h2Family = 0.51 ± 0.16. Regressions of the infection levels of offspring on those of their female parents were not significant. Hence selection of superior individuals requires progeny testing. A 4 by 4 diallel showed significant GCA effects and barely detectable SCA effects. The SCA component was about one third of the GCA component, indicating that inheritance of resistance is largely additive. Reciprocal and maternal effects were not significant. Stability in the pathosystem was attributed to the wide variation in host resistance and some degree of differential interactions between pine and spore genotypes. The highly variable host populations interact with much less variable pathogen populations; the latter possibly caused by the lack of sexual reproduction. As a result, the rate of selection for greater virulence may be matched by the rate of selection for resistance in spite of the much shorter life cycle of the pathogen. m Table of Contents List of Tables x List of Figures xix Acknowledgements xxii 1 G E N E R A L I N T R O D U C T I O N 1 2 L I T E R A T U R E REVIEW: N A T U R A L P A T H O S Y S T E M S 5 2.1 General 5 2.2 Disease In Some Natural Pathosystems 8 2.3 The Evolution of Stable Natural Pathosystems 11 2.3.1 General Theory 11 2.3.2 Stability Strategies of Vertical and Horizontal Pathosystems . . . 12 2.3.3 Theoretical Models 13 2.3.4 Resistance Frequency Distributions among Families Derived from Wild Populations 15 2.4 Genetic diversity and host-pathogen interactions in some stable natural Pathosystems 17 2.5 Genetic Diversity and Stability in Crop Multilines 19 3 C Y T O L O G Y OF D E V E L O P I N G A N D G E R M I N A T I N G SPORES 22 3.1 Introduction 22 3.2 Materials and Methods 25 i v 3.2.1 Cytology of Spore Chains 25 3.2.2 Cytology of Dormant and Germinating Spores 26 3.2.3 Cytology of Germtubes 27 3.3 Results 28 3.4 Discussion 34 4 E A R L Y S Y M P T O M S O N SEEDLINGS A N D G A L L F O R M A T I O N 38 4.1 Introduction 38 4.2 Materials and Methods 40 4.2.1 Materials 40 4.2.2 The Inoculation Chamber 41 4.2.3 Observations 43 4.2.4 Data analysis . 44 4.3 Results . 47 4.3.1 Description of Early Symptoms 47 4.3.2 Frequency of symptoms among families 52 4.3.3 Symptoms and gall formation at the family level 54 4.3.4 Early symptoms and infection of individual seedlings 58 4.3.5 Variation in Gall Formation 60 4.4 Discussion 64 4.4.1 Symptoms 64 4.4.2 Variation in Gall Formation . 65 5 P I N E C L O N E S A N D SINGLE-GALL S P O R E SOURCES 68 5.1 Introduction and Literature Review 68 5.2 Materials and Methods 69 5.2.1 ' Sources and Preparation of the Clones . . 69 v 5.2.2 Spore collection . . 70 5.2.3 Inoculation Technique 70 5.2.4 Collection of Data and Analyses 71 5.2.5 Electrophoretic Variation among Single Galls 72 5.3 Results 72 5.4 Discussion 77 6 V A R I A T I O N A M O N G OPEN-POLLINATED P I N E FAMILIES 80 6.1 Introduction 80 6.2 Literature Review 81 6.3 Materials and Methods 82 6.3.1 Preparation of Seedlings 82 6.3.2 Collection of Spores 83 6.3.3 Inoculation of Seedlings 83 6.3.4 Observations and Analysis of Data 84 6.3.5 Estimates of Narrow Sense Heritability (h 2) 85 6.4 Results 87 6.5 Discussion . 94 6.5.1 The Inoculation Technique 94 6.5.2 Variation among the 40 Open-pollinated Families 94 7 I N H E R I T A N C E OF R E S I S T A N C E : A D I A L L E L CROSS 97 7.1 Introduction 97 7.2 Materials and Methods 99 7.2.1 Preparation of Seedlings 99 7.2.2 Spore Collection and Inoculation 99 7.2.3 Analysis of Data 101 vi 7.3 Results 103 7.4 Discussion 109 8 CONCLUSIONS AND GENERAL DISCUSSION 111 8.1 General Conclusions I l l 8.2 Discussion . 112 8.2.1 Methodology '. 112 8.2.2 The Pathogen - Nuclear Cycle and Genetic Variability 114 8.2.3 The Host 115 8.2.4 The Mode of Inheritance - The core of stability 116 8.2.5 Pathosystem' Stability 118 Bibliography 119 APPENDICES 135 A Experimental Materials used in the Study 135 A.l Stand A 135 A.2 Stand B 137 A.3 Stand C 139 B Analyses of Variance of the Frequencies of Early Symptoms 140 C Resistance Frequency Distributions of 10 Pine Families 147 D A Diallel 150 E Electrophoretic Analysis of Rust Collections 154 E.l Results ." 154 vii Discussion and Conclusions ix List of Tables 4.1 Expected mean square table for ANOVA of early symptom frequency among 10 open-pollinated lodgepole pine families inoculated with west-ern gall rust at two spore loads and two times of inoculation . 46 4.2 Summary of ANOVA of the frequencies of three symptom types produced on lodgepole pine seedlings following inoculations with western gall rust. Analyses were done separately for each symptom type at 2, 4 and 8 weeks following inoculation 53 4.3 Results of analysis of arcsine-square root of % of seedlings of open-pollinated lodgepole pine families showing red flecks, 4 weeks following inoculations with two spore loads of western gall rust 53 4.4 Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the first stage of seedling maturity and high spore load (1.0g/250 seedlings) 55 4.5 Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the second stage of seedling maturity and low spore load (0.1g/250 seedlings) 55 x 4.6 Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the second stage of seedling maturity and high spore load (1.0g/250 seedlings) 56 4.7 Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the second stage of seedling maturity and low spore load (0.1g/250 seedlings) 56 4.8 Correlations between the occurrence of individual symptom types and gall formation (percent infected per family) on lodgepole pine seedlings fol-lowing inoculations with western gall rust. This table presents results of analysis of seedlings at the first stage of maturity (AGE 1) and the en-tries; r values are stratified by spore load and time of observation. Time 4 represents symptom types at the time when the first galls were recorded. r d / = 8 critical=0.6319, p=0.05 57 4.9 Correlations between the occurrence of individual symptom types and gall formation (percent infected per family) on lodgepole pine seedlings fol-lowing inoculations with western gall rust. This table presents results of analysis of seedlings at the second stage of maturity (AGE 2) and the en-tries; r values are stratified by spore load and time of observation. Time 4 represents symptom types at the time when the first galls were recorded. r d / = 8 critical=0.6319, p=0.05 57 4.10 Table of X{df=s) values relating the presence of symptoms and percent infection caused by western gall rust on lodgepole pine. The values were calculated for the 3 separate times of observation and the 2 spore loads. The X( d / =3) PO.OB =7.82 58 x i 4.11 Overall % tests relating the presence of symptoms and gall formation among lodgepole pine seedlings inoculated with western gall rust at the first (AGE 1) of the two stages of seedling maturity that were used in the experiment. The two spore loads and the three separate times of observa-tion were pooled 4.12 Overall x tests relating the presence of symptoms and gall formation among lodgepole pine seedlings inoculated with western gall rust at the second (AGE 2) of the two stages of seedling maturity that were used in the experiment. The two spore loads and the three separate times of observation were pooled 4.13 Results of ANOVA of arcsine-percent infected of 10 open-pollinated lodge-pole pine families inoculated with western gall rust at two stages of seedling maturity and two spore loads 4.14 Results of ANOVA of galls per seedling (square-root transformed) from 10 open-pollinated lodgepole pine families infected with western gall rust at two stages of seedling maturity and two spore loads. Shoot was length analysed as a covariate 4.15 Summary of regressions of offspring (lodgepole pine seedlings) on female parents. The offspring were inoculated with western gall rust at 2 stages of seedling maturity I and II; the independent variable was the number of galls on the female parent and the dependent variables were % infected and the mean number of galls per open-pollinated family. The two spore loads were combined xn 5.16 Results of ANOVA of the number of galls per shoot (adjusted for shoot length) on 16 lodgepole pine clones inoculated with four single gall spore sources of western gall rust collected on trees from the same stand. . . . 74 5.17 Average infection levels caused by four single gall spore sources of western gall rust on 16 lodgepole pine clones derived from two adjacent stands near Prince George 74 5.18 Infection summary for 16 lodgepole pine clones derived from two adjacent stands near Prince George, B. C., and inoculated with four single gall spore sources of western gall rust collected on trees from the same stands. The mean number of galls reported here for each spore source were not not adjusted for shoot length; the covariate 75 6.19 Expected mean square table for the ANOVA model on the 40 open polli-nated lodgepole pine families inoculated with two western gall rust spore sources; coastal and interior 84 6.20 ANOVA of the log of galls per seedling; shoot length analysed as a covariate and based on 40 open-pollinated lodgepole pine families inoculated with two western gall rust spores sources; coastal and interior 87 6.21 The relative susceptibility of 40 open-pollinated lodgepole pine families from the interior of British Columbia to coastal and interior spore sources. Susceptibilty was expressed as % infected and the number of galls per seedling 89 6.22 Summary of regressions of offspring on parents based on 40 open-pollinated lodgepole pine families inoculated with western gall rust from the coast of B. C. The dependent variables were; % offspring infected and mean number of galls per seedling 91 xin 6.23 Summary of regressions of offspring on parents based on 40 open-pollinated lodgepole pine families inoculated with western gall rust from the interior of B. C. The dependent variables were; % offspring infected and mean number of galls per seedling ,91 6.24 Estimates of Narrow Sense Heritabilities (h 2) from forty open-pollinated lodgepole pine families treated with two spore sources of western gall rust, one from the coast and the other from the interior of British Columbia. . 93 7.25 A Diallel Mating Table : number of seedlings per cross 100 7.26 Expected mean squares for an ANOVA of the number of galls per per cross from a lodgepole pine diallel which was inoculated spores of western gall rust 103 7.27 Infections (% infected, mean number of galls per family) caused by western gall rust spores used on the offspring of 4 open-pollinated lodgepole pine parents. The same parents were used to generate a 4 x 4 diallel 104 7.28 Duncan's multiple range test1. Comparisons of mean number of galls per cross from a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust 105 7.29 Percent galled per cross1 in a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust. Reciprocal pairs lumped to generate a half-diallel table 105 7.30 Mean number of galls per seedling for each cross in a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust 106 xiv 7.31 Combining ability analysis; table of ANOVA and variance components using number of galls per seedling (per cross) from a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust 107 7.32 A reduced form of table 7.31 showing mean squares and variance compo-nents for GCA, SCA effects (selfs excluded). Combining ability analysis using the number of galls per cross of a lodgepole pine diallel in which 1-year-old seedlings were inoculated with western gall rust 107 7.33 Estimates of Specific Combining Ability in a lodgepole pine diallel in which 1-year-old seedlings were inoculated with western gall rust 108 7.34 Individual GCA Estimates for the 4 parents in a lodgepole pine diallel in which 1-year-old seedlings were inoculated with western gall rust 108 A.35 A grouping by number of galls per tree (caused by western gall rust), of 120 lodgepole pine trees in Stand A (PG-I) and 102 trees from another 13 x 13 spacing trial (A1) 136 A.36 Lodgepole pine parent trees from Stand A (PG-I 1-12) near Prince George which produced some of the seedlings used in the studies reported here. The number of galls caused by western gall rust were counted in 1984 and the stand average, was about 10 galls 136 A.37 A table of lodgepole pine parent trees from Stand B (PG-II, III and IV) showing the respective number of western gall rust infections (galls) on each of the parents used in the studies of variability and resistance. The Stand average was 40 galls per tree 138 xv A. 38 A table of lodgepole pine parent trees from Stand C (Stump Lake) used an inoculation experiment in which a total of 40 families were inoculated with 2 spore sources of western gall rust. These parents came from a heavily infected stand 139 B. 39 Analysis of arcsine-square root of % of seedlings of open- pollinated lodge-pole pine families showing general red symptoms 2 weeks following inocu-lation with 2 spore loads of western gall rust 141 B.40 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red flecks 2 weeks following inoculation with 2 spore loads of western gall rust 141 B.41 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red streaks 2 weeks following inoculation with 2 spore loads of western gall rust 141 B.42 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing no symptoms 2 weeks following inoculation with 2 spore loads of western gall rust 142 B.43 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing general red symptoms 4 weeks following inocu-lation with 2 spore loads of western gall rust 142 B.44 Analysis of arcsine-square root of % of seedlings of op en-pollinated lodge-pole pine families showing red flecks 4 weeks following inoculation with 2 spore loads of western gall rust 142 B.45 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red streaks 4 weeks following inoculation with 2 spore loads of western gall rust 143 xvi B.46 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing no symptoms 4 weeks following inoculation with 2 spore loads of western gall rust 143 B.47 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing general red symptoms 8 weeks following inocu-lation with 2 spore loads of western gall rust 143 B.48 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red flecks 8 weeks following inoculation with 2 spore loads of western gall rust 144 B.49 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red streaks 8 weeks following inoculation with 2 spore loads of western gall rust' 144 B.50 Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing no symptoms 8 weeks following inoculation with 2 spore loads of western gall rust 144 B.51 Contingency tables based on the presence of early symptoms (stains) and gall formation on seedlings from 10 open-pollinated lodgepole pine families inoculated with western gall rust applied at 2 spore loads, and at the 1*' of 2 stages of seedling maturity. The X2 values were computed for each of the 3 times that symptoms were observed and recorded 145 B.52 Contingency tables based on the presence of early symptoms (stains) and gall formation on seedlings from 10 open-pollinated lodgepole pine families inoculated with western gall rust applied at 2 spore loads, and at the 2nd of 2 stages of seedling maturity. The X2 values were computed for each of the 3 times that symptoms were observed and recorded 146 xvn D.53 Number of seedlings per inoculation run in which progenies of a diallel cross of lodgepole pine were inoculated with spores of western gall rust . xviii List of Figures 3.1 Aeciospores of western gall rust showing the number of nuclei per spore during spore maturation from the youngest (A) to the intermediate (B) and the oldest (C) spores in chains. Mature spores (D) are also shown. All spores were fixed in methanol, stained with DAPI and examined under fluorescent microscopy at x400 29 3.2 Aeciospores of western gall rust collected at the coast of British Columbia 30 3.3 Nuclei in mature spores and germ tubes of western gall rust stained with DAPI after various durations of incubation 31 3.4 Frequency histograms (i-viii) showing changes in the numbers and distri-bution of nuclei in germ tubes of western gall rust spores after 6, 8, 10, 12, 14, 24, 30 and 34 hours of incubation in the dark at 18 - 20°C . . . . 32 4.5 A schematic illustration (not drawn to scale) of the inoculation chamber showing the lid (1), evacuation tube (2), inlet tube (3 and 4) positioned directly above the spore containing beaker (5), clamp (6) and seedlings (7) 42 4.6 Seedlings of lodgepole pine showing general red stains, red flecks, red streaks and no symptoms following inoculations with western gall rust E. harknessii 49 4.7 Changes in the frequency of early symptoms during the 12 weeks following the first inoculation of lodgepole pine seedlings with two spore loads of western gall rust. The two spore loads were combined while the ten families (A-l, A-2, A-10, A-4, A-12; B-l, B-2, B-3, B-4, B-5) were shown separately. 50 xix 5.8 Mean number of galls per shoot for 16 lodgepole pine clones inoculated with 4 single gall spore sources of western gall rust. All clones and spore sources came from two adjacent stands near Prince George. The clones were arranged from left to right along the X axis starting with the highest to the lowest average infection levels. Each line in the graph joins points representing the response (unadjusted mean galls per shoot) of each of the 16 pine clones to a single spore source 76 6.9 Plots of mean number of galls per family (adjusted for shoot length) for 40 open-pollinated lodgepole pine families inoculated with two geographic spore sources of western gall rust; one from the coast and the other from the interior of British Columbia 88 6.10 Frequency histograms illustrating the distribution of lodgepole pine seedlings based on the number of galls per seedling, following the inoculation of 40 open-pollinated lodgepole pine families with two spore sources of western gall rust; from the coast and from the interior of British Columbia. . . . 92 C. l l (1-10).Frequency histograms of 10 open-pollinated lodgepole pine families inoculated with 2 spore loads of western gall rust at 2 stages of seedling maturity. Seedlings treated with both spore loads were combined and each family was described by the distribution of its members by infection classes 147 D. 12 (a - r). Frequency histograms describing parents (a - d) and progenies of a diallel cross (e - r) of lodgepole pine by infection classes (galls per seedling), following inoculations with spores of western gall rust 151 xx E.13 SDS-PAGE profiles ( run on 10% gels) of single-gall spore sources of west-ern gall rust collected on *lodgepole pine from two neighbouring stands near Prince George in British Columbia. The spores were used in an in-oculation experiment which had 16 *lodgepole pine clones. The profile of the 4th spore source was not included 157 E.14 SDS-PAGE of single gall spores of western gall rust collected from *lodge-pole pine trees growing in stands in Prince George, Lighthouse Park (coast) and Richmond (coast) in British Columbia 158 E.15 SDS-PAGE profiles comparing single gall spores of western gall rust grow-ing on lodgepole pine with 1 source collected from scotch pine 159 E.16 SDS-PAGE profiles comparing single gall spores collections of western gall rust growing on lodgepole pine with 2 single spore sources collected from scotch pine 160 xxi Acknowledgements I wish to acknowledge the following institutions for their contributions to this study. The Government of Kenya provided me with a fellowship through the entire course of the study and the National Science and Engineering Research Council of Canada (NSERC) funded my research. The Department of Forest Sciences, University of British Columbia gave me valuable financial assistance in the form of teaching and research assistantships. This study is truly a tribute to their support for education. My deepest appreciation goes to Dr. B. J. van der Kamp who suggested the study and under whose guidance the study was undertaken. His patience, kindness, academic and financial assistance were an immense support during the course of the study. I am also grateful to the members of my supervisory commitee; Drs. R. J. Bandoni, R. J. Copeman, Y. El Kassaby and D. T. Lester who gave valuable advice during the study and for their constructive criticism of the manuscript. Dr. M. Shaw and Mr. Leroy Scrubb permitted me to use their laboratory for electrophoretic work and also gave considerable technical assistance for which I am really grateful. The useful comments and assistance given by Dr. Judy Loo-Dinkins are also acknowledged. This study is dedicated to my parents; Rhoda and Walter Ojwang who instilled the value of education in me from my childhood, and also to my wife Annastancia, and our children Auma and Omondi whose love, encouragement and endurance made the study a worthy effort. xxn Chapter 1 G E N E R A L I N T R O D U C T I O N Lodgepole pine (Pinus contoria Dougl.) is an important timber species in central British Columbia. It is also the most widely distributed conifer in western Canada and the United States and the main native tree species on over 26 million hectares of forest land in the region (Wheeler and Critchfield, 1985). It is also becoming an important timber species well outside its natural range in places such as Sweden and Norway. A major disease of lodgepole pine which causes losses in nurseries and in planted and natural stands, is a stem rust commonly known as western gall rust caused by the fungus Endocronartium harknessii (Moore) Hiratsuka. This rust occurs all over the natural range of lodgepole pine and probably exists as local endemic pathogen populations. Infections take place on the young succulent current year's shoots in the spring. Globose woody galls form on the shoots within 3 to 24 months depending on the physiological age of pine at the time infection. Infections on the main stem cause death in seedlings and young trees and mechanically weaken stems in older trees, often leading to wind breakage at the gall. Branch mortality is another form of damage on older trees, reducing photosynthetic area, and possibly providing infection courts for agents of heartwood decay. Since heavily infected trees may have lower cone yields and may also die, the rust can shape the genetic composition of stands in later generations. Since lodgepole pine regenerates naturally after forest fires into often overly dense stands, the rust may act as a natural thinning agent. However, "the main concern in British Columbia and elsewhere, is for the more intensively managed and well-spaced planted and natural stands. From such stands we 1 Chapter 1. GENERAL INTRODUCTION 2 can ill afford the unplanned loss of stems. The search for resistance therefore becomes inevitable not only in lodgepole pine, but also in other hosts. The rust parasitizes other hard pines such as jack pine (Pinus banksiana Lamb.) and poses a special threat to Monterey pine (P. radiata L.), which is widely planted as an exotic species in the tropical and subtropical regions of Africa, South East Asia, Chile and New Zealand. The lodgepole pine-gall rust pathosystem in British Columbia is natural. Such sys-tems can be used to study the natural variability in host resistance and how natural populations maintain their resident pathogen populations at endemic levels. Variability in the relative susceptibility to western gall rust has been observed between provenances, families and individual trees within stands in British Columbia (van der Kamp 1981, Martinsson 1980). In both natural and planted stands, trees of the same age and stand-ing side by side can show marked differences in the numbers of natural infections. Within the same stands, frequency distributions of trees tend to show the majority of trees with few or no infections while mortality levels are low and estimated to be between 5- 10% (van der Kamp 1981). The host and parasite seem to coexist with no recent indications of epidemic disease buildups. The pathosystem appears to be stable; a condition which may not obtain in artificial stands of the future unless the major factors controlling such stability are well understood and considered in subsequent selection and tree improve-ment practices. A major part of this stability should be attributable to the genetic make up of lodgepole pine populations with respect to resistance to the rust, as well as the genetic structure of the pathogen with respect to virulence. In addition to variability in susceptibility observed in natural pine populations, the development of early symptoms on inoculated seedlings and the possible relationship be-tween such symptoms and resistance is of interest. Some earlier work on fusiform rust (Cronartium quercuum (Berk) Miyabe ex Shirai /. sp. fusiforme ) on slash pine (Pinus el-liotii Engel.) seedlings in southern United States showed that some early symptoms might Chapter 1. GENERAL INTRODUCTION 3 be used to predict the resistance of inoculated seedlings (Miller et al 1976, Walkinshaw 1978, Lundquist and Luttrell 1982, Lundquist et al, 1982). Somewhat similar symptoms have been observed on lodgepole pine seedlings in our nursery and described elsewhere (Allen and Hiratsuka, 1985). These early symptoms, if reliable, could be important in research on western gall rust since their use could considerably reduce the time taken in the screening of seed sources for rust resistance. Chapter 4 describes the extent to which successful infection can be predicted from early symptoms. In order to understand the lodgepole pine-gall rust pathosystem, both the host and the pathogen must be considered. The fungus is a pine to pine rust with only one known spore type, and it has no known alternate host. We cannot therefore make crosses be-tween isolates using pycniospores (spermatia) and carry out standard genetic analyses of its traits. Hiratsuka et al (1966) proposed that karyogamy and meiosis occur in the aeciospores at germination. In this interpretation, the aeciospores behave like teliospores of other well known rusts (Hiratsuka et al, 1966). Since the spores arise from monokary-otic mycelia (True 1938, Hiratsuka et al, 1966), it is not clear how binucleate spores arise without pycnia. If there is no spermatization by pycniospores, then other means of nuclear transfer such as somatogamy can be postulated. Recently, the occurrence of meiosis at spore germination has been questioned (Epstein and Buurlage 1988). As has been shown in other fungi including basidiomycetes, natural processes such as het-erokaryosis and pansexuality (Burnett 1975, Ross 1979) are important sources of genetic variability and can be assumed to play a major role in the absence of a sexual phase. Chapter 3 describes the cytology of immature and germinating spores from coastal B. C. gall rust population. It is possible to demonstrate genetic variability in the rust using techniques such as gel electrophoresis of total proteins, protein fractions and of isozymes. This can be done without knowledge of how the observed variations came about. In this study, the Chapter 1. GENERAL INTRODUCTION 4 electrophoresis of total spore proteins, inoculations of seedlings using single-gall and bulked spore collections (Chapters 5 and 6) were used to demonstrate variation in the rust. The single-gall and bulked spore collections were made from lodgepole pine stands in the interior and the coast of British Columbia. The electrophoretic analysis showed some potential as a method in demonstrating variation in the rust but still needs further improvement. The results are presented in the appendix (Appendix E). Variation in resistance of the pine host was studied by artificial inoculation of open-pollinated families (Chapters 4 and 6), controlled-pollinated families (Chapter 7) and grafted clones (Chapter 5). These studies provided information on the inheritance of resistance. Chapter 2 L I T E R A T U R E REVIEW: N A T U R A L P A T H O S Y S T E M S 2.1 General Even though there have been relatively few extensive studies of natural pathosystems, plant breeders, geneticists, pathologists and laymen recognize the genetic diversity of natural plant ecosystems in which the destructive effects of pathogens are buffered. These ecosystems are therefore the living examples of natural stability in the coexistence of hosts and their parasites. Most of the work has been on conceptual models which mainly attempt to explain the evolution of stable pathosystems while some explain how and why certain types of genes in the hosts and pathogens would confer stability in certain pathosystems. A natural pathosystem must be understood as a system whose behaviour follows certain patterns. These patterns are a result of a set of mechanisms which have a genetic base. The patterns of behavior can be discerned and described whereas the mechanisms which underly the patterns can be discovered by experiment. Such mechanisms are usually described using sub-systems within a system. For example, the variability among and the inheritance of the genes for resistance in a host has been considered in the theory and development of multiline cultivars. Multilines have been advocated partly because they are thought to allow the operation of selection mechanisms that stabilize and keep pathogen races simple (Frey et al, 1977). 5 Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 6 A natural pathosystem such as that of lodgepole pine and western gall rust is con-sidered stable, at least to the extent that both the host and parasite have survived. It follows that such a system has developed patterns of behaviour which make it stable. Such patterns may be seen as changes in the frequencies of host genotypes over time with corresponding changes in the race composition of a parasite through selection to maintain stability. Such patterns of behaviour are numerous and may differ from system to system depending on the organisms involved and the physical and biotic environments. What unifies these systems is that certain rules or patterns of behavior which bring about stability are followed. A pathosystem or any other system which follows such rules devel-oped over time by evolution and selection is said to have an evolutionary stable strategy (ESS) (Maynard Smith and Price, 1973). The genetic basis of host-parasite interactions leading to stability in natural pathosystems in a physical and biotic environment is what man seeks to explore and understand. Most agricultural systems are not stable in the sense described above but they still survive because of the intervention of man. In forest systems with life spans much longer than those of most agricultural crops, such intervention would be quite expensive. For that reason, interfering with a natural forest pathosystem may pose problems which may be quite expensive to deal with. We must therefore describe the stability of such natural pathosystems and their underlying mechanisms. Plant population biologists, pathologists and geneticists have attempted to explain the observed natural stabilities using principles of host- parasite genetics in coevolved systems (Browning 1974 and 1979, Leonard and Czochor 1980, Mettler and Gregg 1969, Nelson 1978, van der Plank 1963 and 1968). This includes theoretical models in which the genetic selection of host and pathogen genotypes are based on gene-for-gene systems (Jayakar 1970, Leonard 1969 and 1977). Two terms commonly used in the genetics of host-parasite interactions which were Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 7 coined by Van der Plank (1963), should be introduced at this time as they are invoked from time to time in this review and in the other sections. These are; vertical and horizontal resistance. Van der Plank (1963) defined vertical resistance as resistance by a host to some, but not all races of a pathogen whereas horizontal resistance was that which was expressed uniformly to all races of a pathogen. It follows that with a horizontal type of resistance, a host genotype with high resistance will remain so to all races of a pathogen and likewise; those with low resistance will maintain that to all races of a pathogen. This is what has been termed as "constant ranking" in a horizontal pathosystem. In vertical resistance, a host genotype may be highly susceptible to one pathogen race and quite resistant to another and so the ranking of host genotypes will depend on the pathogen race used. Since vertical resistance in a host enables a host line to discriminate amongst incoming pathogen races, it was also termed as a race-specific type of resistance by Van der Plank (1963). Furthermore, this type of resistance was postulated to operate on a gene-for-gene basis. The gene-for-gene relationship was conceptualized and demonstrated by Flor (1942) during his study of flax rust. The concept states that, for every gene conferring resistance in a host genotype, there is a corresponding gene that confers virulence to a pathogen genotype. Certain host genotypes would then be attacked only by those pathogen genotypes which carry specific virulence genes designed to match them. This kind of relationship would maintain a dynamic equilibrium between resistance genes in the host and virulence genes in the pathogen in a pathosystem. In artificial agricultural systems, the gene-for-gene relationship can influence the epidemiology of a disease within a growing season. Because of the matching of compatible host and pathogen genotypes in this type of interaction, the first set of inoculum is diluted since only a part will be virulent on available host genotypes. In epidemiological terms, the initial rate of infection is reduced and this delays an epidemic. On the other hand, horizontal resistance is Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 8 race non-specific and does not operate on a gene-for-gene basis. Host genotypes do not discriminate among incoming pathogen races but resistance is expressed phenotypically as reduced rates of tissue colonization, longer latent periods, reduced sporulation, slow rusting and so on. Epidemiologically, this type of resistance reduces the rate of disease spread within a host population. Furthermore, vertical resistance seems to be associated with single or a few genes with large phenotypic effects (mono- and oligogenic) while horizontal resistance, is assumed to be under the control of many genes (polygenic). These additional aspects of the two terms apart from the first definitions based on specificity of hosts and pathogen races have created a lot of controversy. This is because some single genes for resistance act in a manner which is rate reducing and therefore cannot by definition, be considered as conferring vertical resistance (Nelson 1978). Furthermore, gene expression depends on the genetic background into which a gene is introduced (Nelson 1978). The evolutionary significance of vertical and horizontal types of resistance are discussed further in the section on the evolution of natural pathosystems. This chapter reviews four aspects of disease in natural pathosystems. It starts with a section dealing with genetic diversity and stability of natural pathosystems. The second section describes selected natural pathosystems viewed as stable. Section three which is the major one describes and discusses the evolution of stable pathosystems. Section four discusses the genetic structures and possible host-parasite interactions behind some natural pathosystems like the ones mentioned in section two. Section five discusses man's attempts to control genetic diversity through the use of multiline cultivars. 2.2 Disease In Some Natural Pathosystems Two wild and coevolved pathosystems representing two important agricultural diseases were discussed by Nelson (1978). The first one involves Solarium demissum Lindl. which Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 9 is a wild relative of potato and the famous late blight fungus, Phytophthora infestans (Mont.) de Bary. The second one involves two wild grasses Tripsacum L. and Euchlaena mexicana Schrad. (Teosinte) which are related to corn (Zea mays L.) and the well known disease, northern leaf blight of maize caused by Exserohilum turcicum (Pass.) Leonard and Suggs. These two natural pathosystems occur in Mexico which is considered the geographic epicentre in which the hosts and these parasites have coevolved. In each example, the respective host sustains modest amounts of disease and the survival of host and parasite are not threatened. It is not at all clear how coevolution proceeded but the result is that each host and parasite have accumulated a substantial number of genes for resistance and virulence respectively. For example, some lines of S. demissum are immune to some isolates of the late blight fungus while some are not. This shows that genetic variability and some differential host-parasite interactions do exist. An even better studied mixture of wild pathosystems occurs in Israel which is in the general area normally known as the Fertile Crescent. This area is also considered the primary habitat of wild barley (Hordeum spontaneum L.) and wild oats (Avena sterilis L.) (Harlan and Zohary 1966). Some important pathogens of cultivated crops occur on these two wild grains. Examples are, crown rust (Puccinia coronata Corda) of oats, barley rust( P. hordei Fckl.), stem rust (P. graminis Pers.) and powdery mildew (Erysyphe graminis DC. ex Merat). Further readings can be found in Anikster and Wahl (1979), Browning (1974, 1979) Dinoor (1974) and Wahl (1970). For purposes of this review, the A.sterilis - P. coronata pathosystem has been chosen since it is one of the most studied of wild pathosystems. The crown rust fungus alternates between its grass hosts and buckthorn (Rhamnus palaestina L.); its perennial alternate host, in the hills of northern Israel . It is believed that buckthorn has contributed to the high genetic diversity of P. coronata (Wahl et al 1960) and also enables virulent and prevalent races of the rust to survive adverse weather conditions. Dinoor (1974) did race surveys on the rust on Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 10 both buckthorn and oats and found that the rust was more diverse on buckthorn than on oats. In this part of Israel, one finds oats with and without crown rust. Some plants are also infected with E. graminis but are still green and vigorous while others show systemic infections with oat loose smut (Ustilago avenae (P.) Jensen). Even some plants having stem rust remain green and productive. Browning (1979) pointed out that it is surprising that genes for high protein, high oil content, high growth and yields and even high resistance have been extracted over the years from this wild pathosystem despite the prevalence of disease. In artificial crop systems, disease development tends to reduce crop quality and yield. A much less studied pathosystem is the lodgepole pine-gall rust pathosystem. The system appears to be stable. In most stands, most trees have just a few or no infections. Resistance frequency distribution histograms tend to depict a general skew towards high resistance. The modest mortality estimates of 5-10% in natural stands suggest yet unex-plained forces which tend to limit the rust. Although the fungus apparently reproduces asexually, it can be assumed that other mechanisms exist for its genetic recombination. This would enable it to survive on a chiefly outcrossing and therefore variable lodgepole pine host. Such mechanisms occur in a number of plant pathogenic and other fungi (Day 1974). However, asexually propagating pathogen populations may have relatively fewer genotypes than sexually reproducing ones (Leonard 1977) because the best genotypes are selected but not broken and recombined as frequently. We can assume that the pine host and the rust have coevolved and that both parties have a wealth of genetic resources which stabilizes their coexistence. Our main interest is to elucidate the basic genetics of host-parasite interactions operating in these systems. Since the concepts of horizontal and vertical resistance have been illustrated mainly in agricultural systems, it is impor-tant that their relative occurrences and roles be estimated within natural pathosystems Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 11 starting from a relatively small unit like a stand. Some examples of the genetic interac-tions in a natural pathosystem are discussed in section four of this chapter. 2.3 The Evolution of Stable Natural Pathosystems 2.3.1 General Theory Van der Plank (1968) proposed that genes for resistance in a host and similarly genes for virulence in a pathogen are selected against if they are not needed. A classic example of this was provided by maize rust caused by Puccinia polysora Underw. which devastated maize in Africa. Apparently, maize cultivation had gone on for decades and spread all over Africa without the rust. Then the rust, a native and minor pathogen of maize in Mexico was introduced. It was concluded that maize lines had lost their resistance to the rust. This suggested a possible general occurrence of "selection against unnecessary genes for resistance or virulence". This concept provided the basis for many models of balanced host-parasite systems discussed in the next section. Furthermore, Van der Plank (1968) proposed two other concepts; "stabilizing " and "directional" forces of selection to explain how host-parasite populations might evolve on a gene-for-gene basis. He defined "directional" selection as that favouring genes for virulence, that is, selection imposed on a pathogen by the use of resistant host cultivars and "stabilizing" selection as that against unnecessary genes for virulence on susceptible hosts. Leonard and Czochor (1980) and Crill (1977) prefer the use of the term "stabilizing" selection in its original meaning in quantitative traits in population biology. This definition could also be used to describe changes in qualitative traits (Mettler and Gregg 1969) and stabilizing selection would mean opposing forces of selection maintaining alleles at a given locus at stable equilibrium frequencies in a population. With regard to virulence genes, stabilizing selection would represent a combination of forces of selection that reduce the frequency Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 12 of a gene if it is too high and increase it, if it is too low (Leonard and Czochor 1980). In this sense, directional selection would be used to describe forces of selection arising from changes in the environment to force a population to a new equilibrium. The concept of "fitness" of a host and a pathogen genotype is important here because selection results from differences in fitness among individuals of a population. Fitness is defined here as "the relative reproductive success of an individual or individuals; those that survive best and pass on their genes to the greatest number of offspring are the most fit" (Leonard 1977). As a result, genes for superior fitness in any population tend to increase in frequency in that population. The concept of fitness has also been im-portant in the modelling of balanced host-parasite systems. The question as to whether host-parasite populations in gene-for-gene systems evolve as "balanced" or "transient" polymorphisms are not important to this study but have been discussed (Leonard 1977, Leonard and Czochor 1980, Nelson 1978). 2.3.2 Stability Strategies of Vertical and Horizontal Pathosystems After defining horizontal and vertical pathosystems, we should explain how they might have evolved and their respective roles in an evolutionarily stable strategy. In this respect, Robinson (1981) used the terms alloinfection and autoinfection. An infection of host line by a pathogen race in a gene-for-gene system is an alloinfection while infection of the same host by inoculum produced on it is autoinfection. Robinson further concluded that once a pathogen has established itself in a crop by alloinfection, it can only be controlled by the rate reducing effects of horizontal resistance. In other words; vertical resistance controls alloinfection while horizontal resistance controls autoinfection. Robinson (1981) has also discussed the ESS of vertical and horizontal pathosystems. The details are not discussed here but he concluded that vertical pathosystems should evolve exclusively in crops which have both spatial and sequential discontinuity like Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 13 annual crops and deciduous trees. This is because in annual plants, fresh seedlings from vertically protected parents of the previous season will have to be alloinfected and will thus be protected by the vertical genes they carry. This is what Robinson has termed " the recovery of vertical resistance". Though vertical pathosystems have mainly been shown in annual cereal crops, there is some evidence of vertical genes in forest pathosystems such as the white pine-blister rust pathosystems. This is explained by Robinson as caused by the spatial and sequential discontinuity in the alternate hosts of the rust. The ESS of horizontal pathosystems are seen as responses to positive and negative selection pressures on both hosts and pathogen to maintain stability. When these opposing selection pressures are equal, the system is stable. Robinson (1981) has speculated that vertical pathosystems are necessary buffers to a host against decimation by parasites at times when seasonal weather conditions favour parasites. For example, weather conditions supporting an abnormally high growth rate in a parasite would be controlled by a gene-for-gene relationship since hosts with high specific resistance may survive better or delay an epidemic and possibly save a host from decimation. Under most disease conditions, the above argument would suggest that a mixture of vertical and horizontal subsystems in a pathosystem should be a good ESS since genetic complexity and a phenomenon such as host tolerance associated with horizontal pathosystems seem to play a role in endemic levels of disease in natural pop-ulations. 2.3.3 Theoretical Models A number of theoretical models have been developed which attempt to depict how sta-ble pathosystems may evolve ( Jayakar 1970, Leonard 1969, Mode 1958, Mode 1961, Leonard 1977). Most of these models are based on the selection of both host and par-asite genotypes interacting on a gene-for-gene basis. In Person et al (1980), the two Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 14 models are based on polygenic host-pathogen interactions. Three of the models (Jayakar 1970, Leonard 1977 and Person et al 1980) mentioned above are briefly discussed below. The models by Jayakar (1970) and Leonard (1969 and 1977) are somewhat similar because they both are based on gene-for-gene interactions and have also recognized the fact that super resistant hosts or super virulent pathogen genotypes will dominate natural populations unless there are costs associated with resistance or virulence. In the model by Jayakar (1970) both the host and pathogen are haploid and have single genes for resistance or virulence such that; the locus in the host is either R (resistance) or r (susceptibility) and the locus in the pathogen is either V (avirulence) or v (virulent). These genes occur with given frequencies in the population; for example r and V occur at frequencies of pt and p2 respectively. A susceptible host dies upon infection and does not reproduce while the pathogen produces a given number of offspring. Likewise avirulent pathogens on resistant hosts will not reproduce. Given a particular probability that a susceptible host is infected, the proportions of host and pathogen genotypes surviving after one generation can be calculated. It turns out from this model that unless there are fitness costs attached to resistance and virulence, the most virulent pathogen genotypes will dominate. With fitness costs incorporated, balanced polymorphisms can be attained. Leonard's (1977) model illustrates interactions between a diploid host and a haploid pathogen. He assigned a fitness cost for interactions between avirulent pathogens and a susceptible host and this cost represents the degree to which a susceptible host limits sporulation of a pathogen. In a combination of a virulent pathogen and a susceptible host, the fitness of the pathogen would be further reduced from the value for the avirulent-susceptible combination and this would be the cost for virulence. He also assigned fitness costs associated with resistance and susceptibility in the host. With these costs, further mathematical analyses which are beyond the scope of this review show that a balanced polymorphism can be attained. Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 15 Person et al (1980) contend that plant populations show "field resistance " which appears to be evolutionary stable and most likely polygenically controlled. Two theoret-ical models were put forward with two major assumptions. These were that; resistance and pathogenicity are both non- specific and under polygenic control, and for the two interacting populations, variability is continuous and of the type described by a normal curve. The first is an additive model in which they show mathematically that, realized levels of disease would also conform to a normal distribution. It is also assumed that, a shift in the mean of any of the interacting populations; pathogen or host, will result in a relatively smaller shift in the mean of the curve representing realized disease develop-ment. The second model which is a multiplicative one shows a realized host frequency distribution skewed toward host resistance. They (Person et al 1980) concluded that polygenically controlled systems should show constant ranking and should be relatively stable provided that, resistance in the host correlates positively with reproductivity. It is interesting that field observations and experiments with certain host-parasite systems have shown either normal or skewed frequency distributions (Segal et al 1980, Dinoor and Eshed 1984, Burdon 1980a and 1980b). In some instances, different samples from a single host species inoculated with a pathogen will show both types of distributions and more (Dinoor and Eshed 1980). Person et al (1980) did not discuss what the two distributions are likely to indicate especially if both are to be associated with stability of a pathosystem. 2.3.4 Resistance Frequency Distributions among Families Derived from Wild Populations Segal et al (1980) described distribution frequency histograms for the reactions of Avena sterilis to crown rust and of Hordeum spontaneum to powdery mildew. The distribu-tions were normal and skewed respectively. Burdon (1980a) analyzed clones of Trifolium Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 16 repens L. for their response to two diseases. For one disease (Black blotch) caused by Cymadothea trifolii (Pers. ex Fr.), the distribution was normal while for leaf spot (Pseu-dopeziza trifolii) (Biv.) Fckl., the distribution was skewed towards greater resistance. Burdon's (1980a) hypothesis was that these distributions were related to selection pres-sures exerted by pathogens upon their hosts and suggested that high selection pressures produce distributions skewed toward high resistance whereas mild selection pressures pro-duce normal distributions. Dinoor and Eshed (1984) tested H. spontaneum, an annual grass, with a powdery mildew fungus and Limonium siniatum, a perennial wild flower with a rust. The Hordeum system responded variably to the ten mildew cultures used. Some were normal, others skewed toward susceptibility while a third type showed normal distributions but also with a separate class of resistant individuals. For the Limonium system, the amount of disease as well as the frequency distribution curves changed over the season. Over the entire year, the frequency distributions were skewed towards resis-tance. Dinoor and Eshed (1984) argued contrary to Burdon (1980a) that factors other than selection pressures from the pathogen alone may be operating. Severe selection pressure may induce host tolerance and enable a host to react with a normal resistance distribution. This implies that a skewed distribution may also indicate a stable mix-ture of high resistance, tolerance and susceptibility. It is also possible that the level of a pathogen's virulence or aggressiveness interacts closely with environmental conditions which affect infection. Unfavorable conditions can induce increased pathogen aggressive-ness and such a condition will mask the normal resistance spectrum of a host population. In all the above experiments and observations, it was not clear whether the pathogen isolates used differed in virulence. Also, the inoculations did not consider the use of various spore loads. It may well be that frequency distributions depend on inoculum loads and favourable infection conditions in addition to the distribution of resistance in the host. Burdon (1987) has further argued that both his views on selection pressures Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 17 and those of Dinoor and Eshed (1984) have merits. He recognizes four factors which are most likely to affect frequency distributions of infection types in a host population. These include; the history of pathogen attack mainly, frequency, severity and the racial composition of the pathogen. Another one is the genetic basis of resistance which has not been elucidated for most natural populations, and closely associated with this, is the cost of resistance which is partly responsible for balanced host genetic polymorphisms. The fourth one is the nature of the breeding system of the host. About this breeding sys-tem, Burdon argues that in an outcrossing species like T. repens, resistance will likely be selected independently of other characters because of regular genetic mixing. It may fol-low that differences in frequency distributions could be due to selective pressures exerted by different races on host genotypes. On the other hand, in highly selfing or inbreeding species like H. spontaneum and A. sterilis, co-adapted gene complexes are common and a strong selection for one particular aspect of resistance in such a complex may be reflected in short-term non-adaptive changes which in turn may be reflected in the skewness of resistance distribution. 2.4 Genetic diversity and host-pathogen interactions in some stable natural Pathosystems The wild oats and barley populations in Israel again provide examples of genetic diversity governing host responses and virulence of pathogens. The stability of the system was studied using the principles of qualitative genetics of individual host plants and pathogen races and the quantitative genetic aspects using the amounts and distributions of resis-tance in the host populations. The variation in the host was also reflected in the variation in the pathogen population (Anikster and Wahl 1979, Wahl et al 1960, Wahl 1970). In annual crown rust race surveys done over a span of about 20 years, race 276 in the most Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 18 virulent race group; 264 to 276 occured in about 40% of all collections (Wahl 1970). Sur-prisingly, race 202, which was one of the least virulent, was also one of the most common. In the same study, a line transect sampling of wild oat seed showed that only about 30% of the tested seed was resistant to race 276. It was concluded from this that the most virulent race did not dominate and the least virulent had not been eliminated. Browning (1974 and 1979) concluded that only 30% frequency of the host carrying genes for specific resistance to the most virulent race was adequate in combination with non-specific genes, for the protection of such a wild host population. It would appear that both horizontal and vertical types of resistance operate in this indigenous pathosystem. Another example of diversity in a natural pathosystem was examined by Burdon (1987) and shares striking similarities with the wild oat population described above. His experiments showed that there were complex patterns of interactions between Glycine canescens F. J. Herm.; a native Australian legume and Phakopsora pachyrhizi Syd.; a native rust pathogen. Individual host plants showed a diversity of reactions and half the tested individuals were resistant to all races of the rust while others had resistance effective against 11-18% of the races used. No individuals were susceptible to all the races nor were any rust races virulent on all host lines. Also in this pathosystem both virulent and avirulent pathogen races coexisted. Race 1 of P. pachyrhizi was virulent on only 4.5% of the host population while the most virulent; races, 5, 6 and 9 were each virulent on over 35% of the hosts. Furthermore it was illustrated that different isolates could detect only a portion of the different levels of resistance in a single host population. A further analysis of the genetic basis of this resistance revealed that some host lines were protected by one, two or three genes and that at least 12 different resistance genes exist in the population. A pathosystem somewhat similar to the lodgepole pine-gall rust system, is the pine-fusiform rust systems in southern United States where the major hosts are loblolly and Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 19 slash pines. Kinloch and Stonecypher (1969) looked at the variation in susceptibility among open and controlled pollinated progenies of pine parents chosen from wild stands from different sites. Disease severity frequency histograms show tremendous variation among parents within and between test sites. Most sites showed a normal distribution, some yielded distributions skewed towards higher resistance and one skewed towards higher susceptibility. The interesting point was that parent offspring correlations were high and consequently, heritability estimates were quite high (0.65-0.85). It was tenta-tively concluded that the pine population seemed to be segregating at several polymorphic loci based on the high heritabilities and the range and distribution of resistance. 2.5 Genetic Diversity and Stability in Crop Multilines A multiline as denned in modern day agriculture is a " mechanical mixture of plant varieties consisting of a number of phenotypically similar or compatible lines that differ mainly in the resistance genes they carry " (Marshall 1977). They are developed by recurrent backcrossing of chosen host lines to a common parent with certain desired agronomic traits. The type of genes composing a multiline and how often these genes are deployed may differ. Many have argued for the so called "clean crop" approach in which the individual component lines are completely resistant to all the prevalent pathogen races to be controlled (Borlaug 1959). Should the resistance of one component break down, it is replaced by a resistant alternative. The danger is that this approach simultaneously exposes to the pathogen a number of available host genetic resources without consideration of the fact that the genetic resource base may be limited, at least in the short run. Frey et al (1977) called for the so called "dirty crop" approach in which the component lines carry unique single genes for resistance (Marshall and Pryor 1978, 1979), but none of the lines is resistant to all the prevalent pathogen races. Some Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 20 lines will act as non-hosts and reduce the amount of effective initial inoculum and thus reduce the rate of intracrop disease buildup. Frey et al (1977) believe that the dirty crop approach will stabilize the race structure of a pathogen's population with respect to the number of and the relative frequencies of component pathogen races. A number of races will survive and since each host line carries a single or maybe, just a few genes for resistance, stabilizing selection should predominantly keep the pathogen races fairly simple. Eventually, pathogen populations might contain as many races as there are host lines. The hypothesis that pathogen races should remain simple is not yet supported by hard evidence and has therefore been under some criticism. Just one additional virulence gene could enable a pathogen race to attack more than one host line and most likely increase its selective advantage over simpler races (Marshall 1977). A pathogen's race composition may therefore be determined by the relative degrees of selection pressures toward greater virulence as more complex races exhibit virulence on more host lines. This would trigger a counteracting stabilizing selection against races with additional genes. Another possibility is that race complexity may result in the erosion or loss of tolerance of a multiline. Even though the hypothesis that multilines will keep races of a pathogen simple has not been demonstrated, practical questions like the number of host genotypes required to prevent the dominance by a super virulent race is of concern. Calculations (Groth and Person 1977, Marshal and Pryor 1978) have shown that the number of host genotypes will depend on the fitness cost (k) of unnecessary virulence genes. The lower the value of (k) , the higher the number of host genotypes required in a multiline mixture and the more expensive it becomes. The need for accuracy in calculating values of k is therefore important but many questions on how these values should be calculated remain (Burdon 1987). Though the advantages of multilines are recognized despite many pending Chapter 2. LITERATURE REVIEW: NATURAL PATHOSYSTEMS 21 questions, it is the yield capacity and crop uniformity of pureline monocultures that has been more appealing to mechanized agriculture. This has seriously limited the use of multiline cultivars. Chapter 3 C Y T O L O G Y OF D E V E L O P I N G A N D G E R M I N A T I N G SPORES 3.1 Introduction Field observations, inoculation studies and cytological work have demonstrated that E. harknessii produces spores on pine which reinfect pine directly without the requirement of an alternate host (Ouellette, 1965, Wagener, 1964, Zalasky and Riley, 1963, Hirat-suka et al, 1966). Anderson and French (1965) working in eastern Canada reported that, a rust fungus they had identified as Periderrnium harknessii J. P. Moore, infected Castilleja coccinae L. following artificial inoculations using spores produced on Pinus banksiana Lamb. However, inoculation of pine using basidiospores that were produced on C. coccinae was not attempted. No evidence of host alternation has been found in western Canada. Spermatia have occasionally been observed in the wild but they are quite rare and appear to have no essential function (van der Kamp, pers. comm., True 1938). Spermatia have not been observed on inoculated and closely watched nursery seedlings. True (1938) observed that the vegetative mycelia of E. harknessii were uninucleate and that the number of nuclei in germ tubes during host penetration was 1. He also observed that base cells (aeciophores) bearing spore chains could have two, three, and occasionally four nuclei while young aeciospores were usually binucleate and that nuclei tended to group in mature spores with no evidence of fusion. Hiratsuka et al (1966) using spore collections made in Alberta, reported that the 22 Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 23 young aeciospores were binucleate while mature spores were predominantly uninucleate. At germination, the nuclei usually divided twice and the germ tubes became septate and typically divided into four cells, each with one nucleus. These results differed from those reported by True (1938). Hiratsuka et al (1966) and, Hiratsuka et al (1969) concluded that cells of the septate germ tubes were products of meiosis and each represented a basidiospore. Heteroecious rusts develop tortuous non-septate germ tubes of indefinite length having two nuclei which migrate toward their tips (Hiratsuka et al, 1966 , Powell and Hiratsuka, 1969). Their germ tubes are also much longer than those of E. harknessii However, the data of Hiratsuka et al (1966) showed that the typical four- celled germ tube occurred at relatively low frequencies after 72 hours of incubation at 15°C. Only 20% were four-celled, 41% three-celled and 38% two-celled. This is not to say that one should have expected that all germ tubes be four-celled, but that for it to be termed typical, it should have been the majority under favourable laboratory conditions. Following these observations, the aeciospores were renamed aecidoid teliospores (spores which resemble aeciospores but function as teliospores of other rusts). It is on the basis of the above that Hiratsuka (1969) erected the genus Endocronartium to accommodate rust species related to those in the genus Cronartium but having an endo type nuclear cycle. Epstein and Buurlage (1988) using aeciospore collections made in California, stained mature dormant spores and germlings with a DNA- binding 4, 6-diamidino-2-phenylindole (DAPI) fluorochrome. This stain is more sensitive than the HCl-Giemsa stain that Hi-ratsuka et al (1966) had used. They (Epstein and Buurlage, 1988) reported that mature spores were predominantly binucleate (86%) which agreed with True (1938). The num-ber of nuclei in germ tubes increased gradually and asynchronously over time; giving numbers of nuclei such as three, five and seven. Epstein and Buurlage (1988) did not examinee nuclear events during spore development. They concluded that meiosis did not Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 24 occur at germination in their spore collections. The rarity or the absence of spermatia raises one major question. How the dikaryotic or more precisely, the binucleate spores observed by True (1938) and Hiratsuka et al (1966) are produced is the major question. Presumably, the fungus is homothallic. In the absence of meiosis, genetic shuffling would be through mitotic crossover coupled with the agents of heterokaryosis and pansexuality (Day 1974). Meiosis may not be critical to the maintenance of genetic variability of western gall rust. The available literature shows that alternative life cycles for this rust have not been looked at. This is partly because the fungus until recently, had not been successfully grown on artificial culture media (Allen et al, 1988). It follows that studies on hyphal anastomosis, somatogamy and other forms of nuclear transfer have not been done. It is quite possible that the fungus may rely entirely on such systems to achieve the genetic variability it needs to infect a number of its hard pine hosts. If that were the case, it may not require a sexual phase in its life cycle. Hiratsuka et al (1966) based their conclusion that meiosis occurs in the germinating aeciospore on two observations, namely, karyogamy (binucleate immature spores became uninucleate at maturity in his spore collection) and germ tube morphology. The forma-tion of "synaptinemal complexes", a cytological feature during meiosis was not examined. Synaptinemal complexes are visible only through a transmission electron microscope and are best seen at the pachytene stage of meiosis (Westergaard and von Wettstein 1972). Westergaard and von Wettstein (1972) also reported that, the synaptinemal complex in meiotic cells was the vector for chromosome pairing and crossing-over, and was necessary in the formation of chiasmata. The major difficulty as far as western gall rust is con-cerned, would be to select the right stage of spore germination for sectioning and viewing. Several samples of germinating spores would have to be taken over a range of incubation periods. Other than that, it is technically feasible and would provide conclusive evidence Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 25 of meiosis. Since it is possible that karyogamy may occur in spores during their formation and not in mature spores as Hiratsuka et al (1966) had reported, this study looked not only at mature spores, but also at immature aeciospores in chains. These young spores are found in galls at the base of aecia. In addition, DAPI was used to improve the resolution of nuclei within thick-walled aeciospores and also in germ tubes. The determination of the exact time of meiosis using this approach was considered essential before attempting a time-consuming transmission electron microscopy. The approach was to look for the evidence of karyogamy in the form of the reduction in the number of nuclei from two to one at some some stage of spore development; maturation or germination. The next nuclear division would presumably be a meiotic one. The plan was to look for synanptinemal complexes associated with this supposedly meiotic division. It turned out that evidence of karyogamy was not found. 3.2 Materials and Methods 3.2.1 Cytology of Spore Chains In the last week of April 1988, nursery seedlings which had been inoculated with coastal spores 2 years before and bearing young aecia were collected before the peridia had cracked. The galls were collected and the top mature spores were scraped off immediately with a scalpel to leave a thin mat of orange spores at the base of the aecia. Pieces of these mats; roughly 2 by 2 mm, were cut out with as little bark as possible and fixed in glass vials containing 20 ml of 70% ethanol. Fresh samples squashed with a cover slip and examined under the microscope showed spore chains connected by subjunctive cells. Many chains had about five to six spores each and showed aeciospores of different sizes and stages of maturation. Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 26 Spore chains were prepared for viewing under the microscope as follows. Samples were picked out and left to air dry for 5 minutes to allow excess alcohol to evaporate and chopped into smaller sections to enable squashing. Then 2 to 3 droplets of 4,6-diamidino-2-phenylindole (DAPI) were added and allowed to soak into the spore chains for 3 minutes. The DAPI had been prepared by dissolving 50/// of stock solution in 1 ml of 70% ethanol. The alcohol solvent provided better sample penetration by DAPI on our materials than normal distilled water. Cover slips were placed, the sample gently squashed and the edge of the cover slip sealed with cutex nail polish. Viewing was done under epifluorescence using a Zeiss universal microscope using a HBO 100 mercury lamp in conjunction with a BG-12 excitation filter. The number of nuclei in each spore was recorded in each of the 276 chains examined. 3.2.2 Cytology of Dormant and Germinating Spores Spore collections were made in the first week of May 1988 at Lighthouse Park in West Vancouver and at Richmond Nature Park. These two locations are about 20 kilometers apart. Since the latter is a peat bog and the former a wind swept rocky peninsula, they represent fairly different lodgepole pine habitats. Galls which had aecia still bound within yellow to orange and intact peridia were collected. The same day, the galls were brought to the laboratory and the peridia were carefully lifted off. Dry spores were released into a petri dish through a screen by a gentle agitation. A gentle agitation was necessary because only the mature and dry spores were required for this germination experiment. From each of the two locations, 3 galls were chosen and spores were extracted. The spores from each location were bulked and stored at temperatures of 3 — 5°C as two separate mixtures. A total of 1638 dry mature spores were fixed, stained with DAPI and examined. The number of nuclei per spore was recorded and the proportions of binucleate and Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 27 uninucleate spores computed. Only those spores whose nuclei were clearly stained were included while the ones which were in dense clumps or which did not bear fluorescing nuclei were ignored. The first germination experiment used fresh spores. The germination medium was 2% malt extract agar which had been thinly poured (1.5 to 2 mm) at the bottoms of standard petri dishes. Spores were picked up using a paint brush and released onto petri dishes then incubated in the dark at 18 — 20°C. The incubation periods were 0, 6, 8, 10, 12, 14, 24, 30 and 34 hours. For each location, 2 agar plates were used so that each period of incubation had four plates. However, six to eight plates were used in the incubation periods between 8 and 14 hours since this was the period during which from 60 to 85% germination was achieved in earlier trials and was likely to include most of the nuclear division. After each period of incubation, the spores were fixed in formaldehyde vapour. This was done by pouring 3 ml of commercial formaldehyde into a petri dish lid and placing the samples on the agar face down above the formaldehyde. The dish was then sealed all around using masking tape and kept on a clean bench at room temperature for to 12 hours for the formaldehyde to vaporize and penetrate spore and germ tube walls. After this, the fixed samples were stored at 3 — 5°C to await staining. 3.2.3 Cytology of Germtubes The staining of nuclei was done according to the method used by Martin (1987). Fixed samples were taken out of the refrigerator and from each plate, four 10 by 15 mm agar blocks cut out. The blocks were placed on microscope slides and put on a dust free bench for 30 minutes. This allowed some drying, so that the ungerminated and germinated spores adhered better to the agar during staining. Two droplets of DAPI were dispensed from pasteur pipettes onto the samples and after 3 minutes, 10 by 22 mm cover slips were placed and pressed gently. Viewing proceeded Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 28 as described above at a magnification of X400. For each germinating spore, the number of nuclei including nuclei still in the spore bodies was recorded. For each incubation, selected germinating spores were photographed using Kodak Ectachrome 400 film. All photographs were taken at the same magnification of X400. 3.3 Results Of the spore chains examined, 2, 28, 15, 53, and 3 percent had 2, 3, 4, 5 and 6 spores respectively. 24% of all the spore chains examined had spores which were all uninucleate, whereas only 9 % of the chains were all binucleate. The remainder consisted of mixtures of uni, di and trinucleate spores in no particular order. The number of trinucleate spores was negligible. The younger spores in the chains ( Fig 3.1 and 3.2) were not predominantly binucleate and there was no clear progressive reduction in the number of nuclei from the young to the mature spores (Fig 3.1.) Of the mature spores, 41% were binucleate, 57.5 % were uninucleate and 1.5% were trinucleate. Spore germination on 2% malt extract agar began within 4 hours (Fig 3.3) and the percentage of germination increased up to 14 hours. Germination was 15% after 4 hours, 50% after 8 hours and reached a maximum (85-90%) after 12 to 13 hours of incubation. After 6 hours, the average number of nuclei in germ tubes was more than double the number in mature spores. The number of nuclei continued to increase gradually until 34 hours (Fig 3.4 a and b). The mode occurred at 4 nuclei between 8 to 14 hours, shifted to 5 at 24 and 30 hours and was 6 at 34 hours. Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 29 70 60 50 A B D c CD 2. 20 OJ £ 10 0 2 3 1 2 3 . 1 2 3 Number of nuclei per spore 1 § 1 1 IJLJ 2 3 Figure 3.1: Aeciospores of western gall rust showing the number of nuclei per spore during spore maturation from the youngest (A) to the intermediate (B) and the oldest (C) spores in chains. Mature spores (D) are also shown. All spores were fixed in methanol, stained with DAPI and examined under fluorescent microscopy at x400 Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 30 Figure 3.2: Aeciospores of western gall rust collected at the coast of British Columbia A and B. Aeciospore chains showing spores at different stages of maturation; C, a germ tube showing a septum near its base, and D, Mature binucleate aeciospores stained with DAPI and examined under UV light. Magnification A'400 Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES Figure 3.3: Nuclei in mature spores and germ tubes of western gall rust stained with DAPI after various durations of incubation A, a binucleate spore showing a nucleus migrating into a germ tube within 4 hrs of incubation, B, a young germ tube (after 6 hrs) with 2 nuclei and the spore has 2 nuclei as well, C a 5 nuclei germ tube (10 hrs), D an unusual short and branched germ tube (4 nuclei after 4 hrs), E, 7 nuclei (12 hrs); F, 8 nuclei (14 hrs) the 8"1 is out of focus. Magnification A'400. Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 32 Figure 3.4: Frequency histograms (i-viii) showing changes in the numbers and distribution of nuclei in germ tubes of western gall rust spores after 6, 8, 10, 12, 14, 24, 30 and 34 hours of incubation in the dark at 18 — 20°C CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 33 After 6 hours ol incubation (n=108) After 8 hours of incubation (n=118) After 10 hours of incubation (n=147) After 12 hours of incubation (n=232) After 14 hours of incubation (n=418) After 24 hours of incubation (n=207) V TT. 7f-• VI After 30 hours of incubation (n=168) After 34 hours of incubation (n=262) Vlll 1 L I J J jlio 3 4 5 6 7 8 Number of nuclei In germ lubes t 2 3 4 5 e 7 e Number ol nuclei in germ lubes Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 34 3.4 Discussion The demonstration of sexual division at aeciospore germination requires (1) evidence of karyogamy at some stage during maturation or germination, and (2) evidence of meiotic division such as the regular division of the diploid nucleus into two and then four haploid nuclei. At later stages, further nuclear division (mitotic) may occur. With respect to karyogamy, this study shows firstly that more than half of the newly formed immature spores were uninucleate from the time they were formed. Unless karyo-gamy had taken place earlier on in the aeciophores rendering the nuclei diploid, meiosis would be impossible in such uninucleate spores. Secondly, at no stage during spore mat-uration or germination was there a significant reduction in the average number of nuclei per spore nor did the proportion of uninucleate spores ever exceed those found in the youngest immature spores. The proportion of mature spores that were binucleate (41%) was not significantly different from that in the oldest of the immature spores in the chains (35.9%). The evidence at hand does not support the interpretation of Hiratsuka et al (1966). With respect to meiotic division, there was no evidence in the data presented here of regular stepwise division from one to two to four nuclei per spore or germtube following germination. Instead, the number of nuclei increased gradually. However, the variation in the onset of germination between spores presents a difficulty. Germination started within 4 hours and was not complete until about 12 hours. Hence each histogram in Figure 3.4 represents a range of times since germination and at the same time show the gradual increase in the number of nuclei in the germ tubes. Thus in Figure 3.4b, some spores had developed for 10 hours since they germinated while others had grown for as little as 2 hours. Furthermore, the histograms (Fig. 3.4, i-iv) do not represent exactly the same spore population since only germinated spores were sampled. Thus, Figure Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 35 3.4 i describes only the early germinating spores while 3.4 vi represents all spores that germinated. It seems probable that under these conditions, any regular stepwise increase in the number of nuclei associated with meiosis would be masked. Nevertheless, if the second meiotic division occurs in synchrony, the number of germ tubes with 3 nuclei should be low. The variation in the number of nuclei per cell both within and between the immature spore chains suggests asexual reproduction. The number of nuclei per cell during sexual division in fungi is usually well controlled, while it varies considerably in large, asexually produced spores such as certain conidia. E. harknessii aeciospores resemble the latter pattern. It appears that there may be real differences in the numbers of nuclei in aeciospores from one population to another. Epstein and Buurlage (1988) who used mature spores from galls on radiata pine in California showed a ratio of uni- to di- to trinucleate spores of 2:85:14. True (1938) reported that mature spores were usually binucleate. It is also to be noted that Epstein and Buurlage (1988) and True (1938) used spores from a pop-ulation far from those in British Columbia and Alberta and also growing on hosts other than lodgepole pine. It is therefore conceivable that these gall rust populations differ in some respects. It is noteworthy that True (1938), observed grouping of nuclei in mature aeciospores but found no evidence of fusion and also that he observed spores right from the young ones in chains to germlings during host penetration. His observation that there was no nuclear fusion agrees with the results of this study. A repeat of this experiment using spore collections from different parts of the range of western gall rust would be in order. The increase in the number of nuclei per germ tube over time showed a general trend which agreed with the results of Epstein and Buurlage (1988) except that the rate of increase in nuclear numbers was higher in our spore samples than in theirs. It is possible Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 36 that the differences in the rates of nuclear division were due to differences in freshness, the methods of handling and extraction, temperature and so on. Furthermore, the 4-nuclei class was the largest between 8 and 12 hours of incubation in our sample while Epstein and Buurlage (1988) reported the 2-nuclei class to be predominant during the same interval. The fact that the four-nuclei class was still the largest one after 12 hours of incubation could lead one to suggest that it is the class representing the products of meiosis. That would imply that numbers beyond four are produced by mitosis. One can only do that by clearly showing that karyogamy occurs and by ignoring germ tubes with five and six nuclei after only 8 hours. One would also have to ignore odd numbers of nuclei like three and seven which suggest independent nuclear division as Epstein and Buurlage (1988) pointed out. Another problem was that about 60% of mature spores used were uninucleate. No way "was devised to find out whether uni- and binucleate spores germinate the same way. On the other hand, Epstein and Buurlage (1988) used spore samples which were predominantly binucleate came up with results quite similar to the results of this study. It would also appear that consideration should be given to other possible life histories that western gall rust may have but are not yet known. Perhaps karyogamy and meiosis occur elsewhere in the life cycle. The use of transmission electron microscopy may well be used as a confirmatory test to detect meiosis. However, the problem is knowing the right stage of spore development. At the moment, we must limit our conclusions to the materials used. The report by Hiratsuka et al (1966), that young spores are predominantly binucleate and mature ones predominantly uninucleate was not supported by our data. Conse-quently, the observation that karyogamy occurs before spore maturation (Hiratsuka et al 1966) was also not corroborated. The data from germinating spores also produced nuclear numbers not consistent with normal meiotic division. The evidence from the Chapter 3. CYTOLOGY OF DEVELOPING AND GERMINATING SPORES 37 spore samples from the Vancouver region of British Columbia, favours the view that sexual division does not occur during aeciospore formation or germination. If this view is confirmed through further testing, the genus Endocronartium must be redefined or discarded. Chapter 4 E A R L Y S Y M P T O M S O N SEEDLINGS A N D G A L L F O R M A T I O N 4.1 Introduction In the stem and foliage rusts of forest trees, the use of early symptoms in the ranking of resistance reactions is neither as widespread nor as well-developed as in the rusts of cereal crops. Phenomena such as "hypersensitive cell death" in response to infections by obligate parasites like the rusts and mildews have been associated with host plant resistance (Bushnel 1982, Tomiyama 1982). Such cell death or necrosis slows and may stop and restrict fungal growth within the usually tiny necrotic spots. On the other hand, large necrotic lesions or blotches have been associated with susceptible host reactions. Bushnell (1982) pointed out the diversity of host-parasite combinations which have been examined and which have shown conflicting results regarding cell necrosis and fungal inhibition. Symptom types on white pine (Pinus monticola Dougl.) seedlings infected with blister rust fungus (Cronartium ribicola J. C. Fisch. ex Rabenh.) have been described and categorised as needle spot types ( MacDonald and Hoff 1971). These needle spots were described as yellow-brown, yellow, and red and green islands (MacDonald and Hoff 1974). It was speculated that each type was produced by specific genotypic combinations of host and pathogen and represented separate resistance reactions. However, the association of these spots with measurable levels of resistance was not tested. The fungus (C. quercuum ) infects a number of pines in southern U.S. and slash pine 38 Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 39 (P. elliotii Engel) is one of its major hosts. Early symptoms on artificially inoculated seedlings have been used to screen for rust resistance. Lundquist and Luttrell (1982) observed that seedling pigmentation resulting from infection of P. elliotii by C. quercuum was positively correlated with resistance. A red pigmentation on seedlings was the first symptom to appear within 2 weeks of inoculation. This was followed by other types of pigmentation described as dark green or water soaked distinct spots, orange centered, orange reddish, red orange, dark red and purple lesions. Beneath these pigmented lesions, relatively impervious wound periderms were observed and the interpretation was that they seemed to play a role in resistance by separating infected from uninfected tissue and impeding water and nutrient transport to infected cells (Miller et al 1976 , Walkinshaw 1978). The view that these pigmentations were positively correlated with resistance was reaffirmed by Lundquist et al (1982). However, Lundquist and Miller (1984) suggested that the correlation between these symptoms and resistance was not as clear as had been stated earlier. In their histopathological study of fusiform rust, Lundquist and Miller (1984) distinguished between epidermal pigments which appeared within 3 weeks of inoculation and cortical pigments which appeared thereafter. They observed that by the time the epidermal pigment appeared, the fungus had penetrated the epidermis and was in the cortex suggesting that this pigment was not associated with a resistance reaction. This particular observation had been reported by Jackson and Parker (1958). Much earlier, Hutchinson (1935) observed symptoms in the field on P. sylvestris L. infected with the gall forming Peridermium later to be named E. harknessii. He recorded symptoms similar to some of those reported by Lundquist and Luttrell (1982). These were, brownish spots, small swellings, reddish-brown discolourations and numerous spots which were greenish-brown or slightly reddish in the center. He did not relate these symptoms to gall formation. Hoff (1986) described symptoms caused by E. harknessii on ponderosa pine and his data did not show any correlation between stained lesions and Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 40 gall formation. Pigmentations of these sorts had been noted in the University of British Columbia nursery on inoculated lodgepole pine seedlings prior to 1985. The main purpose of this study was to recognise, describe and analyse the variation in frequency of early symptoms between open-pollinated families inoculated at two stages of maturity and two spore loads. The other purpose was to test whether individual symptom types or particular symptom sequences could be used to predict gall formation. 4.2 Materials and Methods 4.2.1 Materials A total of 987 one year-old seedlings from ten open-pollinated mother trees were used in this study. These mother trees were located in the Prince George area of B. C. in two neighbouring stands (Appendix A Tables A.35 - A.37). One stand was a spacing trial named stand A planted in 1967 with a bulk seed collection from the Prince George area. The other, stand B was a natural stand about 25 years old. Both of these stands were infected with western gall rust. Heavily galled trees, with up to 77 galls occurred adjacent to gall free trees. Ten mother trees with varying numbers of galls including 4 gall free ones, were chosen and in the fall of 1984, open pollinated seeds were collected from each. The seeds were pregerminated in petri dishes and planted into 4-piece bullet shaped containers which were colour coded for seedling family identification. The marked and containerised seedlings were raised outdoor in nursery beds. In the last week of April 1985, the 987 seedlings were moved to the University Research Forest near Maple Ridge B. C. to retard shoot elongation and development until spore collection was done. Spores were collected from naturally infected lodgepole pine trees growing in Rich-mond Nature Park, B. C. in Mid-May of 1985. Mature galls from relatively young trees Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 41 of between 15 to 25 years of age were chosen, air dried for a maximum of 24 hours and their spores extracted. Galls already releasing spores were air dried and their spores extracted the same day. During extraction, the bark and the peridium tissues covering the spore surface were lifted carefully with forceps to expose the spores. The spores were then gently brushed or scraped off onto a rayon cloth screen. Dry spores were easily released into a clean petri dish by agitating the veil cloth with a brush. A total of 5g of spore mixture was collected and stored at 0 — 3°C in the dark until inoculation. In the last week of May, all the seedlings were brought from Maple Ridge to the nursery to develop under normal nursery conditions before the first set of inoculations. They were checked daily until virtually all shoots had elongated. When the needles were just emerging from their fascicle sheaths, the seedlings were considered ready for inoculation. Seedlings from each family (mother tree) were divided into two equal groups. The second group was inoculated two weeks after the first so that the effect of the stage of maturity on infections could be tested. At each time of inoculation, the seedlings from each family were further subdivided into two subgroups which received the two spore loads. Thus, there were 10 seedling families, 2 stages of maturity and 2 spore.loads. For each combination of stage of maturity and spore load, there were about 24 seedlings from each family. Furthermore, for each spore load and the stage of maturity, seedlings were arranged in groups of 25 each and taped together. Each group consisted of seedlings randomly chosen from the ten families, so that each family was represented at least once. 4.2.2 The Inoculation Chamber. The inoculation chamber consisted of a wooden cylindrical drum which could be evacu-ated using a vacuum pump (Fig. 4.5). The lid was equipped with two tubes at its centre; one led to a vacuum pump and the other of similar diameter was to let air in after the drum had been evacuated. The inside end of the inlet tube (marked 4) was positioned Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 42 T Figure 4.5: A schematic illustration (not drawn to scale) of the inoculation chamber showing the lid (1), evacuation tube (2), inlet tube (3 and 4) positioned directly above the spore containing beaker (5), clamp (6) and seedlings (7) Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 43 some 5 cm above a 10 ml spore containing beaker. The spore containing beaker sat inside a 1000 ml conical flask which itself was held in place with a clamp attached to the lid. The outside end of the inlet tube was clamped tight, the vacuum pump was switched on and the chamber was evacuated. At a vacuum pressure gauge reading of 10 pounds per square inch, the pump was switched off and the air inlet tube opened to let in a strong jet hitting the spores to form a spore cloud. According to earlier spore deposition trials in the laboratory, spores settled onto the plants below within 2 to 5 minutes. Twenty four hours before the inoculation of seedlings, a spore germination test was done on standard plates of water agar replicated five times at room temperature. As-sessment of germination rates were done after 1-2 hours. Spore loads of 1 g and 0.1 g per 250 seedlings were chosen to represent high and low spore loads respectively. On May 28, 1985 the first group of 500 seedlings was inoculated in the vacuum chamber as described above while the second group was inoculated 10 days later. During each run, five 22 by 22 mm cover slips and five 1 by 1 by 60 mm glass rods coated with thin films of vaseline were placed horizontally and vertically respectively among the seedlings at roughly the same height as the new shoots. These were used to assess levels of spore deposition on the seedlings. The spore cloud generated in the chamber was allowed 5 minutes to settle on the seedlings. The seedlings were removed from the chamber and misted with fine water droplets using a spray bottle. Thereafter, they were placed inside clear plastic tents to maintain high humidity and left at room temperature for 48 hours. The seedlings were then transplanted outdoor into nursery beds 4.2.3 Observations In the first week after transplanting, shoot length of each seedling was measured and recorded. Over the entire growing season, the seedlings were closely examined at least twice a week. Data on symptoms and their development were collected at 2, 4, 8 and Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 44 12 weeks from inoculation. At each time of observation, new symptoms, if any, were recognised and described. At the end of the season, the various symptom categories which had been described were combined into three major types and kept by time of observation. 4.2.4 Data analysis Following the recognition and description of early symptoms, four types of analyses were performed to answer four questions. • The first was whether the percentage (frequency) of seedlings exhibiting particular symptoms was related to family, stage of maturity of seedlings at inoculation and inoculum load. • The second was whether the incidence of infection as percent seedlings with galls after 46 weeks varied significantly between families. This part was reported in detail in the last subsection (4.3.5) of the results section. • The third was whether the frequency of particular symptoms by family was related to the degree of infection. • The fourth question was whether the infection of individual seedlings could be predicted by the presence of particular symptoms appearing on seedlings between 2 to 8 weeks following inoculation. The treatment of seedlings at two stages of maturity with two spore loads was a factorial treatment while the planting in nursery beds was completely randomised. Each treatment combination had at least 23 seedlings from each family. All the effects were considered fixed. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 45 The frequency of individual symptom types by family was analysed using a 3-way analysis of variance model. The frequency of a particular symptom within a family was the number of seedlings with that particular symptom type expressed as a proportion^ of the total number of seedlings in that family. Since such proportion data are not normally distributed and would violate a major assumption in the analysis of variance, an arcsine-square root transformation of the proportion was employed. Residual plots done in SAS package showed that the transformation was appropriate. Analyses of variance were then performed for each time of observation and symptom type. Each analysis was based on a three- factor, fixed effect model with all factors crossed. The factors tested in the analysis were; age (stage of maturity) at inoculation (1 ,2), seedling family (1 to 10) and inoculum load (1 , 2). The linear model was given as : Yijk =:fi + Oi+Bj + 8k + afc + a6ik + 86jk + eijk (4.1) where; Y{jk is the arcsine-square root of p at age of inoculation i, family j and spore load k showing a particular symptom type, p is the general mean, a; seedling age effect, 3j seedling family effect, 8k the inoculum load effect, a/3\j the age by family interaction, ct8ik the age by inoculum load effect, 06jk the family by inoculum load effect and €ijk the error term. The expected mean squares for the above linear model appear in Table 4.1. Variation in the amount of infection among families was assessed using analysis of variance of arcsine p seedlings infected. The linear model was : Yijk = p + Ai + Bj + Ck + AB{j + ACik + BCjk + Eijk (4.2) Where Yijk is the transformed p infected per family, p the general mean, Ai age effect, Bj seedling family effect, Ck inoculum load effect ; ABij is the age by family interaction, ACik the age by inoculum load interaction, BCjk the family by inoculum load interaction Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 46 Table 4.1: Expected mean square table for ANOVA of early symptom frequency among 10 open-pollinated lodgepole pine families inoculated with western gall rust at two spore loads and two times of inoculation.  SOURCE DF EXPECTED MEAN SQUARE AGE (A) 1 + C6crA FAMILY (F) 9 _2 <TE + C§0~p SPORE (S) 1 _2 <TE + Cher's AGE x FAMILY 9 _2 C3crAF AGE x SPORE 1 <4 + C2crAS FAMILY x SPORE 9 + Ci<rFS ERROR 9 and, Eijk is the error term (the 3-way interaction). F-tests for this analysis also followed Table 4.1. To test the degree of correlation between the frequency of a particular symptom type and resistance, the following was done. For each symptom type and each time of observation, a set of four correlations was performed one for each spore load and age, to relate the frequency (percent) of seedlings by family bearing that symptom to the percent infected. The resulting r values were an indication of the degree to which the frequency of the particular symptom type was associated with infection levels in all the families. Optimum conditions (spore load, seedling age and time of observation) for distinguishing between the resistance of families by their early symptoms could then be identified as those which gave the highest r values. The relationship between the appearance of symptoms on a seedling and the forma-tion of galls on that seedling was tested by constructing a series of 2 by 4 contingency tables (Appendix B, Tables B.52 - 53), one for each combination of spore load, age at inoculation, and time of observation, yielding 12 tables describing about 245 seedlings each. In each table, infected and non-infected seedlings were divided into those with and Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 47 without symptoms. The expected frequency of each category was then calculated from the assumption that appearance of symptoms and infection were totally independent and a x value (3 df) was computed. The data on infections were described as follows. Changes in the frequencies of symptom types were illustrated using line graphs. Frequency distribution histograms were used to illustrate distribution of infections per seedling in each family. In addition, regressions of offspring on parents using % of offspring infected on galls per mother tree as dependent and independent variables respectively, were done. 4.3 Results 4.3.1 Description of Early Symptoms The initial intent was to combine the early symptoms (2 to 12 weeks) into a few common patterns or sequences and relate these to families and to subsequent infection. However, virtually every possible pattern or sequence was observed and none of these was partic-ularly common. Hence the attempt to deal with symptoms as few common patterns was abandoned. Instead, symptoms appearing at different times of observation were treated separately. Three major symptom types were recognised and described. These were general red, red flecks and red streaks. In addition, there were seedlings which did not exhibit any external symptoms. The general red stain was the predominant symptom in the first two weeks following inoculations. These stains appeared in large patches covering from half to entire shoot surfaces (Fig. 4.6A). It could later be replaced by, or associated with other types. These red stains could disappear while some of the flecks, spots and streaks persisted and remained visible 3 months later, when the first galls appeared on seedlings. Occasionally, a red stained patch about 2 months old could be sloughed off and replaced by new tissue Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 48 beneath it and in some cases the stains seemed to have just faded away. Flecks were observed against a green plant surface or superimposed on an already red stained surface. They ranged from 1 mm diameter spots to 3 by 1 mm red flecks (Fig. 4.6 B and C). Unlike the general red stain, the flecks appeared 14 or more days after inoculation. About 30 to 45 days later, these spots were much darker compared to their initial light to bright red. Some were clearly necrotic and the plant tissue around them dry and sometimes cracking. Red streaks appeared against a greenish plant surface or superimposed on a red-stained surface (Fig. 4.6 D and E). They measured 5-20 mm in length and were pink-red in the beginning to dark-red toward the end of summer. One to two months from inoculation, some appeared dry and necrotic and would persist over the summer and still be seen on the surfaces of newly formed galls. Fig. 4.7 shows that the frequency of each symptom type changed over time for every family. By twelve weeks, most of the symptoms had disappeared. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION Figure 4.6: Seedlings of lodgepole pine showing general red stains, red flecks, red streaks and no symptoms following inoculations with western gall rust E. harknessii. A. General red stain on a 24-week-old seedling photographed 3 weeks after inoculation. B and C show red flecks (spots of 1 mm in diameter and 1x3 mm flecks) in the upper half of a shoot 4 weeks after inoculation. D shows bright to dark red flecks intermingled with fine red streaks (5-10 mm in length) 4 weeks after inoculation. E shows large streaks (4 weeks after inoc), and F is a seedling 10 weeks after inoculation showing no symptoms; the surface is brown indicating the formation of primary periderm. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 50 Figure 4.7: Changes in the frequency of early symptoms during the 12 weeks following the first inoculation of lodgepole pine seedlings with two spore loads of western gall rust. The two spore loads were combined while the ten families (A-l, A-2, A-10, A-4, A-12; B-l, B-2, B-3, B-4, B-5) were shown separately. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 52 4.3.2 Frequency of symptoms among families Symptoms were much more common following the first set of inoculations whereas older seedlings exhibited relatively fewer symptoms. General red staining was most common two weeks after inoculation and declined rapidly after that in all families (Fig. 4.7). The proportion of seedlings with red flecks or streaks was generally less than 0.20. The frequency of symptom types at one or more of the observation times tested, varied signif-icantly among families, between inoculation times (stage of seedling maturity) and spore loads (Tables 4.2, 4.3; Appendices B.38 - B.50). There were significant family differences in the relative frequencies of symptom types (Table 4.2 and 4.3). For example, families differed significantly with respect to red flecks but not with respect to red streaks at all the three times of observation. In addition to Table 4.2, other tables constructed for all the four symptom types and all the times of observation appear in Appendix B (Tables B.39 - B.50). Age or stage of maturity of seedlings at inoculation was significant with respect to the frequency of the symptom described as general red. Seedlings at the earlier stage of maturity produced more general red stains than their counterparts which had been inoc-ulated 14 days later. However, such general red stains were also observed on uninoculated controls. Their importance in relation to resistance to gall rust was therefore doubtful. Families differed significantly at all the three times of observation with respect to red flecks and the category termed "no symptom". There were no significant differences among families with respect to red streaks. Effects of spore load on symptom production varied for different symptom types and at different times of observation but the variations were not significant at all times of observation (Table 4.2). Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 53 Table 4.2: Summary of ANOVA of the frequencies of three symptom types produced on lodgepole pine seedlings following inoculations with western gall rust. Analyses were done separately for each symptom type at 2, 4 and 8 weeks following inoculation General Red Red Flecks Red Streaks None Time of Obs. Time of Obs. Time of Obs. Time of Obs. Source 1 2 3 1 2 3 12 3 12 3 FAMILY * * * * * * * SPORE * . . . * * FAMILY X SPORE - - - . . . . . . . . . AGE * * * * * * FAMILY X AGE * . - - - . . . . . . . SPORE X AGE * * . . . * _ * ERROR *the asteriks indicate significant effects (p=0.05) at each time of observation; age refers to the stage of seedling maturity at inoculation. Table 4.3: Results of analysis of arcsine-square root of % of seedlings of open-pollinated lodgepole pine families showing red flecks, 4 weeks following inoculations with two spore loads of western gall rust.  SOURCE DF SS MS F-VALUE F-PROB FAMILY 9 0.8817 0.0977 10.2280 0.001 SPORE 1 0.0345 0.0345 3.6047 0.088 FAMILY x SPORE 9 0.1722 0.0191 1.9980 0.158 AGE 1 0.0019 0.0011 0.1138 0.739 FAMILY x AGE 9 0.1601 0.0178 1.8583 0.184 SPORE x AGE 1 0.2375 0.2357 24.8018 0.001 ERROR 9 0.0862 0.0096 Age refers to the stage of seedling maturity at inoculation. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 54 4.3.3 Symptoms and gall formation at the family level Families differed significantly with respect to red flecks at all the 3 times of observation (Table 4.2) but the significance of red flecks with respect to gall formation was not clear. Furthermore, families also differred significantly with respect to % infected and the number of galls per seedling; both summarised in tables 4.4-4.7. The ANOVA tables are reported in greater detail in section 4.3.5. The r values (Tables 4.8 and 4.9) relating the frequency of a symptom and % infection by family showed that very few of the relationships were significant. The poor relationship between symptoms and infection was illustrated by the frequency of red flecks and % infected in the two families B-4 and A-10. Families B-4 and A-10 had the highest proportions of their seedlings exhibiting red flecks. Family A-10 turned out to be one of the most infected, while A-4 was one the most resistant (Tables 4.4 - 4.7). Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 55 Table 4.4: Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the first stage of seedling maturity and high spore load (1.0g/250 seedlings) Family n Galls on Female Parent Mean Galls Per Seedling Percent Infected Mean Candle Length (cm) A-l 25 7 1.00 44.0 6.70 A-2 26 27 3.73 84.0 8.52 A-10 25 0 3.48 80.0 8.56 A-4 24 6 4.79 75.0 11.04 A-12 26 0 2.38 42.0 8.63 B-l 24 0 2.58 71.0 7.67 B-2 24 16 1.58 62.5 6.89 B-3 23 . 4 2.87 87.0 8.76 B-4 25 0 1.16 44.0 6.18 B-5 25 77 2.20 64.0 7.75 Table 4.5: Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the second stage of seedling maturity and low spore load (0.1g/250 seedlings) Family n Galls on Mean Galls Per Percent Mean Candle Female Parent Seedling Infected Length (cm) A-l 25 7 0.84 80.0 10.50 A-2 25 27 1.60 72.0 8.52 • A-10 24 0 1.64 71.0 7.66 A-4 25 6 1.92 68.0 8.32 A-12 26 0 1.20 26.9 9.24 B-l 25 0 0.92 52.0 7.26 B-2 26 16 0.92 65.0 7.66 B-3 24 4 2.00 67.0 9.26 B-4 21 0 0.46 28.0 7.47 B-5 25 77 0.68 72.0 7.50 Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 56 Table 4.6: Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the second stage of seedling maturity and high spore load (1.0g/250 seedlings) Family n Galls on Mean Galls Per Percent Mean Candle Female Parent Seedling Infected Length (cm) A-l 25 7 2.24 56.0 6.70 A-2 25 27 3.12 68.0 7.24 A-10 25 0 2.54 64.0 8.52 A-4 25 6 1.89 76.0 7.50 A-12 25 0 0.81 48.0 8.50 B-l 25 0 1.76 48.0 7.08 B-2 25 16 1.92 52.0 8.44 B-3 25 4 2.96 68.0 9.04 B-4 24 0 0.72 33.0 5.71 B-5 25 77 1.64 40.0 7.30 Table 4.7: Infection Levels on 1-year-old lodgepole pine seedlings from 10 open-pollinated families inoculated with western gall rust spores. This table presents data from the second stage of seedling maturity and low spore load (0.1g/250 seedlings) Family n Galls on Mean Galls Per Percent Mean Candle Female Parent Seedling Infected Length (cm) A-l 24 7 0.33 25.0 9.77 A-2 25 27 0.56 32.0 9.06 A-10 25 0 0.24 20.0 10.26 A-4 24 6 0.50 38.0 7.60 A-12 25 0 0.52 20.0 8.64 B-l 25 0 0.16 16.0 7.78 B-2 24 16 0.21 17.0 7.83 B-3 25 4 0.44 28.0 9.62 B-4 24 0 0.13 12.0 10.12 B-5 24 77 0.38 21.0 8.04 Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 57 Table 4.8: Correlations between the occurrence of individual symptom types and gall formation (percent infected per family) on lodgepole pine seedlings following inoculations with western gall rust. This table presents results of analysis of seedlings at the first stage of maturity (AGE 1) and the entries; r values are stratified by spore load and time of observation. Time 4 represents symptom types at the time when the first galls were recorded. r# = 8 critical=0.6319, p=0.05 Time Spore Load General Red Red Flecks Red Streaks No Stain 1 1 0.61 — — -0.61 1 2 0.23 — - — -0.23 2 1 0.11 0.04 -0.07 -0.52 2 2 0.47 -0.23 -0.03 -0.08 3 1 0.33 0.02 0.15 -0.40 3 2 0.48 0.09 -0.18 -0.18 4 1 -0.59 -0.39 -0.47 0.47 4 2 -0.33 0.18 0.87 0.01 Table 4.9: Correlations between the occurrence of individual symptom types and gall formation (percent infected per family) on lodgepole pine seedlings following inoculations with western gall rust. This table presents results of analysis of seedlings at the second stage of maturity (AGE 2) and the entries; r values are stratified by spore load and time of observation. Time 4 represents symptom types at the time when the first galls were recorded. rdf=s criticab=0.6319, p=0.05 Time Spore Load General Red Red Flecks Red Streaks No Stain 1 1 0.03 — — -0.12 1 2 0.77 — — -0.77 2 1 0.08 0.19 -0.24 -0.12 2 2 0.73 -0.42 -0.08 -0.76 3 1 0.21 ' 0.41 -0.22 0.06 3 2 -0.39 0.15 -0.44 0.05 4 1 0.33 0.27 0.57 -0.18 4 2 0.21 • 0.08 — -0.11 Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 58 4.3.4 Early symptoms and infection of individual seedlings The relationship between the presence of a symptom and gall formation on individual seedlings was shown by the % values in Table 4.10 to be inconsistent and in most cases, not significant. In general, seedlings bearing early symptoms were more likely to become infected than those which did not (Tables 4.11 and 4.12). However, only 5 of the 12 % values (Table 4.10) were significant at p=0.05, and even in those cases, the differences (percent infected) of seedlings in various symptom classes, were not very large. Tables B.51 and B.52 in Appendix B show the twelve 2 by 4 contingency tables on which these were based. Table 4.10: Table of X(<#=3) values relating the presence of symptoms and percent infection caused by western gall rust on lodgepole pine. The values were calculated for the 3 separate times of observation and the 2 spore loads. The X(df=3) Po.os=7-82 Time Spore X (Age 1) X (Age 2) Load 1 1 13.46 9.16 1 2 9.46 0.65 2 1 11.11 15.19 2 2 2.41 1.96 3 1 5.14 4.09 3 2 5.76 7.77 Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 59 Table 4.11: Overall x tests relating the presence of symptoms and gall formation among lodgepole pine seedlings inoculated with western gall rust at the first (AGE 1) of the two stages of seedling maturity that were used in the experiment. The two spore loads and the three separate times of observation were pooled. Symptoms Infection J. Stained Not Stained Row Total Galled 284 (272) 30 (42) 314 Not Galled 143 (155) 36 (23) 179 Column Total 427 66 493 Seedlings regardless of family, inoculum load and time of observation, were divided into two groups; stained and not stained; expected values in parentheses and X(d/=i) value = 12.23; the X(df=i) critical=:3.84. Table 4.12: Overall x tests relating the presence of symptoms and gall formation among lodgepole pine seedlings inoculated with western gall rust at the second (AGE 2) of the two stages of seedling maturity that were used in the experiment. The two spore loads and the three separate times of observation were pooled. Symptom—* Infection J, Stained Not Stained Row Total Galled 170 (157) 24 (37) 194 Not Galled 231 (244) 69 (56) 300 Column Total 401 93 494 Seedlings regardless of family, inoculum load and time of observation, were divided into two groups; stained and not stained; expected values in parentheses and X(df=i) value = 8.69; X(df=i) criti-cal=3.84. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 60 4.3.5 Variation in Gall Formation In the first week of August 1985; about 3 months following inoculation, most of the general red stains had disappeared from the seedlings and only a few seedlings (about 20 to 30%) per family had either red flecks or red streaks some of which showed necrosis. Most stem surfaces had turned brownish in colour, indicating periderm formation. At the same time, small swellings, sometimes rough and producing resin showed on some seedlings. Almost all of these swellings became galls a mere two to three weeks later in the first week of September, 1985. Epicormic shoots formed below or above lesions in bunches all around the stem were associated with gall formation. Virtually all seedlings which had epicormic shoots on or near stained lesions developed galls. In some cases, the epicormics appeared after gall formation. Most gall were positioned just below shoot tips. Some seedlings had galls on an entire shoot and as many as 10 separate galls were observed on 8 cm shoots whereas seedlings with multiple shoots supported 20 or more galls. Figures in Appendix C (Table C.53) show that for each of the ten families, and both spore loads, the frequency distributions were skewed toward low infection levels. Analysis of arcsine-percent infected showed significant family differences (Table 4.13). The analysis using candle length as a covariate showed that covariance analysis was justified. The entries in the ANOVA (Table 4.14) are therefore based on adjusted values. Differences in galls per seedling were significant (p=0.05) with respect to the stage of maturity, spore load and seedling family (Table 4.14). At the higher spore load, seedlings at both stages of maturity incurred higher levels of infections, both as % infected and the mean number of galls per seedling than those inoculated with the lower spore load (Tables 4.4 to 4.7). In general, seedlings inoculated at the second stage of maturity showed lower infection levels than those inoculated 2 weeks before them (Tables 4.4 to Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 61 4.7). Families A-12 and B-4 derived from gall free parents had significantly lower mean number of galls than the rest of the families. Other families from gall free parents; A-10 and B-l were not significantly different from the rest. The age x spore and the spore load x family interactions were significant (p=0.05) whereas the age x family interaction was not. The regressions of offspring on parents (Tables 4.15), showed poor relationships be-tween the number of galls on parents and the % infected and mean galls per seedling of open-pollinated offspring. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 62 Table 4.13: Results of ANOVA of arcsine-percent infected of 10 open-pollinated lodgepole pine families inoculated with western gall rust at two stages of seedling maturity and two spore loads SOURCE DF SS MS F-STAT F-PROB FAMILY 9 0.565 0.063 5.167 0.015 SPORE 1 0.481 0.481 39.550 0.001 AGE 1 0.516 0.516 42.478 0.001 SP x FAM 9 0.182 0.020 1.667 0.241 SP x AGE 1 0.203 0.203 16.614 0.004 AGE x FAM 9 0.176 0.019 1.613 0.086 ERROR 9 0.097 0.012 Table 4.14: Results of ANOVA of galls per seedling (square-root transformed) from 10 open-pollinated lodgepole pine families infected with western gall rust at two stages of SOURCE DF SS MS F-STAT F-PROB FAMILY 9 42.986 4.776 7.765 0.001 SPORE 1 31.157 31.157 50.653 0.001 AGE 1 99.653 99.653 162.010 0.001 SP x FAM 9 12.473 1.386 2.253 0.017 SP x AGE 1 9.479 9.479 15.412 0.001 AGE x FAM 9 9.362 1.040 1.691 0.086 SP x AGE x FAM 9 5.059 0.562 0.914 0.513 ERROR 946 466.023 Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 63 Table 4.15: Summary of regressions of offspring (lodgepole pine seedlings) on female par-ents. The offspring were inoculated with western gall rust at 2 stages of seedling maturity I and II; the independent variable was the number of galls on the female parent and the dependent variables were % infected and the mean number of galls per open-pollinated family. The two spore loads were combined. • Stage of Maturity Error DF F-value Significance (PR>F) R 2 AGEI 1 8 0.73 0.419 0.083 AGEI 2 8 0.09 0.770 0.011 AG EI I 1 8 0.31 0.593 0.037 AGE 11 2 8 0.18 0.684 0.022 1 is (%) Infected; 2 is the mean number of galls per seedling, AGE I and II are the two stages of maturity of seedlings at inoculation Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 64 4.4 Discussion 4.4.1 Symptoms The diversity of early symptoms possibly indicates a corresponding diversity of genetic resources for host reaction and/or resistance. Because a bulked spore collection was used in this experiment, it is impossible to tell whether the various host responses that were categorised as distinct symptoms were elicited by distinct pathogen genotypes or are the constitutive property of host genotypes and therefore non-specific all rust genotypes. The poor correlation between the frequency of symptom types and infection, under all the conditions tested, suggests that early symptoms cannot be used to predict resistance at the family or individual seedling level. Instead, progeny tests must be maintained for about 1 year following inoculation in order to allow distinct gall to develop. The results of this study agree with those reported by Hoff (1986) in which no significant correlation between symptoms and resistance of ponderosa pine to western gall rust was found. Early symptoms are probably a host resistance reaction to penetration and colo-nization by the rust. Since red flecks and streaks appeared after inoculation and often developed into necrotic spots isolated by necrophyllactic periderms (Mullick 1977), they probably represent resistance reactions. The high inoculum loads and favourable infec-tion conditions used in this experiment may explain why these resistance mechanisms were often ineffective. It may be that the resistance reactions associated with some of these early symptoms are of the rate reducing type; effective only against a proportion of all penetrations. Thus, there are several reports in the literature of periderm formation in response to bark invasion by Cronartium rusts, and instances of development of fun-gal mycelium beyond such periderms (Jackson and Parker 1958, Lundquist and Miller 1984). The favourable infection conditions used in this study ensured multiple penetra-tions shown by the fact that there were several separate red flecks per seedling. A number Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 65 of these may have circumvented the resistance mechanisms, leading to the formation of one or more galls on many of the symptom-bearing seedlings. The favourable conditions were used in order to obtain infection rates that would distinguish among families. The overall infection level achieved in this study, 51 % , was considerted near-optimum for this purpose. Such conditions are however, much more favourable for infection than in the field. Valuable resistance may be masked by such inoculum loads. Many asymptomatic seedlings remained free of infection indicating that there may also be resistance reactions which do not give rise to early symptoms. Such resistance mechanisms may operate early during the penetration of host cuticle, thus preventing the invasion of internal host tissue. 4.4.2 Variation in Gall Formation The two spore loads used in this study did not significantly change the resistance ranking of the 10 families at either stages of maturity at inoculation (age x family interaction, spore x family interaction in Tables, 4.13 and 4.14). This result is similar to that of Blenis and Pinnel (1988) who showed that family ranks were not significantly changed with increasing spore loads. The levels of infection among the ten families demonstrated genetic variability with respect to resistance. The variability was largely attributable to two families (Tables 4.4 and 4.5), which at age 1 had less than 50% of their members infected at the high spore load and less than 30% with the lower spore load. At age 2 the differences were much less pronounced. It was shown that some families derived from gall free parents produced progenies as susceptible as those from infected parents. Thus, the infection observed in the field was not consistent with that resulting from artificial inoculations. In the fusiform rust studies, Lundquist et al (1982) concluded that some form of resistance breakdown occurs, especially under extremely high inoculum loads in the laboratory. The differences Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 66 between spore mixtures used in these experiments and those in natural stands should partly explain the observations. The number of infections counted per seedling was partly affected by the saturation of a candle with spores. A 6.0-cm candle for example, could only bear a limited number of separate galls and would therefore bear no more infections despite a further increases in spore load. Therefore, a larger seedling could have appeared more susceptible than a smaller one even if the two were genetically the same with respect to resistance. Fur-thermore, galls formed close to each other because of saturation could not be easily distinguished. It follows that, on some seedlings, the total number of separate infections was underestimated. Skewed host frequency distributions favouring low levels of infection present a dif-ficulty because the factors that cause them have not been explained. The results of Burdon (1980a) on the reaction of Trifolium repens to a disease is an example of such a distribution. The theoretical multiplicative model (Person et al, 1980) also predicted such a skew and hypothesized that it could be evolutionarily stable. The interpretations of skewness and its place in stable host-parasite systems are interesting but beyond the scope of this study. One point to be made is that Burdon (1980a) did not use disease isolates known to differ in virulence. He assumed a non-specific interaction between the pathogen and host. It is therefore possible that a distribution skewed toward low levels of infection could be an artifact of experimentation. These could be caused by factors such as the selection of a small sample of pathogen isolates which were virulent on just a portion of the resistance spectrum of a host population. Burdon (1980a) concluded that high selection pressure on a host by a pathogen favours selection for host resistance and thus the skewness and a mild selection pressure produces a normal resistance distribu-tion. That hypothesis is not easy to test in the gall rust system and it seems unlikely that lodgepole pine is under a high selection pressure from gall rust in B. C. Chapter 4. EARLY SYMPTOMS ON SEEDLINGS AND GALL FORMATION 67 The conclusions from this chapter are as follows: • Variation in the frequency of various symptoms produced by seedlings within and between families following inoculations with E. harknessii can be interpreted as caused in part by genetic variability in the host. • The early symptoms were poor indicators of gall formation • Families were significantly different with respect to the mean number of galls per seedling and % galled, providing further evidence of genetic variability in resistance to the rust. • The ranking of families by resistance (mean number of galls per seedling) did not change significantly between the two spore loads and two inoculation times. • A 2-week delay in inoculation of seedlings ready for inoculation markedly decreased infection levels and showed that spore release in synchrony with the state of matu-rity of host tissue is critical for infection. Chapter 5 PINE C L O N E S A N D SINGLE-GALL S P O R E SOURCES 5.1 Introduction and Literature Review-As regards research in genetics and breeding, the importance of clones in the estimation of genetic parameters has long been recognised (Burdon and Shelbourne 1973, Libby 1973). Burdon and Shelbourne (1973) showed how total genotypic variance, environ-mental variance and genotype-environment interactions can be estimated in clonal ex-periments. There is no doubt that the separation of genetic from environmental variances is necessary in genetic experiments, and that the recognition of genotype-environment interactions leads to the matching of genotypes to sites. In disease resistance breeding, the mean response of a parent for a given trait can be efficiently estimated through its clones (Burdon and Shelbourne 1973). This mean response is an estimate of the genotypic value of the parent and can be compared with that of the progenies derived from open-pollination of the cloned parents. In general, superior to open-pollinated and controlled-pollinated progenies when the description of the genotypes of parents is criti-cal. Since it is difficult to cross isolates of western gall rust for genetic analysis, studies of the variation in virulence and protein (electrophoretic) profiles of single-gall collections can be used. Electrophoretic techniques have been used in the study of variation in fungi (Chesson et al 1978). Howes et al (1982) and Kim et al (1984) studied variation in polypeptide patterns among isolates of the stem rusts of wheat and oats caused by P. 68 Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 69 gramminis f. sp. tritici and P. gramminis f. sp. avenea respectively. The technique was used in the taxonomy of rust species also in the genus Puccinia which infect species of knapweed (Kim and Mortensen 1986, Watson et al 1981). Isozyme analysis of single-gall spores of western gall rust collected in California and Oregon showed very little variation in the fungus over a wide geographical range (Vogler et al 1987). The results (Vogler et al 1987) suggested that the gall rust populations along the coasts of Oregon and California may be homogeneous. The aim of this experiment was to test whether the variation and ranking observed among parent trees in the field could be reproduced through artificial inoculations. The second aim was to test the variation in virulence and electrophoretic profiles among rust spore sources collected from two adjacent pine stands and also to test the interactions between pine clones and single-gall spore sources. 5.2 Materials and Methods 5.2.1 Sources and Preparation of the Clones The clonal materials were grafts derived from two stands in Prince George. One was a 25-year-old natural stand (Appendix A, Table A.37) and the other an experimental plantation (Tables A.35 and A.36) established with seed collected from Prince George timber supply area. Both stands were heavily infected with western gall rust and the mean number of galls per tree was about 40 and 10 respectively. The natural stand provided locally occurring genotypes whereas the plantation provided a wider mixture of genetic material from the Prince George area than one could ordinarily get from a local stand. In the spring of 1983, scion materials from a number of trees from both the natural and planted stands mentioned above were collected and brought to Vancouver for grafting. Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 70 The root stocks were 1-year-old seedlings derived from open-pollinated parents in the same stands. In 1987, the smallest clone consisted of one plant with 9 shoots and 65 cm of shoot length. Hence, each spore source could be replicated at least twice on each clone. These nine shoots would be the equivalent of 9 seedlings with an average shoot length of about 7 cm. The grafted trees were placed in cold storage until spores matured in the field. The four-year-old clones had flushed by the second week of June 1987, 12 days after coming out of cold storage. The total number of shoots suitable for inoculation was recorded. The shoots from each clone were divided into four equal groups to be inoculated with each of the four spore sources. Shoots were randomly assigned to spore source so that a single spore source would not be clumped on adjacent shoots. Shoots were marked using coloured wires to identify the spore source that each would receive. A total of 10 shoots from 6 of the clones were marked as untreated controls. A day before inoculation, germination tests on 2% agar were conducted for each spore source. 5.2.2 Spore collection Four large single galls were collected in the last week of May 1987, two from each of the two stands in Prince George described above. Spores were extracted, kept in glass vials and stored dry at —3 to — 5°C for 11 days until the clones were ready for inoculation. 5.2.3 Inoculation Technique All shoots except the ones receiving a particular spore source were covered with clear 40 by 50 cm plastic bags during inoculations. Dry spores were spread evenly at the bottom of a petri dish then picked using a soft paint brush and spread on a shoot from the tip down. The inoculated shoot was then misted with distilled water, enclosed in polythene bags and tied at the shoot base to maintain high humidity. The plants were Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 71 watered and kept at room temperature with the polythene bags on for a total of 48 hours. Thereafter, the bags were removed and the plants transferred to the nursery where their pot containers were dug into nursery beds. 5.2.4 Collection of Data and Analyses The treatment of clones with 4 spore sources was a factorial arrangement followed by a completely randomised planting in nursery beds. The pine clones and the four spore sources were considered random. The number of galls per shoot was recorded 10 months after inoculation. The mean number of galls per clone by spore source was analyzed using a two-way analysis of variance (ANOVA) model. Due to the possibility that the number of infections per shoot might be influenced by shoot length, the model treated shoot length, X, as a covariate. The linear model was constructed as below and covariance analysis and multiple comparison of clonal means were performed using University of British Columbia ANOVAR statistical package. Yijk =n + Ai + Bj + ABij + Xijk + Eijk (5.3) where; Y^. is the number of galls per shoot adjusted for shoot length, fi the general mean, A{ the spore source effect; Bj the clonal effect, AB^ the spore by clone interaction , X^k the covariate and E^k is the error term. A linear regression of the overall infections per clone (mean of all the four spores) on the number of infections on their respective parent trees was performed using the SAS statistical package. Thus the R value would indicate the degree to which the clones could reproduce the levels of infections observed on their parent trees. Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 72 5.2.5 Electrophoretic Variation among Single Galls Protein extraction was done according to the method described by Gabriel and Ellingboe (1982). Protein estimation was done using the method of Lowry et al (1951) and the accuracy of the method was improved as recommended by Siegel (1976). Gels were stained with silver using the protocol by Merril et al (1981). 5.3 Results The first galls appeared on the clones 14-16 weeks from inoculation which was in the second to fourth week of October. However, fully formed galls did not appear until 8 to 10 months later. Analysis of variance showed significant clone by spore interactions with respect to galls per clone (Table 5.16), so the means in Tables 5.17 and 5.18 could not be inter-preted simply. Mean galls per shoot for the 16 clones plotted against each spore source (Figure 5.8) illustrated the interaction. For example, spore source 2 infected the high-est proportion of all the shoots on which it had been used (Table 5.17). However, the same spore source caused low infection levels on clone 11 which was highly susceptible to source 1. Three spore sources 1, 2 and 3 produced similar levels of infection on clone 3 but spore source 4 caused twice as many infections on the same clone. The resistance ranking of pine clones therefore depended on the spore source used. Spore sources 2 and 3 had the highest and lowest mean infection levels respectively (Table 5.17). Clones 5 and 7 whose parent trees were uninfected in 1982 were not infected by any of the spore sources (Table 5.18) but 3 other clones from uninfected trees were infected. Two of the 16 cloned parents; B-5 and B-8 which had the highest number of infections of 77 and 42 galls respectively, were no more susceptible than those which had between 5 Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 73 and 16 galls. For that reason they significantly reduced the coefficient of determination of clone on parent regression. Without the two parents, the regression of the mean number of galls per clone on the total number of infections on their respective parent trees was markedly improved and showed a positive and significant relationship (p=0.001) with an rdf-i2 v alue of 0.60. However the significant relationship after the exclusion of the two parents should be interpreted with caution. Further testing using more cloned parents and single-gall inocula is needed. There was some electrophoretic variation (Fig E.13) among the four single-gall spores used in the inoculation of the 16 pine clones. As in the other spore sources examined, variation occurred mainly in three regions. However, most bands were diffuse and made comparisons between profiles difficult and inconclusive. The results were therefore re-ported and discussed in Appendix E. Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 74 Table 5.16: Results of ANOVA of the number of galls per shoot (adjusted for shoot length) on 16 lodgepole pine clones inoculated with four single gall spore sources of SOURCE DF SS MS F-STAT F-PROB SPORE 3 2.806 0.935 5.7721 0.01 CLONE 15 13.356 0.890 5.4941 0.01 SPORE x CLONE 45 7.271 0.162 1.787 0.01 ERROR 249 22.514 0.090 1 The main effects were tested against the interaction term. Table 5.17: Average infection levels caused by four single gall spore sources of western gall Spore Number of Mean Shoot % of shoots Mean Galls Source Shoots Length (cm) infected per Shoot 1 82 9.8 58.5 1.5 2 79 10.4 67.1 1.9 3 79 9.8 41.8 0.7 4 74 9.7 63.5 1.8 Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 75 Table 5.18: Infection summary for 16 lodgepole pine clones derived from two adjacent stands near Prince George, B. C., and inoculated with four single gall spore sources of western gall rust collected on trees from the same stands. The mean number of galls reported here for each spore source were not not adjusted for shoot length; the covariate Clone Parent Tree Galls on Parent Number of Shoots Mean Shoot Length % galled Mean Galls per Shoot Spore Source Overall Mean 1 2 3 4 1 (A-5) 18 15 12.9 66.7 2.3 2.3 1.0 2.3 1.7 2 (B-5) 77 18 8.9 50.0 0.2 0.5 1.3 1.4 0.8 3 (B-8) 42 17 8.2 29.4 0.0 0.5 0.4 1.5 0.6 4 (B-3) 4 11 10.2 36.0 1.5 0.0 0.3 0.0 0.4 5 (A-6) 0 15 10.4 0.0 0.0 0.0 0.0 0.0 0.0 6 (A-3) 25 13 7.4 92.3 2.3 5.0 3.0 1.3 2.8 7 (B-4) 0 9 7.2 0.0 0.0 0.0 0.0 0.0 0.0 8 (B-6) 5 10 13.7 30.0 1.0 0.5 0.0 0.0 0.4 9 (A-4) 0 9 11.3 77.8 2.0 3.5 0.5 0.0 1.8 10 (A-2) 27 72 11.5 81.9 2.5 2.9 0.9 3.2 2.4 11 (B-7) 1 12 9.2 25.0 3.0 0.0 0.0 0.0 0.8 12 (A-7) 10 13 8.6 38.5 0.0 3.5 0.3 2.3 1.2 13 (A-8) 13 20 9.4 90.0 1.8 1.8 1.6 4.8 2.5 14 (A-l) 0 25 10.1 64.0 0.6 1.1 1.0 1.0 0.9 15 (B-10) 0 24 9.8 62.5 1.0 1.2 0.6 1.6 1.1 16 (B-9) 11 31 7.3 54.8 2.0 2.6 0.4 0.6 •1.4 Means 14.6 19.6 9.7 49.9 1.3 1.6 0.7 1.2 1.2 Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 76 5 -e-Figure 5.8: Mean number of galls per shoot for 16 lodgepole pine clones inoculated with 4 single gall spore sources of western gall rust. All clones and spore sources came from two adjacent stands near Prince George. The clones were arranged from left to right along the X axis starting with the highest to the lowest average infection levels. Each line in the graph joins points representing the response (unadjusted mean galls per shoot) of each of the 16 pine cloneo to a single spore source Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 77 5.4 Discussion It took 44 to 46 weeks from inoculation for the clones to clearly show galls as opposed to 12-14 weeks in younger 1-year-old seedlings. This suggests that there was an influence of the physiological age of shoots on the rate of gall development. In addition, the number of galls per shoot was considerably lower than the numbers normally achieved on 1-year-old seedlings inoculated under similar conditions. The clones were made 4 years before, using scion materials taken from parents that were about 26 years old. Thus at the time of inoculation, the clones were physiologically 30 years old. This evidence to an extent corroborates the interpretation by Zagory and Libby (1985) on tissue maturation effects on resistance among radiata pine clones to western gall rust. The variation in infection levels among spore sources in each clone can be explained by at least three factors. One factor could be spore quality related to viability after application on host tissue. A slight contamination of a spore source by some hyperparasite could have been magnified since spores were brushed on host surfaces then moistened and incubated at temperatures ideal for the germination of many fungal spores. Genetic differences among the spore sources and clones may also have accounted for the variation observed. Genetic factors could have influenced the ability of each spore source to cause viable infections, which in this system would be the formation of galls, which would eventually sporulate and cause further infections. Apart from the two uninfected clones, each of the four spore sources caused at least one infection on at least 10 of the remaining 14 clones. Most spore sources therefore differed only in virulence as measured by the proportion of the challenged tissue infected, or as the mean number of infections per shoot for each spore source (Table 5.17). The clone-spore source interactions observed in the experiment also explained, part of the variation since the virulence of a spore source relative to others depended on the Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 78 clones looked at, and the resistance of a clone relative to other clones also depended on the spore source. Furthermore, each spore source and clone might have interacted differently with the environment that was provided in the laboratory during inoculation and in the nursery beds in which the clones were grown. The significant variation in virulence among the four spore sources poses some prob-lems relating to the survival of spore sources of low virulence. Western gall rust according to the most recent evidence, seems to reproduce asexually. The rust may consist of a few separate races that occur in localised clonal populations. Source 3 produced less than half as many galls per shoot as any of the other sources in spite of the fact that the same number of spores which did not differ significantly in germination capacity were used in the inoculation. It is possible that source 3 might compensate for low virulence by high spore production. However, if one assumes that the number of spores produced by a spore line is proportional to the number of galls of that spore line, then source 3 should not survive very long since it cannot reproduce as rapidly as the other lines. Starting with the same number of galls for each spore line, the number of galls of source 3 would decline by 50% in each generation relative to the other spore sources, so that it would disappear (less than 1% of the other spore lines) in seven generations which would be considerably shorter than the life of a stand of lodgepole pine. However the interactions suggest that certain trees in the pine population or in neighbouring popultions may be very suscep-tible to source 3, in the manner in which clone 11 was susceptible to source 1. A large sample of clones might detect such trees. There may also be spore source-environmental interactions so that source 3 might yield much higher infection rates relative to the others under different conditions. The spore-clone interactions did not necessarily imply a classical major gene model but presented some evidence of major gene effects. More parents with larger numbers of clones per parent treated with more spore sources are required. However, the interactions Chapter 5. PINE CLONES AND SINGLE-GALL SPORE SOURCES 79 are important for the survival of lodgepole pine since a rust isolate highly virulent on one genotype may only be weakly virulent on another. In an earlier paper, Robinson (1979) has discussed various factors including controllable experimental error which can give non-genetic interactions and thus lead to misinterpretations. It is, however, not clear whether the demonstration of a typical vertical or a horizontal pathosystem is necessary for the explanation of stability. The parents represented by clones 5 and 7 have been tested at least three times in our nursery through their open-pollinated progenies and they seem to be consistently more resistant than the others. The influence of a few highly resistant genotypes in a natural pathosystem combined with less resistant ones is can lead to stability as was discussed by Browning (1979). The overall clone-parent regression was positive but similar regressions for individual spore sources varied and were much lower. This regression suggests that despite spore-clone interactions, tests on clones should be done using a bulked sample of spore sources to which a tree would normally be exposed. The main conclusions from this study are listed below : • The greater the physiological age of a shoot, the longer it takes for infections to show up as galls. • The four spore sources in this study showed overall differences in virulence rather than in pathogenicity. • There were interactions between spore sources and pine clones. Chapter 6 V A R I A T I O N A M O N G OPEN-POLLINATED PINE FAMILIES 6.1 Introduction In many genetic and selection experiments, progenies of open-pollinated parents have been used. Parents from a population can be sampled to represent different levels in the expression of a certain trait. The heritability of that trait can then be tested through open-pollinated progenies. In these tests, two things are usually important to a breeder and geneticist. The first is the degree to which half-sib or open- pollinated offspring resemble their particular parents. This relationship between sets of parents and their offspring can be tested using the method of parent-offspring regressions (Becker 1984, Falconer 1981). The slope of the regression line is often used to calculate narrow sense heritability (h 2) of the trait in question (Falconer 1981). The R value indicates the degree of correlation between parents and their offspring. The h 2 value can also be estimated from the ratios of family to total phenotypic variances as shown in Becker (1984). The second point is that in an open-pollinated situation, the female parent is allowed to mate with a random set of male parents. The degree to which such a parent combines with these male parents either to increase or decrease the expression of the chosen trait can therefore be estimated. This is what has been termed the general combining ability (GCA) of a parent (Griffing 1956a). For example, a tree showing a high level of rust resistance can be pollinated with a number of different male parents. If its progenies consistently show high levels of resistance, such tree has a high general combining ability 80 Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 81 for resistance. In a polycross, where a number of parents are involved, individual GCA values can be calculated for each parent (Griffing 1956a) after which the parents can be ranked accordingly. In general, GCA has been used for many years in the selection of seed parents. From open-pollinated parents, the variability of a trait and the GCA of the parents can be estimated. We can also estimate narrow sense heritabilities (h 2) through parent-offspring regressions and also through genetic variances derived from the expected mean squares in analysis of variance. The regressions can be used to judge whether parents can be selected on the basis, of their field appearance alone or whether progeny testing will be necessary. 6.2 Literature Review Variability in susceptibility to western gall rust has been demonstrated among individual trees. This has been shown over the years on at least two species of pine which the rust parasitizes. Some of the earlier studies of variability were on scotch pine (Pinus sylvestris L.) by Hutchinson (1935) and True (1938). These studies concentrated mainly on the histology of infection and addressed the connection between external stains on inoculated shoots and the nature of the underlying fungal and host tissues. Experiments on the variability in susceptibility to western gall rust using half-sib families have been reported for Pinus radiata (Old et al 1986, Vogler et al 1987) and for Pinus ponderosa (Hoff 1986). In the work reported by Old et al (1986) half- and full-sib seedlings were used and narrow sense heritabilities were estimated. Some studies on the susceptibility of lodgepole pine to gall rust also indicate a great deal of variation among individual trees within stands (van der Kamp 1981). Martinsson (1980) compared open-pollinated provenances by their relative susceptibilities to western gall rust in British Columbia and Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 82 the Yukon territory in Canada. There was variation among the provenances and also among members of the same provenance. Other than studies on variability, no major work has been done on the genetics of resistance to this rust in lodgepole pine. The aim of this study was to demonstrate and describe genetic variability among a much larger group of families than was used earlier in the symptom study, and to estimate narrow sense heritabilities. A second aim was to test the effects of 2 geographic spore sources of western gall rust on the families. Results from this experiment were considered important in revealing in part, the major patterns of inheritance of resistance to western gall rust. 6.3 Materials and Methods 6.3.1 Preparation of Seedlings A total of 40 parents were chosen in 1982 for this particular experiment. They represented the range of infection levels observed on their stands of origin. Of those, nine parents came from Stand A and 26 parents came from Stand B both near Prince George and described in Appendix A (Tables A.35 - A36). The remaining five parents came from Stump Lake (Appendix A.38) which is about 400 km South East of Prince George and some 40 km South of Kamloops. The number of galls on each parent was recorded. Open- pollinated seeds were collected from these trees in 1985. In April 1986, the seeds were pregerminated then transplanted into soil-filled styroblocks. The styroblocks had 80 cavities each. A total of 33 styroblocks were used and in each, each seedling family was represented twice. The planting pattern of the families in the first styroblock was completely random then the same pattern was kept the for all of the 33 blocks. The planted styroblocks were then set on nursery beds and watered regularly. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 83 6.3.2 Collection of Spores Two spore collections; one from the coast and the other from Prince George were made in the spring of 1987. Both spore lots were derived from a number of galls. The coastal mixture consisted of two equal volumes of spores from Lighthouse Park in Vancouver and from Richmond Nature Park. The other mixture came from the two stands from Prince George described above. Before inoculation, the two spore collections, from the coast and Prince George, weighing 2 g each were mixed with equal weights of fine ground and washed silica to aid act as a diluent and to facilitate handling. 6.3.3 Inoculation of Seedlings The one-year-old seedlings had flushed by the third week of May and were ready for inoculation. The 33 styroblocks were divided into two equal groups of 16 and the 1 extra block was used as the uninoculated control. Germination tests on 2% water agar were conducted on the two spore mixtures 24 hours prior to inoculation. Inoculations then proceeded in the chamber as described in section 2.2 of chapter 5. Since only 2 styroblocks could go into the chamber at a time, the spore silica mixtures were divided into 8 portions of 0.5 g. For each 0.5 g portion, 2 runs of 0.25 g each were made. After the first run, the positions of the styroblocks in the chamber were changed before the other 0.25 g of spores were applied. Spore deposition was monitored by placing horizontally and vertically oriented vaseline-coated cover slips among seedlings. After a complete run, the seedlings were misted with distilled water, placed inside clear plastic tents and incubated at room temperature for 48 hours then transferred to the nursery. The bases of the styroblocks were buried in nursery beds. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 84 Table 6.19: Expected mean square table for the ANOVA model on the 40 open pollinated lodgepole pine families inoculated with two western gall rust spore sources; coastal and interior •_• S O U R C E D F E X P E C T E D M E A N S Q U A R E SPORE (S) FAMILY (F) FAMILY x S PORE (F*S) E R R O R (E) 1 39 39 1721 <TE + C4(T|,f -f C 5 t r | °E + C2<Ts*F +C3aF 6.3.4 Observations and Analysis of Data The first galls on inoculated seedlings appeared 12-14 weeks from inoculation. The final assessment was made in spring of 1988, 48 weeks from inoculation. Shoot or candle length, the number of galls per seedling and % infected by family were recorded. Since there were a number of seedlings without infections, the variable; number of galls per seedling was transformed as logw(xi + 0.5) to normalise residuals. Spore and seedling family effects were considered random. Analysis of covariance was performed with shoot length as the covariate, following the linear model below. Yijk ^fi + Ai + Bj-r ABij + Eijk (6.4) where; Yijk is the number of galls per tree transformed as log (a;,- + 0.5), fi the general mean, A{ the effect spore source i; Bj the effect of the j th family, AB^ the effect of the interaction between spore source i and the j th family, and Eijk is the error term. The expected mean squares for the above ANOVA model are given in Table 6.19 where C\ to C 5 are the coefficients for each component in the table. Seedling family and spore source were assumed random. Regressions of offspring on the 40 parents were conducted to relate both the % of progeny infected and the mean number of galls per family as the dependent variables to Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 85 the number of galls on respective female parents. Regressions of offspring on parents were also performed using 10 of the 16 clones described in chapter 5. These 10 clones were derived from parents which were among the 40 open-pollinated parents used in this experiment. The independent variables were the number of galls counted on the parents in the field and also the mean number of galls per shoot on their respective clones. The dependent variables were the % of progeny infected and the mean number of galls per family. Comparisons of R 2 values between regressions of offspring on field parents and regressions of offspring on clones of the same parents were made. Infection frequency distributions were also constructed for each combination of spore source and open-pollinated family. Two histograms combining all families were con-structed for each of the 2 spore sources. 6.3.5 Estimates of Narrow Sense Heritability (h 2) To eliminate the effect of shoot length on the number of galls per seedling, two things were done. Covariance analysis was performed as described. The common slope of the regression was then used to adjust each value. A SAS program for the estimation of variance components in an unbalanced design was used since there were slight variations in the numbers of seedlings per family. Individual tree heritability (hj) was estimated as below: h2 = 4-o-j. 1 0-1 + O-LiT + 0~F where, hj is the (h 2) based on individual tree data, <J2F is the family component, o~\^F is the spore source by family interaction component and o~2E is error including non-additive genetic and environmental components of variance. The numerator of the fraction is Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 86 multiplied by 4 since the family component of half-sibs estimates 1/4 of the total additive genetic variance. Family heritability hF was estimated using the general formula below : 2 .2 _ ° > F °A + C 2° L r + al C3 ^  C3 ^ aF where C2 and C3 the coefficient for the o~2SxF and Cz<rF variance components. The standard errors for both the individual and family estimates of h 2 were estimated in SAS according to the approximation formula in McCutchan et al, (1986). Since two spore sources were used in the experiment, the ANOVA model was split into two identical 1-way layouts so that two hj estimates were made to investigate the variation of h 2 between the two spore sources. The formula was simply : 2 _ Aa2F a F -r crE The standard error for the one-way layouts were estimated by computing the variance of an F-distribution ( Mood et al, 1974). Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 6.4 Results 87 Mean germination rates with their standard deviations after 12 hours for the coastal and interior spore sources were 66% ± 5 and 60% ± 7 respectively. Spore deposition rates after each run averaged at 35 ± 12 spores per mm 2 of the cover slips. Analysis of variance of galls per seedling (transformed) indicated significant spore source by family interactions (Table 6.20). The coastal spore source appeared more virulent than the interior source in 33 of all the 40 farnilies tested. The range of infection levels was wider with the coastal than with the interior spore source (Fig. 6.9, Table 6.21). Coastal spores also caused more infection than interior sources in terms of % infected and the mean number of galls per seedling (Table 6.21 and Fig. 6.9). At high levels of infections as those caused by coastal spores, the number of galls per seedling gave a better separation of families by their relative levels of resistance than percent infected. The measure of resistance as percent infected per family, ranged from 53 — 100% for coastal spores as opposed to 35 — 100% for interior spores. Table 6.20: ANOVA of the log of galls per seedling; shoot length analysed as a covariate and based on 40 open-pouinated lodgepole pine families inoculated with two western gall rust spores sources; coastal and interior SOURCE DF SS MS F-STAT F-PROB SPORE 1 18.765 18.766 45.034 0.001 FAMILY 39 31.709 0.813 1.951 0.050 SPORE * FAMILY 39 16.253 0.417 2.558 0.050 ERROR 1720 280.402 0.163 The main effects in the table were tested for significance using the interaction term. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 88 Coastal Interior Spore Source Figure 6.9: Plots of mean number of galls per family (adjusted for shoot length) for 40 open-pollinated lodgepole pine families inoculated with two geographic spore sources of western gall rust; one from the coast and the other from the interior of British Columbia. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 89 Table 6.21: The relative susceptibility of 40 open-pollinated lodgepole pine families from the interior of British Columbia to coastal and interior spore sources. Susceptibilty was expressed as % infected and the number of galls per seedling Family Galls On Parent Tree n Coastal Spores Interior Spores % Infected Galls1 n %Infected Galls1 1(C3) 3 18 100.00 15.11 20 90.00 6.25 2(B-9) 0 17 100.00 14.00 19 73.68 5.42 3(A-11) 6 22 90.90 15.90 26 96.15 7.92 4(A-4) 0 25 88.00 10.84 29 100.00 6.86 5(B-26) 0 29 96.55 12.89 30 80.00 4.80 6(A-10) 0 25 92.00 10.68 21 80.95 5.43 7(C4) 2 24 91.67 9.63 29 96.55 6.69 8(B-4) 0 27 62.96 5.88 26 80.76 4.73 9(B-11) 0 21 90.47 9.33 19 89.47 5.47 10(A-2) 27 22 100.00 12.81 26 88.46 6.15 ll(B-29) 0 22 59.10 5.18 25 84.00 3.76 12(B-28) 0 24 100.00 11.50 27 96.29 5.70 13(B-18) 0 23 86.95 8.00 26 96.15 7.96 14(B-1) 0 25 96.00 10.12 23 73.91 3.87 15(B-19) 11 23 91.30 10.34 25 92.00 5.48 16(B-20) 0 30 90.00 10.73 22 77.27 5.18 17(B-15) 108 26 96.15 11.69 29 93.10 6.96 18(B-24) 0 25 96.00 11.80 29 86.21 7.21 19(A-8) 13 30 96.67 12.33 30 93.33 5.77 20(B-21) 11 30 93.33 12.90 30 93.30 6.23 21(B-13) 33 25 96.00 12.80 23 91.30 7.78 22(B-16) 0 26 100.00 10.23 24 95.83 5.12 23(B-5) 77 24 87.50 7.67 23 95.65 6.26 24(B-8) 42 20 100.00 10.05 24 88.89 5.22 25(A-9) 11 18 94.44 9.83 21 100.00 6.19 26(B-14) 1 12 75.00 5.58 21 95.23 3.62 27(B-17) 112 21 95.23 8.95 22 81.82 4.68 28(B-23) 27 28 53.57 4.50 22 81.82 3.68 29(A-12) 0 25 76.00 7.24 17 35.30 1.24 30(C-5) 14 17 82.35 • 8.29 20 85.00 3.90 31(C-2) 30 23 95.65 8.17 12 66.67 4.42 32(B-6) 5 8 100.00 10.50 5 100.00 11.20 33(B-27) 0 28 75.00 6.96 23 86.95 4.61 34(B-12) 0 25 92.00 11.62 24 91.67 6.04 35(A-1) 0 22 90.09 11.19 18 72.22 4.56 36(B-2) 16 27 100.00 12.11 23 86.96 6.87 37(B-15) 108 23 95.65 12.32 29 79.31 3.83 38(A-6) 0 16 81.25 7.53 19 78.95 4.53 39(B-25) 17 22 90.10 8.57 23 82.61 4.78 40(C-1) 16 5 100.00 12.60 7 100.00 5.71 Mean galls per seedling Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 90 The forty families separated into two rather distinct groups. Six families (8, 13, 23, 28, 32 and 33) had nearly equal infection levels with both interior and coastal spore sources (Fig. 6.9). The other 34 families showed higher infection levels with coastal than with interior sources. The average infection levels for the two groups were about equal for interior spores but for coastal spores, the group of six families showed a much higher average resistance than the remaining 34 families. The interactions in the ANOVA (Table 6.20) can therefore be attributed to the behaviour of the two groups of families. The parent-offspring regressions performed for each of the 3 stands and summarised in Tables 6.22 and 6.23 were mostly non-significant. The regressions of open-pollinated offspring on the 10 parents used both in the clonal and this experiment were also non-significant. However, regressions of open-pollinated offspring on the clones were all positive and much higher. The regression of mean galls per open-pollinated family on the mean number of galls per clone gave the highest positive and significant regression (#2=0.61, p=0.05) for coastal spores. The R2 value for interior spores was 0.01. Regres-sions of percent infected per open-pollinated family on galls per clone gave non-significant but positive R2 values of 0.28 and 0.13 for coastal and interior spores respectively. The resistance frequency distributions caused by coastal spores had a large zero class but otherwise approached normality with a tail to the right (Fig. 6.10). The distribution produced by interior spores was somewhat skewed toward low infection levels. A set of 80 frequency histograms; 40 for each spore by family combination, produced a variety of shapes. For coastal spores, 24 families were near- normal, 8 were bimodal, 1 was skewed toward low infection and 7 had flat or irregular distributions. For interior spores, only 11 families showed normal distributions, 15 were skewed toward low infection, 7 were mixed, 1 was skewed toward susceptibility, 5 were flat or irregular and one was ignored because it had only 5 seedlings. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 91 Table 6.22: Summary of regressions of offspring on parents based on 40 open-pollinated lodgepole pine families inoculated with western gall rust from the coast of B. C. The dependent variables were; % offspring infected and mean number of galls per seedling. Stand Error DF F-value Significance (PR>F) R 2 A1 7 8.31 0.023 0.540 A2 7 1.43 0.271 0.161 B1 24 1.79 0.194 0.069 B2 24 0.93 0.345 0.040 C1 3 0.00 0.00 0.000 C2 3 1.12 0.368 0.271 1(%) Infected ; 2Mean number of galls per seedling Table 6.23: Summary of regressions of offspring on parents based on 40 open-pollinated lodgepole pine families inoculated with western gall rust from the interior of B. C. The dependent variables were; % offspring infected and mean number of galls per seedling. Stand Error DF F-value Significance (PR>F) R 2 A1 7 0.56 ' 0.478 0.074 A2 7 0.20 0.665 0.028 B1 24 0.00 0.949 0.0002 B2 24 0.00 0.962 0.0001 C1 3 3.72 0.149 0.554 C2 3 3.49 0.159 0.538 Infected ; 2Mean number of galls per seedling Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 92 Treatment with Coastal Spores (Spore 1) (n =903) 12 0 1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829 Treatment with Interior Spores(Spore 2) (n =898) 14 r— — 0 1 2 3 4 5 6 7 6 9 1011121314151617181920212223242526272629 Number of Galls Per Seedling Figure 6.10: Frequency histograms illustrating the distribution of lodgepole pine seedlings based on the number of galls per seedling, following the inoculation of 40 open-pollinated lodgepole pine families with two spore sources of western gall rust; from the coast and from the interior of British Columbia. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 93 Heritability estimates (Table 6.24) had fairly small standard errors. The estimate for family heritability was higher than that for individual trees whereas, the two hj and hp estimates for each spore source were significantly different. Table 6.24: Estimates of Narrow Sense Heritabilities (h 2) from forty open-pollinated lodgepole pine families treated with two spore sources of western gall rust, one from the coast and the other from the interior of British Columbia. HERITABILTY (h 2) ESTIMATE STD. ERROR /^(Individual)0 h 2F(Family) b /^(Coastal) 1 /^(Coastal) 1 h 2 (Interior) 1 hp (Interior) 1 0.21 0.51 0.58 0.74 0.31 0.64 0.10 0.16 0.14 0.06 0.10 0.08 ° and b are estimates based on variance components from the two way ANOVA, whereas 1 denotes estimates of individual and family h2 based on the one-way ANOVA layouts for each spore source. Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 94 6.5 Discussion 6.5.1 The Inoculation Technique In the symptom study (in 1985), progeny from ten open-pollinated trees were treated with coastal spores. In 1987, nine of these trees were among the 40 parents whose open-pollinated progenies were divided into two groups, one treated with coastal and the other with interior spores. The spore load used in 1985 was 1.0 g and in 1987 0.25 g. Simple correlations, using both percent infected and the mean number of galls per family were performed between the treatment in 1985 and the two in 1987. There was only one significant correlation with an r value of 0.83, i(df-7)=0.79, p=.01. This correlation was between the % infected in 1985 and 1987 treatments in which coastal spores were used. The r value when the % infected on seedlings treated with coastal spores in 1985 and those treated with interior spores in 1987 were correlated was 0.53 and was not significant (p=0.05). This demonstrates that using the same spore provenance, the inoculation procedure can give reproducible results. The much lower correlations between infections caused by coastal and those caused by interior spores may be evidence of virulence differences between coastal and interior spores. 6.5.2 Variation among the 40 Open-pollinated Families The interactions (Table 6.20) appeared to be caused in part by the six families that appeared to be equally susceptible to the two spore sources as compared to 34 other families that were more susceptible to the coastal spore source (Fig. 6.10). It is possible that the six families carried resistance mechanisms (and therefore genes) that were as effective against interior spore sources as the other 34 families. The same genes or different genes carried by the six families are much more effective against coastal spore sources than those carried by the other 34 families. The genes for resistance to Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 95 coastal spores in the six families may have arisen through migrations from coastal pine populations without being accompanied by virulent coastal spore sources. They may also have arisen locally through selectively neutral mutations. One could also suggest that the 34 pine families in the interior have coevolved with their local rust populations and have accordingly accumulated resistance genes to them but not to coastal rust populations (van der Kamp 1988a). Experiments by van der Kamp (1988b) showed that coastal spore sources caused higher infection levels on interior pine provenances than on coastal provenances. Furthermore, coastal pine provenances were more resistant to interior spores than interior provenances. It is however not easy to explain why coastal spore sources apparently carry genes enabling them to be highly virulent on interior spore sources whereas interior spores sources have not shown the same behaviour. The use of percent infected as a measure of relative resistance between families can be misleading when high inoculum loads are used as is common in laboratory inoculations. This was illustrated in the high percentages of infection by coastal spores in this study and corroborates van der Kamp (1988b) in which galls per tree was a more sensitive measure of resistance between families when high proportions of seedlings were infected. Two families treated with the same amount of spores can show nearly the same proportion of their members infected but if one family has three times the mean number of infections per seedling as the other, the two families most likely have different levels of resistance. At high inoculum levels in artificial inoculations, the probability of infection for each seedling is increased to levels well above those in the field. At this point even normally resistant seedlings may develop one or two infections. Thus if families are compared mainly by their proportions infected, there be will a tendency to underestimate the resistance of some families. Narrow sense heritabilities based on the number of galls per seedling were substantial, especially when calculated for interior and coastal spores separately. This indicated that Chapter 6. VARIATION AMONG OPEN-POLLINATED PINE FAMILIES 96 additive inheritance plays a role in this pathosystem. However the presence of a few major genes segregating at various loci can also give additive genetic variance. For that reason the heritability values do not negate the influence of major genes. In this study using a much larger number of parents than the 10 families described in chapter 4, parent-offspring regressions were still poor. This was so because some parents which had remained gall free for a number of years in the field produced offspring which were as susceptible as those from infected parents. This has been a common result before and during the course of this entire study and some other observations may explain it. The clones described in chapter 5 showed that 3 out 5 from uninfected ortets yielded susceptible ramets. The two clones which remained uninfected also yielded the most resistant open-pollinated offspring (Table 6.21). It follows that a number of uninfected parents may merely be escapes. Some trees may also have escaped infection because they flush earlier than others resulting in tissue maturation during spore release. Factors other than genetically determined resistance must play a major role in the degree of infection in the field. In most cases the sources of inoculum that infect trees in nature are not the same sources used in the artificial screening of the progenies of selected parents. This may also partly explain why natural infection levels of trees are not reflected in the artificial inoculations of open-pollinated progenies. The above paragraph suggests that to select superior parents efficiently, clones and/or open-pollinated offspring should be tested. This conclusion is supported by the fact that, the relationship between clonal performance and that of open-pollinated offspring was much better than between open-pollinated offspring and parents in the field. The data support the conclusion that the number of galls on a parent is a poor indicator of the performance of its clones and its sexual offspring performance. Chapter 7 I N H E R I T A N C E OF R E S I S T A N C E : A D I A L L E L CROSS 7.1 Introduction One of the many mating designs that has been used extensively in basic genetic studies and breeding of agronomic crops is the diallel (Sprague and Tatum, 1943). It was designed as a scheme in which single crosses are made between p inbred lines in all possible combinations such that, P 2 is the number of all possible crosses. Various modifications of the diallel, their appropriate statistical treatments and genetic interpretations have been described and discussed (Griffing 1956a and 1956b, Kempthorne 1956, Hayman 1954b, Baker 1978). Several genetic interpretations can come from the analysis of a diallel provided that, certain assumptions are met. These assumptions (Hayman 1954a) are: 1. that the parents used are homozygous and come from an inbreeding parental pop-ulation with no change in allele frequencies; 2. that nonallelic genes act independently (no epistasis); 3. that there is linkage equilibrium (independent distribution of genes among parents); and 4. no multiple allelism The presence of epistasis, which would violate assumption 2, leads to unreliable es-timates of both additive and dominance variance ( Matzinger and Kempthorne 1956). 97 Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 98 The lack of the independent distribution of genes will result in the over- or underesti-mation of dominance variance (Hayman 1954b). Subject to these assumptions, a diallel analysis can be an efficient method of estimating a number of genetic parameters includ-ing narrow sense heritabilities (h 2) in one generation of mating. General and Specific combining abilities of parents in hybrid combination can be estimated and equated with additive and non-additive genetic variances respectively. Maternal and reciprocal effects, the interactions of genotype and environment (Griffmg 1956, Baker 1978) and the mode of gene action can also be deduced from the same analysis (Hayman 1954b). All those estimates are indispensable in the understanding of how a trait is inherited whether in the wild or in artificial agrosystems. They can also partly explain how the stability of a trait like disease resistance can be maintained. The diallel cross has been used to a lesser degree in forest genetics and breeding than in agriculture (Boyle 1986, Kriebel et al 1972, Libby et al 1969, Morgenstern 1974, Mullin 1985, Samuel et al 1972). Relatively few diallel crosses have been used in the study of disease resistance and Wilcox (1982) is one of the few examples. In this study, a 4 by 4 diallel consisting of parents bearing various numbers of galls has been used to demonstrate genetic variability in resistance to gall rust and to investigate the main underlying patterns of inheritance. The natural pine-gall rust pathosystem seems stable and will likely continue to be so. Two hypotheses are introduced here which may explain the apparent stability. H0: Resistance in the natural pine-gall rust pathosystem is controlled by minor genes that act in an additive manner. Hi: Resistance in the pine-gall rust pathosystem is controlled by a few major and dominant genes, which, if present, impart considerable resistance, such that any popula-tion of pine seedlings can be divided into two groups, one resistant, and one susceptible. Hence the frequency of resistant and susceptible seedlings should approximate Mendelian Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 99 ratios. Using the diallel material, the significance of only General Combining Ability effects would suggest that resistance to gall rust can be explained by additive genetic variance. This would point to a polygenic mode of inheritance. On the other hand, if both general and specific combining abilities are significant, then we may have a mixture of additive and non-additive genetic effects. Both types of pathosystems have been described in chapter 2. 7.2 Materials and Methods 7.2.1 Preparation of Seedlings The parents which were used in the diallel cross came from the natural stand designated as B and described in the Appendix A (Table A.37). These were; B-l, B-2, B-4 and B-8 and were chosen to give two equal groups of resistant and susceptible parents. They had 0, 16, 0 and 54 galls respectively by the spring of 1982. In the spring of 1983, they were crossed in all possible combinations given as a 4 by 4 diallel (Table 7.25). In addition, open-pollinated seed was collected from each of the four parents. Cones were harvested and extracted in the summer of 1984. In the spring 1985, the seeds were pregerminated and planted in colour-coded containers. The number of seeds from selfed progenies were much lower than the others (Table 7.25). 7.2.2 Spore Collection and Inoculation During the third week of April 1986, fresh spores were collected from 2-year-old seedlings which had been artificially inoculated with a mixture of coastal spores collected from Richmond Nature Park. A total of 2 g was collected, sieved, and mixed with 8 g of washed fine-ground silica powder as a diluent so that each run would consume 1 g of Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 100 Table 7.25: A Diallel Mating Table : number of seedlings per cross FEMALE MALE WIND B-l B-2 B-4 B-8 B-l 23 42 36 13 46 B-2 45 2 33 23 41 B-4 39 38 13 44 41 B-8 10 0 16 0 39 the spore-silica mixture. The mixture was stored at temperatures of 3 — 5°C for a week before inoculation. By the first week of May 1986, the seedlings were ready for inoculation. Since all the seedlings could not go into the inoculation chamber at the same time, 8 inoculation runs were made. Seedlings from each mating were divided into groups of 4 to 8 to go into separate inoculation runs. In each inoculation run, all the matings, whenever possible, were represented. Appendix D.54 shows the number of seedlings per cross present in each of the 8 inoculation runs. The inoculations were done in the inoculation chamber described in chapter 5. At the end of each run, 5 candles were randomly picked to estimate spore deposition and germination rates. The seedlings were thereafter misted with distilled water, placed inside clear plastic tents and kept at room temperature (18 — 20°C) for 48 hours. After this, they were transferred to the nursery and planted. Seedlings from each of the 8 inoculation runs were planted into 8 completely randomized blocks. The numbers in each block and for each family per block were in most cases, unequal. The second week following inoculations, the number of shoots or candles and the total length of candles for each seedling were measured and recorded. The number of galls per seedling was recorded 40 weeks after inoculation. Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 101 7.2.3 Analysis of Data Since the variable; galls per seedling was not normally distributed due to the large zero classes in all of the four progeny groups, each data entry was log-transformed as, log(xi + 1) before analysis. Nevertheless, the large zero class renders any attempts at transformation somewhat ineffective. Using a general least squares method, analysis of covariance of a completely random-ized block design with unequal numbers per cell was performed with shoot length as a covariate. This first analysis enabled a traditional multiple comparison of family means from all the crosses using Duncan's multiple range test before performing a more detailed combining ability analysis. The 2-way linear model for the analysis was as below. Y(ij)k = fi + A{ + Bj + ABij + E{ij)k (7.5) Where; • Fjjfcis galls per seedling (adjusted for shoot length), • Ai is the effect of the ith cross (family), • Bj the effect of the jth block, • AB^ the interaction between the ith cross and the jth block, • E(ij-)k is the residual term. A more detailed genetic analysis with combining abilities, reciprocal, and maternal effects was performed using a general least squares method by Schaffer and Usanis (1969). This program was written for the analysis of diallel designs and handles unbalanced completely randomized block designs with missing values. The full linear model for analyzing the diallel is given below. Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 102 YijU =fi + Bi + GjiGk) + Sjk + Mj + Rjk +G*Bik + Eijkl (7.6) Where; • Yijki is the number of galls of the I th tree of the jk th cross in the i th block, • fi is the general mean, • B{ is the effect of the i th block, • Gj(Gk) is the general combining ability (GCA) of the jth female (k th male) line, • Sjk is the specific combining ability (SCA) for the j(k) ih cross, • Mj is the maternal effect for the j th parent, • Rjk is the reciprocal effect of the j{k) th cross, • G * Bik is the GCA by block effect, and • Eijki is random error. Since the parent trees in the diallel had been chosen to represent a range of field resistance, the genotypic effects of the parents were treated as fixed and the blocks as random. The expected mean square table was therefore constructed for a mixed effect ANOVA model as shown in Table 7.26. The diallel programme assumes that all the effects are random, but our genotypes were considered fixed. So to adjust the coefficients to account for the fixed effects, the matrix of coefficients produced by the programme was inverted and then multiplied by the vector of variance components to estimate a new set of variance components. Furthermore, the analysis was done such that statistically non-significant effects were deleted from the Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 103 Table 7.26: Expected mean squares for an ANOVA of the number of galls per per cross from a lodgepole pine diallel which was inoculated spores of western gall rust. Source of Variation DF Expected Mean Squares B L O C K 7 o% + K12<T2BL GCA 3 aE + KWO-Q^BL + KUO~GCA S E L F 3 cr | + if 80*G*BL + K9&SELF S C A 2 <rE + Kea-Q^BL + K7a2SCA M A T 3 a2E + KAa2G,BL + K5cr2MAT R E C P 2 t r | + K2O-Q^BL + K^O-^ECP G X B L 21 aE -f K\0~QXBL E R R O R 335 al linear model. An example would be the reduction of the full diallel table to a half diallel if reciprocal effects were not significant. Each cross would then have larger -numbers of seedlings used in the estimation of variance among seedling families. After the above analyses, selfs were excluded then values for general and specific combining abilities (GCA and SCA respectively) for each of the four parents and their crosses were estimated using the method by Griffing (1956a). 7.3 Resul ts There was a general skew toward low infection levels among the crosses of the diallel (Ap-pendix D.54). Despite the large zero class in most families (Appendix D.54), the variation in galls per seedling was continuous and there was no clear separation of seedlings into groups of resistant and susceptible. The 4 parents ranked as; B-2, B-8, B-l, B-4 in order of increasing resistance (Table 7.27). In chapter 6, it was shown that at high spore loads or high infection levels, the zero class became much smaller (Fig. 6.10) and resistance distribution approached normality. Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 104 The ordinary two-way analysis of covariance of the diallel (selfs excluded) using the number of galls per seedling indicated significant differences among crosses (P=0.05). Blocks had no significant effect. A test on the equality of variances showed that the variances were not statistically different. Based on a Duncan's Multiple Range Test, the means of each of the 5 reciprocal pairs were not significantly different. The analysis was therefore repeated after grouping reciprocal pairs to generate a half-diallel consisting of six crosses with no selfs. Families or crosses still remained significant (Table 7.28). They ranked in order of decreasing resistance as follows; B-l x B-4, B-2 x B-4, B-4 x B-8, B-l x B-8, B-2 x B-8 and B-l x B-2 . A summary on infections as % galled (Table 7.29) showed crosses involving parent B-4 with the highest resistance. The same trend was shown when the mean number galls per seedling (Table 7.30) was considered. Table 7.27: Infections (% infected, mean number of galls per family) caused by western gall rust spores used on the offspring of 4 open-pollinated lodgepole pine parents. The same parents were used to generate a 4 x 4 diallel.  Family n % Infected Mean Galls Std. Dev. B-l 46 72.0 1.98 2.14 B-2 41 69.0 2.96 3.59 B-4 41 37.0 1.00 1.53 B-8 39 72.0 2.59 2.73 Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 105 Table 7.28: Duncan's multiple range test1. Comparisons of mean number of galls per cross from a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust  FAMILY PARENTS n MEAN STD DEV 1 B-l * B-2 85 3.75a 3.00 2 B-l * B-4 75 1.03b 1.63 3 B-l * B-8 23 3.00ab 2.35 4 B-2 * B-4 73 1.18b 1.97 5 B-2 * B-8 23 3.17a 2.53 6 B-4 * B-8 60 2.35b 3.02 means followed by the same letters are not statistically different Table 7.29: Percent galled per cross1 in a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust. Reciprocal pairs lumped to generate a half-diallel table FEMALE MALE PARENTAL MEAN B-l B-2 B-4 B-8 B-l 35 76 41 86 59.5 B-2 - - 38 91 68.3 B-4 - 15 59 38.0 B-8 - - - - 78.0 selfs of parents B-l and B-4 are included since each has over 10 seedlings Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 106 Table 7.30: Mean number of galls per seedling for each cross in a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust. MALE FEMALE B-l B-2 B-4 B-8 MEAN B-l 1.13 3.38 0.86 3.84 2.30 B-2 4.02 1.50 1.76 3.17 2.61 B-4 1.18 0.63 0.31 2.78 1.23 B-8 1.90 - 1.18 - 1.54 Mean 2.06 1.83 1.02 3.26 1.97 Table 7.31 shows the full-diallel ANOVA table. GCA effects were highly significant at P=0.01. SCA effects were significant only at P=0.10. Since maternal and reciprocal effects were not significant, reciprocal pairs were grouped and the data was reanalyzed as a half-diallel, with each- cross now having more seedlings. The full ANOVA table was then reduced to Table 7.32 and again, GCA effects were highly significant at P= 0.01. For SCA effects, the critical-F value (p=0.05) was 3.0 while the actual-F value was quite close at 2.98. On putting confidence limits on the variance component for SCA effects, the interval did not include zero, suggesting that grouping of reciprocal sets had improved the precision. It was concluded that SCA effects were present. The SCA value in column 5 of Table 7.32 was about 1/3 the GCA value (same table). The SCA values were much closer to each other and the differences were not so obvious (Table 7.33). In Table 7.34, individual GCA values were clearly different. Parents B-4 and B-8, were the most and least resistant respectively and were good combiners but in opposite directions. Parent B-4 combined well for relatively high resistance while parent B-8 combined well for low resistance. Parent B-l which had no galls in 1982 but had at least 1 gall in 1986 did not differ significantly from parent B-2 which had 16 galls by 1982. The smallest value (-1.2593 in Table 7.34) indicates the lowest level of infection (parent B-4) while the highest value of 0.6574 was for parent B-8. Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 107 Table 7.31: Combining ability analysis; table of ANOVA and variance components using number of galls per seedling (per cross) from a diallel of lodgepole pine in which 1-year-old seedlings were inoculated with spores of western gall rust.  SOURCE DF SS MS VAR. COMP. STD DEV MEAN 1 1689.135 1689.135 - -BLOCK 7 38.904 5.558 -0.1363 0.1767 GCA 3 385.459 128.486** 0.7031 0.5986 SELF 3 72.915 24.305 1.4110 2.0691 SCA 2 45.870 22.935 0.3620 0.3898 MATR 3 58.929 19.643 0.0833 0.0672 RECP 2 8.097 4.049 -0.0566 0.0530 G*L 21 167.886 7.995 0.0598 0.1548 ERROR 335 2364.803 7.059 7.0591 0.5438 TOTAL 377 4832.000 ** significant at p=.01 Table 7.32: A reduced form of table 7.31 showing mean squares and variance components for GCA, SCA effects (selfs excluded). Combining ability analysis using the number of galls per cross of a lodgepole pine diallel in which 1-year-old seedlings were inoculated with western gall rust. SOURCE DF SS MS VAR. COMP. STD DEV MEAN 1 1726.327 1726.327 BLOCK 7 23.834 3.405 -0.1440 0.0507 GCA 3 400.609 **133.536 1.0092 0.8442 SCA 2 44.614 *22.306 0.3567 0.3798 ERROR 326 2439.615 7.483 7.4835 0.5844 * significant at p=.05; ** significant at p=.01 Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 108 Table 7.33: Estimates of Specific Combining Ability in a lodgepole pine diallel in which 1-year-old seedlings were inoculated with western gall rust.  FEMALE MALE B-l B-2 B-4 B-8 B-l - 0.4543 -0.2843 -0.1665 B-2 - - -0.1676 -0.2855 B-4 - - - 0.4532" B-8 - - - -Table 7.34: Individual GCA Estimates for the 4 parents in a lodgepole pine diallel in PARENT GCA RANK B-l 0.2854 2 B-2 0.3599 3 B-4 -1.2953 1++ B-8 0.6574 4+ is most resistant ;+ least resistant) Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 109 7.4 Discussion • The variation in infection within families was generally continuous; seedlings did not separate into clear groups of resistant and susceptible. The combining ability analyses show that: • Effects of GCA are more important than SCA effects at least for the 4 parents tested. • Two of the parents have high GCA values but express resistance in opposite direc-tions. • That SCA effects apparently exist but are relatively small. The relative resistance of the four parents as seen through their progenies from open-pollination was reflected in the relative resistance of the progenies in the diallel cross. According to the individual GCA estimates, the parents in the diallel were be ranked as; B-8, B-2, B-l and B-4 in order of increasing relative resistance. This only differed slightly from the ranking obtained from the open-pollinated seedlings of B-2, B-8, B-l and B-4 in the same order of resistance. This agreement between the two sets of results would suggest that the offspring reflected the relative resistance and ranks of their female parents. This relationship has not been significant as was evident in the poor parent-offspring regressions in chapters 4 and 6. The GCA and SCA effects should ideally be interpreted in terms of additive and non-additive genetic effects. Such direct interpretations would be highly restricted in the light of the assumptions concerning diallels mentioned in the introduction section (Baker 1978, Griffing 1956a, Gilbert 1958). However, estimates of SCA and GCA effects in the diallel were valid even if the assumptions were not fully met. The diallel cross had to Chapter 7. INHERITANCE OF RESISTANCE : A DIALLEL CROSS 110 be used so that tests of reciprocal and maternal effects could be made in addition to the that GCA effects were relatively more important than SCA effects in this sample of parents (Baker 1978). Consequently, it also indicated that additive genetic effects were likely more important than dominant genetic effects. In general, requirements such as the use of inbred lines in forestry are frequently violated while lack of epistasis cannot be easily assumed. Authors, especially those in tree breeding (Boyle 1986, Morgensten 1974, Mullin 1985), indicate that diallels are suitable not only for GCA and SCA estimates but for estimates of variance components including the components for selfing, reciprocal, and maternal effects. The question remains as to whether we can explain the natural stability with GCA effects alone which would imply the predominant role of additive genetic variance in the inheritance and subsequently, the stability of resistance of the pine stands sampled. The GCA and SCA estimates and their interpretetion in terms of population parameters is limited by the small sample size of parents and because the parents came from the extreme ends of the resistance distribution. For that reason, inferences drawn would be valid mainly for the sample of parents and not the population from which they came. Another limitation was the use of bulked spores because no single gall among those collected could have sufficed as the sole source of inoculum. This could have masked spore by family interactions. other regular ones. It should be noted that the ratio was 0.85. This indicated Chapter 8 CONCLUSIONS AND GENERAL DISCUSSION 8.1 General Conclusions The major conclusions of the study can be divided into four parts namely; methodology, the pathogen, the host and the pathosystem. On methodology, the following can be stated: • Early symptoms cannot be used to predict the resistance of families or individual trees. • The number of galls per tree or shoot was a more sensitive measure of resistance, and yielded analysable results over a wider range of infection intensity than percent infected. • One-dimensional SDS-PAGE showed some promise as a method for fingerprinting single-gall spore collections. With respect to the pathogen, the following are the major conclusions: • Coastal populations of E. harknessii reproduce asexually as was shown in the stud-ies on the cytology of immature and mature spores, and germ tubes. • Inoculation studies revealed some variation in virulence among pathogen prove-nances as well as within local pathogen populations. With respect to the host, the major conclusions are listed below: 111 Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 112 • There was considerable genetic variation within the Prince George host popula-tions with respect to early symptom development (variable symptom expression) following inoculation. • There is considerable genetic variation within the host with respect to resistance. • This genetic variation appeared to be largely additive. Therefore, breeding for resistance is possible as was shown by the estimates of narrow sense heritability and the combining ability analyses. • Non-additive genetic variance was detected and accounted for a small part of the to-tal genetic variance. The non-additive genetic variance component could have been caused by dominance effects and possibly by epistatic interaction among genetic loci. With respect to the pathosystem: • The wide variation among parents with respect to resistance and the possible major role of polygenic inheritance of resistance combined with little variation relative to the host in the pathogen can explain a major portion of the stability of the pine-gall rust pathosystem. This need not exclude the role of monogenic inheritance in certain portions of the pathosystem. 8.2 Discussion 8.2.1 Methodology Throughout this study, one-year-old or older seedlings were used rather than younger seedlings in their first year of primary and indeterminate growth. Upon inoculations using the same volume of spores, older seedlings will support many more infections per Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 113 seedling because they have much greater lengths of young stem tissue that is susceptible to infection. The number of galls per seedling can therefore be the yield variable when older seedlings are used, whereas on young seedlings, percent infected is the yield variable as more than one or two galls are uncommon or difficult to count. Furthermore some resistance reactions may not be active during primary growth in the first year. For that reason, it is recommended that at least one-year-old seedlings be used in such studies. Early symptoms following inoculations did not give a good prediction of gall forma-tion. For most of the conditions tested, gall formation and the presence of a symptom were independent. This was shown by the lack of significant correlations between symp-tom frequency and family infection levels (r values), and the lack of associations between symptom appearance and gall formation on individual seedlings (X 2 values). However, the symptoms were a result of inoculation with the rust since control seedlings did not develop such symptoms. Similar symptoms caused by fusiform rust on southern hard pines were interpreted as evidence of resistance and families which exhibited these symp-toms were considered more resistant (Lundquist et al 1982, Lundquist and Luttrell 1982, Lundquist 1984). Reports from Hiratsuka's group (pers. comm.) in Alberta, Canada support the view that some of these symptoms may indicate successful fungal penetration of host tissue. In this study a number of symptomatic seedlings formed galls. Asymp-tomatic seedlings also formed galls but to a lesser extent than symptomatic ones. One interpretation is that these symptoms are resistance reactions since they have been as-sociated with the formation of necrophyllactic periderms but that they fail to stop gall formation when infection conditions are optimum and spore loads are high. Another is that they are produced in response to host penetration and are not in themselves the product of active resistance. At the moment, the interpretation that they are a resistance reaction is supported mainly by their association with necrosis and periderm formation and because they are developed after inoculations. Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 114 The results of electrophoresis (Appendix E) were not considered conclusive because of the diffuse appearence of most polypeptide bands. However variation occurred only in a few regions which, may support the conlusion that the fungus has no sexual cycle. This leads to the interesting question about how populations of the rust can achieve genetic variability which is important for survival in a genetically and physically diverse host environment. As our evidence on germtube cytology suggests the lack of a sexual life cycle, the rust may consist of a few local dominant pathogen genotypes. This would be so because a new virulent genotype arising by mutation, pansexuality or migration can build up with little or no loss of its genetic make up if normal sexual segregation does not occur. Without sexual reproduction, parasexual phenomenon such as heterokaryosis may be one way in which the fungus may develop genetic variability. The mycelia of two adjoining but otherwise separate, and genetically different infections may occasionally exchange nuclear material. 8.2.2 The Pathogen - Nuclear Cycle and Genetic Variability The results of nuclear staining of spore chains and germ tubes were different from those of Hiratsuka et al (1969), but were close to those of Epstein and Buurlage (1988). A possibility which should be investigated is that Hiratsuka et al (1966) and Hopkin et al (1988), have sampled a western gall rust population distinctly different from those in British Columbia and also from California. In a recent paper, Hopkin et al (1988) reiterated, the earlier conclusion that the uninucleate cells of the germ tubes are func-tional basidiospores. This was based mainly on the observation of appresorial swellings on branched germ tubes which led them to believe that each cell of the germ tube may be capable of independent infection. Considering a number of questions already raised, the lack of evidence of karyogamy in spore chains and the variable numbers of nuclei in germ tubes, it seems inappropriate to base such conclusions on swellings which occur Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 115 quite frequently among fungi in culture. In this study, evidence of variability among pathogen spore sources was provided by the markedly different infection levels caused by coastal and interior bulked spore sources on interior trees (Chapter 6). Further evidence came from the pine clone-spore source interactions (chapter 5). The recent paper by van der Kamp (1988) (cited in the discussion in chapter 6 ) and the apparent difference in virulence between coastal and interior spores in this study presented evidence that some geographically separate spore sources may have evolved virulence genes to their local host populations without losing their virulence to distant host populations. A comparison of coastal and interior spore sources yielded the greatest distinction observed in this study. 8.2.3 The Host In discussing a stable natural pathosystem, we should note that stability is influenced to some degree by the overall abiotic and biotic environment of which it is a part. This environment includes microenvironmental variation within local systems which influence initial infection conditions and further disease development. Examples of such conditions are, crown size, shape, vertical crown position and aspect which may influence moisture and temperature conditions during spore germination. Others such as proximity to a virulent isolate or proximity to a heavily galled tree may also play a role. The problem is that some of these sources of variation may be difficult or impossible to reproduce in an experimental setting. The effects of such factors are therefore difficult to take into account when describing a stable natural pathosystem. Inoculation experiments are usually conducted on young seedlings under conditions which yield infection levels much higher than those observed in the field, possibly masking some sources of variation. Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 116 The genetic variability that has been demonstrated in the host will act in the com-plex environment described above. This variability has been shown independently using clones, open- and controlled-pollinated material. The occurrence of a number of distinct symptom types or combinations of symptom types shortly after inoculation can be in-terpreted to mean that host reactions and therefore resistance is controlled by a number of genetic mechanisms. This can be said without any opinion on the number of genes involved or their mode of action; whether specific, non- specific, additive or non-additive and so on. That the various symptom types occurred within a single open pollinated fam-ily further indicates intra-family variation. The diallel cross experiment, though based on quite a small sample of parents, showed that GCA ( due to additive genetic effects) accounted for much more of the total phenotypic variance than SCA effects. 8.2.4 The Mode of Inheritance - The core of stability The significance of GCA effects in the diallel suggests that any model of stability in the lodgepole pine-gall rust pathosystem is influenced by additive genetic variance. Coupled with the significant h 2 values, it provides indirect evidence of a polygenic system of inheritance. Consequently, we reject the hypothesis that a few major genes with dominant effects would solely describe the genetic structure in the pine-gall rust pathosystem and account for its stability. To illustrate a polygenic system, family B-4 and A-12 have shown consistent levels of resistance above others in at least two independent inoculations in which separate bulk spore collections were used. The resistance of these two parents can be interpreted as evidence of constant ranking. Furthermore, parents B-4 and B-8 showed high individual GCA values for high and low resistance respectively. Parent B-4 combined well with the other two parents for improved resistance but in its combination with B-8, the resistance was less than average. This shows some additive genetic effects but may also include some dominance. Parent B-8, which shows high susceptibility to the Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 117 rust, may at the same time possess other fitness attributes that will ensure its survival on the host population. In a natural stand it will pass on genes for susceptibility, thus allowing the rust to survive and reducing selection pressure on the pathogen by the host. However, one problem is that the small sample of parents in the diallel could not permit any firm conclusion that polygenic as opposed to monogenic inheritance is the major factor underlying the stability of the pathosystem. A different picture of the genetic structure of the host emerges from the work with clones and single gall inocula. So far, this experiment provided the strongest evidence for the presence of single or a few major genes resulting in clone-spore interactions. The two clones B-4 and B-12 were not infected at all. The two parents may have dominant effects as B-4 showed in the diallel experiment. This does not illustrate a classic gene-for-gene model but for the survival of both host and pathogen, it may have important practical implications for stability as was discussed earlier on. What was clearly depicted is that there is considerable genetic diversity in the host but less diversity in the pathogen and possibly several interacting loci which could be responsible for the clone by spore interactions in chapter 5. The evidence on inheritance of resistance showed that the lodgepole pine population most likely maintains resistance to western gall rust from one generation to the next through a polygenic in conjunction with a mono- or oligogenic system of inheritance. The evidence of mono- or oligogenic modes of inheritance was given by the SCA effects in the diallel cross and the clone-spore interactions. Consequently, resistance may also be passed on in a non-additive manner. One may also argue that non-additive effects could be caused by epistatic interactions among different loci or by dominant single genes or a few genes with high resistance to the prevalent isolates of the pathogen. Chapter 8. CONCLUSIONS AND GENERAL DISCUSSION 118 8.2.5 Pathosystem Stability The wide variation in resistance of lodgepole pine within stands and also between prove-nances is an important feature in host-parasite interactions. Evidence from both lodge-pole pine and western gall rust suggest that the highly variable host poulation interacts with a much less variable pathogen population. Since selection depends to a great extent on variation, lodgepole pine would respond to selection pressure imposed by the rust and other factors at faster rates than the rust would to similar levels of selection pressure. This would be an important strategy for the genetic stability of pine because there are at least 20 generations of the rust for every lifespan of a pine stand. Pine can therefore match the relatively more rapid changes in the rust by its quicker response to selection. The variability in the host seems to be regulated by both polygenic and mono- or oligogenic systems of inheritance. The stability of a polygenic system has been postulated for natural pathosystems by a number of authors (Nelson 1978, Person et al, 1980). There was also evidence of variation in the rust but it appeared to be relatively much less than in pine. In pine, the two clones which were not infected at all may represent genotypes with high resistance. Such trees can be important for the survival of host populations. Browning (1974, 1979) described and discussed the importance of resistance based on a predominantly polygenic host population mixed with a smaller sub-population of host genotypes carrying major genes or genes conferring high resistance to the most virulent local pathogen isolates. Such genes will protect a host population from decimation during conditions optimum for infection; a condition discussed by Robinson (1979, 1987). Bibliography [1] Allen, A. and Y. Hiratsuka. 1985. Artificial inoculation of young seedlings of lodge-pole pine with Endocronartium harknessii. Can. J. Bot. 63:1168-1170. [2] Anderson, G. W. and D. W. French. 1965. Western gall rust in the lake states. For. Sci. 11: 139-141. [3] Anikster, J. and I. Wahl. 1979. Coevolution of the rust fungi on Graminacea and Lilieacea and their hosts. Ann. Rev. Phytopathol. 17:367-403. [4] Baker, R. J. 1978. Issues in diallel analysis. Crop Sc. 18:533-536. [5] Becker, W. A. 1984. Manual of quantitative genetics. 4f/lEd. Academic Enterprises Publ. ppl88. [6] Beckman J. S. and M. Soller. 1986. Restriction fragment length polymorphisms and genetic improvement of agricultural species. Euphytica 35: 111-124. [7] Blenis P. V. and H. D. Pinnel. 1987. Effect of inoculum concentration oi Endocronar-tium harknessii on the infection of lodgepole pine. Can. J. For. Res. 18:1123-1126. 119 Bibliography 120 [8] Borlaug, N. E. 1959. The use of multiline or composite varieties to control airborne epidemic diseases of self-pollinated crop plants. In First International Wheat Gen-eral Symposium Proceedings, Winnipeg, Manitoba. ppl2-26. [9] Boyle, T. J. B. 1987. A diallel cross in black spruce. Genome 29:180-186. [10] Browning J. A. 1974. Relevance of knowledge about natural ecosystems to develop pest management programmes for agrosystems. Proc. Am. Phytopathol. Soc. 1:191-199. [11] Browning J. A. 1979. Genetic protective mechanisms of plant-pathogen populations. Their coevolution and use in breeding for resistance. In Biology and Breeding for Resistance to Arthropods and Pathogens of Cultivated Plants. Texas Agric. Expn. Stn. Publ. 1451, 605 PP. [12] Burdon J. J. 1980. Variation in disease resistance within a population of Trifolium repens. J. Ecol. 68:737-744. [13] Burdon J. J. 1980. Intraspecific diversity in a natural population of T. repens. J. Ecol. 68:717-735. [14] Burdon J.J. 1982. The effect of fungal pathogens on plant communities. In the plant community as a working mechanism, ed. E. I. Newman 99-112. Oxford: Blackwell, Bibliography 121 128pp. [15] Burdon J. J. 1987. Diseases and plant population biology. Cambridge University Press.pp208. [16] Burdon R. D. and C. J. A. Shelbourne. 1973. The use of vegetative propagules for obtaining genetic information. N. Z. J. For. Sci. 4(2):418-425. [17] Burnett, J. H. 1975. Mycogenetics. An introduction to the general genetics of fungi. John Wiley and Sons. 375pp. [18] Bushnell, W. R. 1982. Hypersensitivity in rusts and mildews. In Plant infection: The physiological and biochemical basis, Y. Asada et al.(eds). Japan Sci. Soc. Press, Tokyo/Springer-Verlag, Berlin,pp. 97-116. [19] Chesson, A., J.J. Morgan, and R. C, Codner. 1978. Comparative electrophoretic study of proteins of Acremonium-like Hyphomycetes. Trans. Brit. Mycol. Soc. 70: 345-361. [20] Crill, P. 1977. An assessment of stabilising selection in crop variety development. Ann. Rev. Phytopathol. 15: 185-202. Bibliography 122 [21] Day , P. R. 1974. Genetics of host-parasite interaction. W. H. Freeman, San Fran-cisco, pp 238. [22] Dinoor, A. 1974. The role of wild and cultivated plants in the epidemiology of plant diseases in Israel. Ann. Rev. Phytopathol. 12:413-436. [23] Dinoor, A. and N. Eshed. 1984. The role and importance of pathogens in natural plant communities. Ann. Rev. Phytopathol. 22: 443-466. [24] Epstein L. and M. B. Buurlage. 1988. Nuclear division in germinating aeciospores and its taxonomic significance for the western gall rust fungus, Peridermium hark-nessii. Mycologia 80:235-240. [25] Falconer D. S. 1981. Introduction to quantitative genetics. 2nd. ed. Longman Group Ltd., London and New York, pp.340. [26] Fleming R. A., C. Person. 1982. Consequences of polygenic determination of resis-tance and aggressiveness in nonspecific host-parasite relationships. Can. J. Plant. Pathol. 4:89-96. [27] Flor, H. H. 1942. Inheritance of pathogenicity in Melampsora lini . Phytopathology. 32:653-669. Bibliography 123 [28] Frey K. J., J. A. Browning and M. D. Simmons. 1977. Management systems for host genes to control disease loss. Ann. New York Acad. Sci. 287:255-274. [29] Gabriel D. W. and A. H. Eliingboe. 1982. High resolution two- dimensional elec-trophoresis of protein from congenic wheat lines differing by single resistance genes. Physiologic. PI. Path. 20:349-357. [30] Gilbert, N. E. 1958. Diallel cross in plant breeding. Heredity 12:477-492. [31] Griffing, B. 1956a. Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. Biol. Sc. 9:463-493. [32] Griffing, B. 1956b. A generalized treatment of diallel crosses in quantitative inheri-tance. Heredity 10:31-50. [33] Groth, J. V. and C. 0. Person 1977. Genetic interdependence of host and parasite in epidemics. Ann. New York Acad. Sci. 287:97-106. [34] Harlan, J. R. and D. Zohary 1966. Distribution of wild wheats and barley. Science 153:1074-1080. [35] Hayman, B. I. 1954a. The analysis of variance of diallel tables. Biometrics, 10:235-244. Bibliography 124 [36] Hayman, B. I. 1954b. The theory and analysis of diallel crosses. I. Genetics 39:789-809. [37] Hiratsuka, Y. 1969. Endocronartium, a new genus for autoecious pine stem rusts. Can J. Bot. 47: 1493-1495. [38] Hiratsuka, Y. ,W. Morf and J. M. Powel. 1966. Cytology of the aeciospores and ae-ciospore germ tubes of Peridermium harknessii and P. stalactiforme of the Cronar-tium coleosporoides complex. Can. J. Bot. 44:1639-1643. [39] Hoff R. J. 1986. Susceptibility of pine populations to western gall rust-Central Idaho. USDA For. Serv. Inter. Res. Stn. Ogden, Utah; Res. Note INT-354. [40] Hopkin A. A., J. Reid, Y. Hiratsuka and E. Allen. 1988. Initial infection and early colonization of Pinus contorta by Endocronartium harknessii (western gall rust). Can. J. Plant Pathol. 10:221-227. [41] Howes, N. K. and W.K. Kim and R. Rohringer. 1982. Detergent soluble polypep-tides extracted from uredospores of four physiologic races of Puccinia graminis f. sp. tritici. Physiologic. Plant Pathol. 21:361-366. Bibliography 125 [42] Hutchinson W. G. 1935. Resistance of Pinus sylvestris to a gall-forming Perider-mium. Phytopathology 25:819-843. [43] Jackson, L. W. R.,and J. N. Parker. 1958. Anatomy of fusiform rust galls on loblolly pine. Phytopathology 48:637-640. [44] Jayakar, S. C. 1970. A mathematical model for interaction of gene frequencies in a parasite and its host. Theor. Pop. Biol. 1: 140-164. [45] Kempthorne, 0. 1956. The theory of the diallel cross. Genetics, 41:451-459. [46] Kim, W. K., J. W. Martens and N. K. Howes. 1984. Electrophoretic analysis of detergent soluble polypeptides of nine races of Puccinia graminis f. sp. avenae and their relation to P. graminis f. sp. tritici . Can. J. Plant Pathol. 6:111-118. [47] Kim, W. K. and K. Mortensen. 1986. Differentiation of Puccinia jaceae, P. centau-rea, P. acroptili and P. carihami by two-dimensional polypeptide mapping. Can. J. Plant Pathol. 8(3): 233-240. [48] Kinloch, B. B. Jr. and R. W. Stonecypher. 1969. Genetic variation in susceptibility to fusiform rust in seedlings from a wild population of loblolly pine. Phytopathology. 59:1246-1255. Bibliography 126 [49] Kriebel, H. B.,G. Namkoong and R. A. Usanis. 1972. Analysis of genetic variation in 1-, 2-, and 3-year old eastern white pine in an incomplete diallel cross experiments. Silvae Genetica 21: 44-48. [50] Leonard, K. J. 1969. Genetic equilibria in host-pathogen systems. Phytopathology 59: 1858-1863: [51] Leonard, K. J. 1977. Selection pressures and plant pathogens. Ann. New York Acad. Sci. 207-222. [52] Leonard, J. J. and R. J. Czochor. 1980. Theory of genetic interactions among pop-ulations of plants and their pathogens. Ann. Rev. Phytopathol. 18:237-258. [53] Libby W. J., Stettler R. F.and F. W. Seitz. 1969. Forest Genetics and Forest tree breeding. Ann Rev. Gen. 3:469-494. [54] Libby W. J. 1973. The use of vegetative propagules in forest genetics and tree im-provement. N. Z. J. For. Sci. 4(2):440-447. [55] Lowry, 0. H., Rosenbrough, N. J., Farr, A. L. and Randall, R. J. (1951). Protein measurement with Folin phenol reagent. Journal of Biological Chemistry 193: 265-275. Bibliography 127 [56] Lundquist, J. E., T. Miller and H. R. Powers Jr. 1982. A rapid technique for deter-mining resistance of slash pine to fusiform rust. Phytopathology. 72: 613-615. [57] Lundquist, J. E., and E. S. Luttrell. 1982. Early symptomatology of fusiform rust on pine seedlings. Phytopathology 72:54-57. [58] Lundquist, J. E. and T. Miller. 1984. Development of stem lesions on slash pine seedlings infected by Cronartium quercuum f. sp fusiforme. Phytopathology 74:514-518. [59] MacDonald, G. I. and R. J. Hoff. 1971. Resistance to Cronartium ribicola in Pinus monticola : Genetic control of needle-spots-only resistance factors. Can. J. For. Res. 1:197-202. [60] MacDonald, G. I. and R. J. Hoff. 1974. Resistance to Cronartium ribicola in Pinus monticola : an analysis of needle-spot types and frequencies. Can. J. Bot. 53:2497-2505 [61] Marshall, D. R. 1977. The advantages and hazards of genetic homogeneity. Ann. New York Acad. Sci. 287:1-20. [62] Marshall, D. R. and A. J. Pryor. 1978. Multiline varieties and disease control I. The "dirty crop" approach with each component carrying a unique single resistance gene. Bibliography 128 Theoretical and Applied Genetics 51:177-184. [63] Marshall, D. R. and A. J. Pryor. 1979 Multiline varieties and disease control. II. The "dirty crop" approach with each component carrying two or more genes for resistance. Euphytica 28:145-159. [64] Martin, B. 1987. Rapid identification of Rhizoctonia spp. associated with diseased turfgrasses. Plant Disease 71:41-49. [65] Martinsson, 0. 1980. Stem rusts in lodgepole pine provenance trials. Silvea Genetica 29:23-26. [66] McCutchan, B. G., J. Y. Ou and J. G. Namkoong. 1985. A comparison of unbalanced designs for estimating heritability in perennial tree crops. Theoretical and Applied Genetics 71:536-544. [67] Matzinger, D. G. and Kempthorne O. 1956. The modified diallel table with partial inbreeding and interaction with environment. Genetics 41:822-833. [68] Maynard Smith, J. and G. R. Price. 1973. The logic of animal conflict. Nature 246: 15-18. Bibliography 129 [69] Merril, C R., D. Goldman, S. E. Sedman and M. H. Ebbert. 1981. Ultrasensitive stain for proteins in polyacrylamide gels shows regional variation in cerebral fluid proteins. Science 27:1437-1438. [70] Mettler, L. E. and T. G. Gregg. 1969. Population genetics and evolution. Prentice Hall, Engelwood Cliffs, N. J. pp212. [71] Miller, T., E. B. Cowling, H. R. Powers Jr. and T. E. Blalock. 1976. Types of resistance and compatibility in slash pine seedlings to Cronartium fusiforme . Phy-topathology 66:1229-1235. [72] Mode, C. J. 1958. A mathematical model for the coevolution of onligate parasites and their hosts. Evolution 12:158-165. [73] Mode, C. J. 1961. A generalized model of a host-pathogen system. Biometrics 17:386-404. [74] Mood, A. M., Graybill F. A. and Boes D. C. 1974. Introduction to the theory of Statistics. 3rd Ed. pp 564. [75] Mullick, D. B. 1977. The non-specific nature of disease in bark and wood during wounding, insect and pathogen attack. Rec. Adv. in Phytochem. 11:395-440. Bibliography 130 [76] Mullin, T. J. 1985. Genotype-nitrogen interactions in full-sib seedlings of of black spruce. Can. J. For. Res. 15:1031-1038. o c [77] Nelson, R. R. 1978. Genetics of horizontal resistance to plant diseases. Ann. Rev. Phytopathol. 16:359-378. [78] Ouellete G. B. 1965. Cronartium stalactiforme on Pinus banksianae (Jack pine) in Quebec. PI. Dis. Reptr. 49:909. [79] Old, K. M., W. J. Libby, J. H. Russell and K Elridge. 1986. Genetic variability in susceptibility of Pinus radiata to western gall rust. Silvae Genetica 35:145-149. [80] Person, O, Fleming, R., Cargeeg, L. 1980. Non-specific interactions based on poly-genes. In disease resistance to diseases and pests in forestry trees. (Heybrook, H. M., Stephan, B. R., Von Weissenberg K. eds). Proc. 3rd Int. Workshop of genetics of host-parasite interactions in forestry, Wageningen, The Netherlands, Pudoc. p318-325. [81] Robinson, R. A. 1979. Permanent and impermanent resistance to crop parasites; A re-examination of the pathosystem concept with special reference to rice blast. Z. Planzenzuchtg. 83, 1-39. Bibliography 131 [82] Robinson, R. A. 1981. Ecological aspects of disease resistance. In plant disease con-trol : Resistance and susceptibility, ed. by R. C. Staples, G H. Toenniessen, p236-258. New York Wiley Publ. pp399. [83] Robinson, R. A. 1987. Host management in crop pathosystems. McMillan Publ. 263pp. [84] Ross, I. K. 1979. Biology of the fungi. McGraw-Hill Publ. 499pp. [85] Samuel, C. J. A.,R. C. B. Johnstone and A. M. Fletcher. 1972. A diallel cross in sitka spruce. Assessment of first year characters in an early glasshouse test. Theoretical and Applied Genetics 42:53-61. [86] SAS Institute. SAS user's guide: statistics. Cary, NC: SAS Institute; 1984. 584 p. [87] Segal, A., J. Manisterski, G. Fischbeck and I. Wahl. 1980. How plant populations defend themselves in natural ecosytems. In: J. G. Horsfall and E. B. Cowling (Eds.) : Plant disease V. How plant defend themselves. Academic Press, New York. p. 75-102. [88] Siegel I. H. 1976. Biochemical Calculations. 2nd Ed. J. Wiley and Sons Inc., New York. pp. 333-337. Bibliography 132 [89] Sprague, G. F. and L. A. Tatum. 1942. General versus specific combining ability in single crosses of corn. J. Am. Soc. Agron. 34:923-932. [90] Tomiyama, T. 1982. Hypersensitive cell death: Its significance and physiology. In Plant infection: The physiological and biochemical basis, Y. Asada et al.(eds). Japan Sc. Soc. Press, Tokyo/Springer-Verlag, Berlin, pp. 329-344. [91] True R. 1938. Gall development on Pinus sylvestris attacked by the Woodgate Peri-dermium and morphology of the parasite. Phytopathology 28:24-26. [92] van der Kamp, B. J. 1981. The incidence and impact of western gall rust, stalacti-form rust and atropelis canker in managed stands of lodgepole pine in the interior of B. C. In B. C. Min. For., Res. Branch, Victoria, B. C. 56pp. [93] van der Kamp, B. J. 1988a. Western gall rust in British Columbia. In proceedings of the thirty-fifth annual western international forest disease work conference; August 1988. pp 8-13. [94] van der Kamp B. J.1988b. Susceptibility of lodgepole pine provenances to geograph-ically separate western gall rust spore sources. Can. J. For. Res. 18:1203-1205. [95] Van der Plank, J. E. 1963. Plant diseases: Epidemics and control. Academic Press, New York. 349pp. Bibliography 133 [96] Van der Plank, J. E. 1968. Disease resistance in plants. Academic Press, New York 206p. [97] Van der Plank, J. E. 1978. Genetic and molecular basis of plant pathogenesis. New York: Springer 167pp. [98] Vogler, R. D., B. B. Kinloch, F. W. Cobb, Jr.,and T. L. Popenuck. 1987. Isozyme patterns in Endocronartium harknessii from several pine hosts and stands in Cali-fornia. Phytopathology 77:1242.(Abstract) [99] Wagener, W. W. 1964. "Facultative heteroecism" Was it demonstrated in Perider-mium harknessii in 1919-20?. Mycologia 56:782-84. [100] Wahl, I., A. Dinoor, J. Halperin, and S. Schreiter. 1960. The effect of Rhamnus palaestina on the origin and persistence of oat stem rust races. Phytopathology 50: 562- 567. [101] Wahl, I., 1970. Prevalence and geographic distribution of resistance to crown rust in Avena sterilis. Phytopathology 60:746-749. [102] Walkinshaw, C. H. 1978. Cell necrosis and fungus content in fusiform rust infected loblolly ,longleaf and slash pine seedlings. Phytopathology 68:1705-1710. Bibliography 134 [103] Watson, A. K., D. Schroeder, and I. Alkhoury. 1981. Collection of Puccinia species from diffuse knapweed in eastern Europe. Can. J. Plant Pathol. 3: 6-8. [104] Westergaard, M. and von Weittstein, D. (1972). The synaptinemal complex. Ann. Rev. Gen. 6: 71-110. [105] Wheeler, N. C. and W. B. Critchfield. 1985. The distribution and botanical charac-teristics of lodgepole pine: Biogeographical and management implications. In Lodge-pole Pine. The species and its management. Ed. by D. A. Baumgartner, R. G. Kre-bill, J. T. Arnott and G. F. Weetman. pp 1-14, [106] Wilcox, M. D. 1982. Genetic variation and inheritance of resistance to dothistroma needle blight in Pinus radiata. N. Z. J. For. Res. 12(1): 14-35. [107] Zagory, D. and W. J. Libby. 1985. Maturation-related resistance of Pinus radiata to western gall rust. Phytopathology 12:1443-1447. [108] Zalasky, H. and C. G. Riley 1963. Infection tests with two caulicolous rusts of jack pine in Saskatchewan. Can. J. Bot. 41:459-65. Appendix A Experimental Materials used in the Study A.l Stand A All the stands from which the experimental seed were collected are located at Buckhorn Lake (some 30 Km south east of Prince George area) with the exception of 5 parent trees which were located at a natural stand near Stump Lake near Kamloops, S.E. British Columbia. Stand A is a 13 x 13 spacing trial among many other spacing trials and composed of trees derived from a bulk seed collection from the Prince George Area. Compared to a local stand at Buckhorn Lake, it can be considered as a wider sample of the Prince George area. Tables A.35 show groupings of the 120 trees in Stand A and another 13 x 13 spacing replicate A 1 (102 trees) according to the amount of infection measured as the number of galls per tree. Table A.36 shows the actual parent trees from stand A that were used in the study. Below is a table showing the number of galls on each of the parent trees which were used in the study. 135 Appendix A. Experimental Materials used in the Study 136 Table A.35: A grouping by number of galls per tree (caused by western gall rust), of 120 lodgepole pine trees in Stand A (PG-I) and 102 trees from another 13 x 13 spacing trial (41) Infection Group Stand A Stand A1 Number Galls/tree 0 62 59 1 - 3 33 22 4 - 6 10 8 7- 9 5 2 10 - 12 3 3 i 12 5 6 Table A.36: Lodgepole pine parent trees from Stand A (PG-I 1-12) near Prince George which produced some of the seedlings used in the studies reported here. The number of galls caused by western gall rust were counted in 1984 and the stand average was about 10 galls Parent Number Total Number Of Infections A-l 0 A-2 27 A-3 25 A-4 0 A-5 18 A-6 0 A-7 10 A-8 13 A-9 11 A-10 0 A - l l 6 A-12 0 Appendix A. Experimental Materials used in the Study 137 A.2 Stand B Stand B is also located near Buckhorn Lake and is separated from stand A by a road. Stand B was regenerated naturally following a forest fire in 1961. In 1982, the number of galls per tree was counted by backdating over a ten year period for all the 370 trees in the Stand (van der Kamp, 1988). The average number of branch infections in galls per tree was 36 while the mean number of stem galls per tree was 0.44. For all the ten years, stem galls averaged at 1.3 % of all galls. The number of branch and stem galls on each parent from stand B used in this study is shown in the table A.37 Appendix A. Experimental Materials used in the Study 138 Table A.37: A table of lodgepole pine parent trees from Stand B (PG-II, III and IV) showing the respective number of western gall rust infections (galls) on each of the parents used in the studies of variability and resistance. The Stand average was 40 galls per tree Parent Progeny Galls Number Code Per Tree B-l PG-II A 0 B-2 PG-II B 16 B-3 PG-II C 4 B-4 PG-II D 0 B-5 PG-II E 77 B-6 PG-II F 5 B-7 PG-II G 1 B-8 PG-II H 42 B-9 PG-II I 0 B-10 PG-II J 0 B-ll PG-II N 0 B-12 PG-II 0 0 B-13 PG-III A 33 B-14 PG-III B 1 B-15 PG-III C 108 B-16 PG-III D 0 B-17 PG-III E 112 B-18 PG-IV A 0 B-19 PG-IV B 11 B-20 PG-IV C 0 B-21 PG-IV D 11 B-22 PG-IV E 2 B-23 PG-IV F 27 B-24 PG-IV G 0 B-25 PG-IV H 17 B-26 PG-IV I 0 B-27 PG-IV J 0 B-28 PG-IV N 0 B-29 PG-IV 0 0 Appendix A. Experimental Materials used in the Study 139 A.3 Stand C Stand C located near Stump Lake; South of Kamloops is also a naturally regenerated and heavily infected stand. Since only 5 parents were selected from it, it is not as important as the two stands from Prince George. The 5 trees and their individual infections by 1982 are in table A.38. Table A.38: A table of lodgepole pine parent trees from Stand C (Stump Lake) used an inoculation experiment in which a total of 40 families were inoculated with 2 spore sources of western gall rust. These parents came from a heavily infected stand Parent Number Total Number Of Infections C-l 16 C-2 30 C-3 3 C-4 2 C-5 14 Appendix B Analyses of Variance of the Frequencies of Early Symptoms 140 Appendix B. Analyses of Variance of the Frequencies of Early Symptoms 141 Table B.39: Analysis of arcsine-square root of % of seedlings of open- pollinated lodgepole pine families showing general red symptoms 2 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.5664 0.0629 5.5373 0.009 Inoculum Load 1 0.0002 0.0002 0.0147 0.873 Age of Seedling 1 0.2468 0.2468 21.7168 0.001 Fam. x Inoc. 9 0.0989 0.0109 0.9671 0.519 Fam. x Age 9 0.1254 0.0139 1.2259 0.383 Inoc. x Age 1 0.0502 0.0502 4.4152 0.063 Error 9 0.1023 0.0114 Table B.40: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red flecks 2 weeks following inoculation with 2 spore loads of western gall rust  Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.3185 0.0354 3.0832 • 0.054 Inoculum Load 1 0.0105 0.0105 0.9118 0.367 Age of Seedling 1 0.0085 0.0054 0.7439 0.415 Fam. x Inoc. 9 0.1420 0.0158 1.3749 0.321 Fam. x Age 9 0.1276 0.0142 1.2357 0.379 Inoc. x Age 1 0.0297 0.0129 1.1304 0.317 Error 9 0.1033 0.0115 Table B.41: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red streaks 2 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.1475 0.0169 0.6383 0.7434 Inoculum Load 1 0.0639 0.0169 0.6381 0.4498 Age of Seedling 1 0.0191 0.0191 0.7427 0.4152 Fam. x Inoc. 9 0.0898 0.0099 0.3888 0.9122 Fam. x Age 9 0.1533 0.0170 0.6633 0.7252 Inoc. x Age 1 0.0000 0.0000 0.0000 0.9476 Error 9 0.2311 0.0257 Appendix B. Analyses of Variance of the Frequencies of Early Symptoms 142 Table B.42: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing no symptoms 2 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.6218 0.0691 7.2489 0.004 Inoculum Load 1 0.0073 0.0073 0.7629 0.409 Age of Seedling 1 0.0777 0.0777 8.1570 0.018 Fam. x Inoc. 9 0.0932 0.0104 1.0868 0.452 Fam. x Age 9 0.2149 0.0239 2.5048 0.094 Inoc. x Age 1 0.0845 0.0845 8.6856 0.015 Error 9 0.0858 0.0095 Table B.43: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing general red symptoms 4 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.1869 0.0208 1.7582 0.2063 Inoculum Load 1 0.0686 0.0686 5.8023 0.0380 Age of Seedling 1 0.2927 0.2927 24.7782 0.0008 Fam. x Inoc. 9 0.0982 0.0109 0.9240 0.5460 Fam. x Age 9 0.0792 0.0088 0.7453 0.6661 Inoc. x Age 1 0.1145 0.1145 9.6900 0.0122 Error 9 0.1063 0.0118 Table B.44: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red flecks 4 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.8817 0.0979 10.2280 0.0012 Inoculum Load 1 0.0345 0.0345 3.6047 0.0876 Age of Seedling 1 0.0011 0.0011 0.1138 0.7386 Fam. x Inoc. 9 0.1722 0.0191 1.9980 0.1583 Fam. x Age 9 0.1610 0.0178 1.8583 0.1844 Inoc. x Age 1 0.2376 0.2376 24.8018 0.0008 Error 9 0.0862 0.0096 Appendix B. Analyses of Variance of the Frequencies of Early Symptoms 143 Table B.45: Analysis of arcsine-square root of % of seedlings of op en-pollinated lodge-pole pine families showing red streaks 4 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.1281 0.0142 0.6015 0.7702 Inoculum Load 1 0.1570 0.1570 6.6328 0.0290 Age of Seedling 1 0.1744 0.1744 7.3677 0.0231 Fam. x Inoc. 9 0.2676 0.0297 1.2562 0.3695 Fam. x Age 9 0.1926 0.0214 0.9040 0.5587 Inoc. x Age 1 0.0356 0.0356 1.5048 0.2503 Error 9 0.2131 0.0237 Table B.46: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing no symptoms 4 weeks following inoculation with 2 spore loads of western gall rust  Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.3743 0.0416 4.4615 0.0187 Inoculum Load 1 0.2008 0.2008 21.5258 0.0013 Age of Seedling 1 0.3640 0.3640 39.0500 0.0002 Fam. x Inoc. 9 0.0537 0.0059 0.6399 0.7422 Fam. x Age 9 0.1402 0.0156 1.6708 0.2277 Inoc. x Age 1 0.0133 0.0133 1.4228 0.2630 Error 9 0.0839 0.0093 Table B.47: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing general red symptoms 8 weeks following inoculation with 2 spore loads of western gall rust  Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.5669 0.0629 2.9845 0.0597 Inoculum Load 1 0.0121 0.0121 0.5745 0.4731 Age of Seedling 1 0.1614 0.1614 7.6461 0.0213 Fam. x Inoc. 9 0.2067 0.0229 1.0879 0.4510 Fam. x Age 9 0.1889 2.0988 0.9945 0.5032 Inoc. x Age 1 0.0018 0.0018 0.0847 0.7684 Error 9 0.1899 0.0211 Appendix B. Analyses of Variance of the Frequencies of Early Symptoms 144 Table B.48: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red flecks 8 weeks following inoculation with 2 spore loads of western gall rust  Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.46920 0.05213 3.2039 0.0492 Inoculum Load 1 0.00001 0.00001 0.0033 0.9100 Age of Seedling 1 0.06793 0.06793 4.1747 0^ 0692 Fam. x Inoc. 9 0.30340 0.03371 2.0718 0.1462 Fam. x Age 9 0.17851 0.01954 1.2008 0.3946 Inoc. x Age 1 0.44096 0.44096 8.8556 0.0152 Error 9 0.14645 0.01627 > Table B.49: Analysis of arcsine-square root of % of seedlings of open-pollinated lodge-pole pine families showing red streaks 8 weeks following inoculation with 2 spore loads of western gall rust  Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.2971 0.0330 1.2887 0.3555 Inoculum Load 1 0.3592 0.3592 14.0232 0.0046 Age of Seedling 1 0.0022 0.0022 0.0865 0.7664 Fam. x Inoc. 9 0.3532 0.0395 1.5320 0.2672 Fam. x Age 9 0.1888 0.0209 0.8188 0.6149 Inoc. x Age 1 0.0567 0.0567 2.2137 0.1686 Error 9 0.2306 0.0256 Table B.50: Analysis of arcsine-square root of % of seedbngs of open-pollinated lodge-pole pine families showing no symptoms 8 weeks following inoculation with 2 spore loads of western gall rust Source of Variation DF Sum of Squares Mean Squares F-Value F-Prob Family 9 0.6189 0.0688 6.6238 0.0052 Inoculum Load 1 0.1398 0.1398 13.4699 0.0052 Age of Seedling 1 0.0156 0.0156 1.5050 0.0902 Fam. x Inoc. 9 0.2378 0.2642 2.5447 0.2503 Fam. x Age 9 0.4460 0.0049 0.4773 0.8573 Inoc. x Age 1 0.1232 0.1232 11.8711 0.0073 Error 9 0.0934 0.0104 Appendix B. Analyses of Variance of the Frequencies of Early Symptoms 145 Table B.51: Contingency tables based on the presence of early symp-toms (stains) and gall formation on seedlings from 10 open-pollinated lodgepole pine families inoculated with western gall rust applied at 2 spore loads, and at the I s' of 2 stages of seedling maturity. The X 2 val-ues were computed for each of the 3 times that symptoms were observed and recorded. Infection Age Spore Time Symptom Type Row X 2 Load Obs. Red Red Red No Stain Total Value General Flecks Streaks - 1 1 1 55 (66.5) 0 0 31(19.5) 86 + 136 (124.5) 0 0 25 (36.5) 161 191 0 0 56 247 13.462 - 1 1 2 23 (25.4) 19 (23.3) 11 (15.0) 33 (22.3) 86 + 50 (47.6) 48 (43.7) 32 (28.0) 31 (41.7) 161 73 67 43 64 247 11.108 - 1 1 3 9 (11.5) 30(30.6) 11(15.0) 36 (28.9) 86 + 24 (21.5) 58(57.5) 32 (28.0) 47 (54.1) 161 33 88 43 83 247 5.142 - 1 2 1 51 (62.3) 0 0 46 (34.7) 97 + 107 (95.7) 0 0 42 (53.3) 149 158 88 246 9.461 - 1 2 2 21 (23.7) 19 (21.3) 13 (13.8) 44 (38.2) 97 + 39 (36.3) 35 (32.7) 22 (21.2) 53 (58.8) 149 60.0 54.0 35.0 97.0 246 2.406 - 1 2 3 11 (17.0) 17 (17.7) 9 (10.3) 60 (52.0) 97 + 32 (26.0) 28 (27.3) 17 (15.7) 72 (80.0) 149 43 45 26 132 246 5.763 Plus (+) and minus (-) signs denote infected (galled) and uninfected seedlings respectively. Contingency tables with no seedlings falling under the symptom category; red flecks or streaks are based on 1 instead of 3 degrees of freedom. *J/=i=3.84, Xj,=3=7.82, p=0.05. Appendix B. Analyses of Variance of the Frequencies of Early Symptoms 146 Table B.52: Contingency tables based on the presence of early symp-toms (stains) and gall formation on seedlings from 10 open-pollinated lodgepole pine families inoculated with western gall rust applied at 2 spore loads, and at the 2n d of 2 stages of seedling maturity. The X2 val-ues were computed for each of the 3 times that symptoms were observed and recorded. Infection Age Spore Time Symptom Type Row X2 Load Obs. Red Red Red No Stain Total Value General Flecks Streaks - 2 1 1 68 (76.7) 1(1.8) 1(1.8) 41(30.8) 86 + 104 ( 95.3) 3(2.2) 3 (2.2) 28 (38.2) 161 191 4 4 69 247 9.164 - 2 1 2 31 (40.6) 24 (25.4) 9 (11.6) 47 (33.4) 111 + 60 (50.4) 33 (31.6) 17 (14.4) 28 (41.6) 138 91 57 26 64 247 15.188 - 2 1 3 8 (10.3) 17 (21.4 14 (14.7) 72 (64.6) 111 + 15. (12.7) 31 (26.6) 19 (18.3) 73 (80.4) 138 23 48 33 83 249 4.094 - 2 2 1 142 (144.3) 0 0 47 (44.7) 189 + 45 (42.7) 0 0 11 (13.3) 56 187 88 245 0.653 - 2 2 2 79 (83.3) 43 (41.7) 21 (20.8) 46.0(43.2) 189 + 29 (24.7) 11 (12.3) 6 ( 6.2) 10.0(12.8) 56 108 54 27 97 245 1.967 - 2 2 3 6 (6.9) 41 (46.3) 3 (4.6) 139 (131.0) 189 + 3 (2.1) 19 (13.7) 3 (1.4) 31 (38.9) 56 9 60 6 132 245 7.767 Plus (+) and minus (-) signs denote infected (galled) and uninfected seedlings respectively. Contingency tables with no seedlings falling under the symptom category; red flecks or streaks are based on 1 instead of 3 degrees of freedom. XL=1=3.84, XJ/=3=7.82, p=0.05. A p p e n d i x C R e s i s t a n c e F r e q u e n c y D i s t r i b u t i o n s o f 10 P i n e F a m i l i e s Figure C.ll: (1-10).Frequency histograms of 10 open-pollinated lodgepole pine fami-lies inoculated with 2 spore loads of western gall rust at 2 stages of seedling maturity. Seedlings treated with both spore loads were combined and each family was described by the distribution of its members by infection classes c CD 3 ? a C s » Family 1 (A-1) at Age 1 Both Spore Loads Combined 1 2 3 4 6 6 7 1 0 10 Gals Per Seedling (n - SO) Family 2 (A-2) at Age 1 Both Spore Loads Combined 3 4 6 I ( 9 10 11 12 Galls Per Seedling (n • 51) Family 3 (A-10) at Age 1 Both Spore Loads Combined fe-c CD 2 3 4 5 6 7 3 0 Galls Per Seedling (n - 49) 10 11 12 Family 1 (A-1) at Age 2 Both Spore Loads Combined S 4 6 6 7 t » 10 11 Gals Per Seedling (n - 50) Family 2 (A-2) at Age 2 Both Spore Loads Combined » 10 11 12 Gate Per Seedling (n>50) Family 3 (A-10) at Age 2 Both Spore Loads Combined 0 I 2 3 4 ( 6 7 S « 1 0 11.12 Galls Per Seedling (n - 50) 147 Appendix C. Resistance Frequency Distributions of 10 Pine Families 148 Family 4 (A-4) at Age 1 Both Spore Loads Combined Family 4 (A-4) at Age 2 Both Spore Loads Combined & c a> 8 ! § • • • 1 J » 4 S ( 7 t « Gals Par Seedling (n • 49) 11 12 0 1 » » 4 * • 7 • » 10 11 U Gab Per Seedling (n-48) Family5(A-12)atAge1 Both Spore Loads Combined Family 5 (A-12) at Age 2 Both Spore Loads Combined Family 6 (B-1) Age 1 Both Spore Loads Combined Family 6 (B-1) at Age 2 Both Spore Loads Combined 1 1 TS] v Appendix C. Resistance Frequency Distributions of 10 Pine Families 149 c O) 3 c % te o ST to Family 7 (B-2) at Age 1 Both Spore Loads Combined Family 7 (B-2) atAge 2 Both Spore Loads Combined "T3 1 4 0 1 2 J « 5 t 7 ( s 10 11 It 0 1 2 3 4 6 C 7 3 0 10 11 12 Galls Per Seedling (n . 50) G a ) t t P e r Seedling (n . 49) Family 8 (B-3) at Age 1 Both Spore Loads Combined Family 8 (B-3) at Age 2 Both Spore Loads Combined 13 TBI 0 1 2 3 4 5 6 7 3 0 10 11 12 e i 2 3 4 S S 7 * « 1 0 1 1 1 2 Gab Per Seedling (n • 47) Family 9 (B-4) at Age 1 Both Spore Loads Combined Gals Per Seedling (n • 50) Family 9 (B-4) atAg e 2 Both Spore Loads Combined T7 T E 1 2 3 4 S S 7 t • 10 11 12 Galls Per Seedling (n - 50) Family 10 (B-5) atAge 1 Both Spore Loads Combined 1 j » 4 i 0 7 i » H > l 1 1 2 Gab Per Seedling (n > 46) Family 10 (B-5) at Age 2 Both Spore Loads Combined TS "2TJ 3 4 S • 7 • • Gals Per Seedling (n - 60) 3 4 I 6 7 • f Gab Per Seedling <n - 49) Appendix D A Diallel Table D.53: Number of seedlings per inoculation run in which progenies of a diallel cross of lodgepole pine were inoculated with spores of western gall rust I I INOCULATION RUN MALE FEM. 1 2 3 4 5 6 7 8 1 1 4 3 4 4 5 4 - -2 1 4 3 3 5 5 5 7 9 3 1 4 4 4 5 5 6" 7 3 4 1 4 5 1 2 4 4 6 5 5 5 7 7 2 2 2 3 2 3 4 5 4 5 5 7 1 4 2 1 3 4 4 4 5 5 5 6 7 2 3 4 3 4 4 5 6 7 8 3 3 3 4 5 4 3 4 4 5 5 5 6 7 8 1 4 4 2 4 2 4 4 4 4 8 2 - - -3 4 4 1 4 5 1 - - -4 4 Seedlings from each inoculation run were planted in the nursery as separate blocks. 150 Appendix D. A Diallel 151 Figure D.12: (a - r). Frequency histograms describing parents (a - d) and progenies of a diallel cross (e - r) of lodgepole pine by infection classes (galls per seedling), following inoculations with spores of western gall rust c D Distribution of Seedlings Parent B-1 (Open-Pollinated), n=46 Distribution of Seedlings Parent B-2 (open-pollinated), n=41 2 3 4 S 6 7 Number of Galls Per Seedlings 3 4 5 Number of Galls 6 7 s Per Seedling Distribution of Seedlings Parent B-4 (Open-pollinated), n=41 Distribution of Seedlings Parent B-8 (Open-pollinated), n=39 1 2 3 4 5 6 7 8 Number of Galls Per Seedling 10 11 2 3 4 5 6 7 B 9 Number of Galls Per Seedling 10 Appendix D. A Diallel 152 Distribution of Seedlings Female B-1 x Male B-4, n=43 >. CJ c CD 1 2 3 * 5 6 7 Number of Galls Per Seedling Distribution of Seedlings Female B-1 x Male B-1, n=23 1 2 3 4 5 6 7 0 Number of galls per seedling Distribution of Seedlings Female B-2 x Male B-1, n=38 ST 3 4 5 6 7 Number of Galls per Seedling Distribution of Seedlings Female B-4 x Male B-2, n=33 & e 8" §1 2 3 4 S 6 7 Number of Galls Per Seedling Distribution of Seedlings Female B-1 x Male B-8, n=14 •A Si c p s An i 2 3 4 5 6 7 8 Number of Galls Per Seedlings Distribution of Seedlings Female B-1 x Male B-2, n=49 w SI man f ke 1 2 3 4 5 6 7 0 Number of Galls per Seedling Distribution of Seedlings Female B-2 x Male B-4, n=38 2 3 4 5 6 7 6 Number of Galls Per Seedling Distribution of Seedlings Female B-4 x Male B-1,n=32 2 3 4 5 6 7 Number of Galls Per Seedling Appendix D. A Diallel 153 Distribution of Seedlings Female B-4 x Male B-4, n=13 Distribution of Seedlings Female B-4 x Male B-8 n,16 60 60 & c CD w 20 • 20 1 2 3 4 S 6 7 8 Number of Galls Per Seedling Distribution of Seedlings Female B-8 x Male B-1, n=9 2 3 4 S 6 7 « Number of Galls Per Seedling Distribution of Seedlings Female B-8 x Male B-4, n=44 1 2 3 4 5 6 7 Number of Galls Per Seedling 2 3 4 5 6 7 Number of Galls Per Seedling Appendix E Electrophoretic Analysis of Rust Collections E.l Results A typical electrophoretic profile of total detergent soluble protein of this rust had about 30 major bands (Fig E.13). Most bands were confined to the upper two thirds of the gel, falling within the 42 to the 200 kilodalton (50-200 kDA) range. Four spore sources from the 1987 collections which were used in' the inoculation of pine clones in another experiment (chapter 5) showed some differences in the 36, 45 and the 97 kDA regions (Fig. E.13). One source was cut off because of difficulty in resolution. At a glance, the 3 spore sources had quite similar protein profiles but lanes bi and b2 show a band at about the 97 kDA region which the others did not show. Lane c± and c2 and also did not have the strong doublets (double bands) at the 42 kDA region which the other ones show. Other regions in which differences occurred are marked by arrows. Figure E.14 shows profiles of 3 spore collections run in duplicate. Lanes c and d are is a source from Prince George, e and f is from Richmond Nature Park and g and h is from Lighthouse Park. The source from Prince George does not have the prominent band (about 195 kDA) shown by the e,f and g,h pairs. Further down at about the 95 kDA, the g-f pair show a clear doublet which does not appear in the c-d pair. The doublet in the lanes g and h corresponding to those in e and f, (95 kDA) are wider apart suggesting differences in molecular weight which should explain differences in mobility. Further down to the 42 kDA region was another doublet in lanes e and f showing differences in 154 Appendix E. Electrophoretic Analysis of Rust Collections 155 mobility in comparison with the g-h pair in the same region. The 3-4 pair does not show the same doublet. Appendix E. ElectrophoTetic Analysis of Rust Collections 156 Figure E.15 are two profiles; lanes b and c which came from galls on lodgepole pine in Richmond Nature Park and d which was an isolate from scotch pine also from Richmond. While b and c look quite similar, the one from scotch pine shows 3 instead of two bands at about the 42 and the 50 kDA regions (arrows). The gross banding pattern also looks different. In another gel (Fig. E.16) however, an isolate (lane c) from lodgepole pine resembles the one from scotch pine (lane f) at the 42 kDA region. A second isolate from a different scotch pine tree (lane g) resembled the ones from lodgepole pine. Appendix E. Electrophoretic Analysis of Rust Collections 157 Figure E.13: SDS-PAGE profiles ( run on 10% gels) of single-gall spore sources of western gall rust collected on *lodgepole pine from two neighbouring stands near Prince George in British Columbia. The spores were used in an inoculation experiment which had 16 *lodgepole pine clones. The profile of the 4"* spore source was not included Lanes a and b, e and f do not show the strong doublet that lanes c and d show at about the 97 kDA region (arrows). Furthermore, lanes a and b have strong double bands at the 42 kDA. Below the 42 kDA region (both lanes e and f, arrows), there is a band which does not show on the others. Lane g is a crude form of Bovine serum albumin (BSA). Figure E.14: SDS-PAGE of single gall spores of western gall rust collected from *lodgepole pine trees growing in stands in Prince George, Lighthouse Park (coast) and Richmond (coast) in British Columbia Lanes a and b are Bio-Rad low and high molecular weight standards respectively, run on 8% gels. Lanes c-d, e-f and g-h are duplicate pairs from Prince George, Richmond and Lighthouse Park respectively. Differences occur at A (195 kDA) where the (c-d) pair are missing the strong band (arrow). At about the 97 kDA region (B), the (c-d) pair do not show the doublet which is clear in the e-f pair. At the same region (B), the (g,h) pair show two bands but which arc much wider apart (arrows). At about the 70 kDA region (C) ,the g-h pair do not show a band which is present in the e-f pair. At the 42 kDA region (D), the g-h pair have two separate bands while for the e-f pair, there is a strong doublet. Appendix E. Electrophoretic Analysis of Rust Collections 159 a b e d Figure E.15: SDS-PAGE profiles comparing single gall spores of western gall rust growing on lodgepole pine with 1 source collected from scotch pine Lane a shows Bio-Rad high molecular weight standards, b and c are gall rust isolates from lodgepole pine in Richmond and look quite similar; d is from scotch pine in Richmond. Differences occur at the 42 kDA and 50kDA regions which show triplets in the scotch pine isolate but show doublets on *lodgepole pine isolates (arrows). Resolution of bands above 66 kDA were not possible. Appendix E. Electrophoretic Ansdysis of Rust Collections 160 Figure E.16: SDS-PAGE profiles comparing single gall spores collections of western gall rust growing on lodgepole pine with 2 single spore sources collected from scotch pine Lane a shows Bio-Rad SDS-PAGE low molecular weight standards, lane b which is a lodgepole pine isolate resembles one isolate from scotch pine (f-g pair) at the 42 and 50 kDA regions. On the other hand, the other isolate from scotch pine (h) shows gross banding patterns similar to those from lodgepole pine (Arrows). Appendix E. Electrophoretic Analysis of Rust Collections 161 E.2 Discussion and Conclusions In general, there were not many differences. Many of these bands were quite fine and made it difficult to resolve differences. This can be a problem in electrophoresis when total proteins as opposed to fractions of proteins or certain specific enzyme systems are used as the bases of analysis. The rust showed some electrophoretic variation among isolates within the same stand but the differences were only located at a few regions. The differences described above were detected at three main regions. These were the 42, 95 and 97 kDa regions. The very fine and closely placed bands might have concealed genuine differences among isolates. All the same, a number of geographically separated spore isolates showed remarkable similarities. At the moment there is no direct connection between protein profiles and differences in the levels of infection caused by equal volumes of separate single gall inocula. The differences observed among the four spore sources used in the inoculation of clones in another experiment cannot be interpreted in terms of their electrophoretic profiles before further testing of the spore sources on the same or other clones. The results have simply shown that some variation can be detected even though the profiles differed by only a few bands. It is possible that two-dimensional SDS-PAGE might have revealed more differences but this remains to be seen. Based only on the method of analysis used here, the rust population appears to have very little variation but this remains to be seen as the use of other techniques may give different results. It may be useful to use other materials like germinating spores and if that does not work well, we can employ restriction fragment length polymorphisms (rflps) (Beckman and Soller 1986,). Even though the isolate from scotch pine had a unique protein profile, at least one isolate from lodgepole pine from the same locality of Richmond resembled it. If this had Appendix E. Electrophoretic Analysis of Rust Collections 162 not been observed, one would have suggested a host influence on the protein profile of rust isolates. The results suggest that the differences so far observed, are most probably dependent on the rust and not the host. The main conclusions from this study are as follows: • There were detectable but inconclusive differences among single spore sources within and between stands using one-dimensional SDS-PAGE. • The differences did not seem to involve many bands even though they occurred among single galls within and between stands and also between provenances. • The differences observed were most probably dependent on the rust and not the host, though the influence of the host could not be ruled out. • There were more similarities even among geographically separate spore isolates than there were differences. • The importance of the observed differences with respect to virulence has yet to be elucidated. o A better approach with an improved protein treatment procedure is required. 

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.831.1-0098173/manifest

Comment

Related Items