h. . r3. .3... 'r.;. ."‘-y4 VYfii‘na 0. ~ni~ :1. Icil .7Il. bl 0'9. - 0.5 . hy§vlt to T .: . v ...&I.tL A|ltLau u?.: all: .I O E . [It I. EQ.§'Q V I 2: Jnfi. E .(T 5 ‘ u {3 . o ' g. .l o 1 I?! I . v . 0 o O . .1» .. k...|\.~...rr.li. \m...LLu......A..?o , 2:. . .n .1. 1' 'l' ll,‘ ‘. GAN STATEU UENIVR llllgllllllllllll lllll llllllllllllllllllllll L 1293 01026 3402 LIBRARY Michigan State Unlverslty This is to certify that the thesis entitled GENETIC RELATIONSHIPS BETWEEN PLANT ARCHITECTURE, SEED SIZE AND ALLOZYMES IN COMMON BEAN (PHASEOLUS VULGARIS L.) presented by Mary Elizabeth Malburg has been accepted towards fulfillment of the requirements for MaSterS degree in PBG - CSS / Major professor Date 5’ / XI (/72 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE ll RETURN BOX to rornovo this chockout from your rocord. TO AVOID FINES rotum on or baton doto duo. DATE DUE DATE DUE DATE DUE MSU Io An Afflrmotlvo Action/Equal Opportunity Institution Wanna-m GENETIC RELATIONSHIPS BETWEEN PLANT ARCHITECTURE, SEED SIZE AND ALLOZYMES IN COMMON BEAN (PHZSEOLUS VULGARIS L.) BY Mary Elizabeth Malburg A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Plant Breeding and Genetics Program Department of Crop and Soil Sciences 1992 -32€7 f7 41—) w 6’7 ABSTRACT GENETIC RELATIONSHIPS BETWEEN PLANT ARCHITECTURE, SEED SIZE AND ALLOZYMES IN COMMON BEAN (PHASEOLUS VULGARIS L.) BY Mary Elizabeth Malburg Genetic relationships between upright plant architecture, seed size, and a diaphorase (DIA) allozyme variant were investigated in F} populations derived from crosses between navy bean (20 g/IOO seeds) and pinto bean (4O g/1OO seeds) Phaseolus vulgaris lines differing for type I, II and III growth habit and DIA genotype. Upright plant architecture, a highly heritable trait, segregated independent of seed size. The results provide evidence that, due to yield component compensation, the pinto ideotype differs from the navy architype for such pod traits as number of pods per plant or number of seeds per pod. The Diap-zms allele was present in 71% of pinto, navy, and great northern genotypes with type II plant architecture. The allele, however, was not linked to the architectural complex or the architectural traits investigated. Due to random drift, the Diap-st allele may be maintained in upright genotypes without being associated with a locus or linked loci governing upright plant architecture. ACKNOWLEDGMENTS I express my sincere gratitude to my major professor, Dr. Jim Kelly, for his guidance and suggestions concerning this research and for sharing his knowledge of common bean. I also thank my committee members, Dr. David Douches and Dr. Amy Iezzoni, for their input and critical review of this thesis. In addition, I thank Jerry Taylor and Norm Blakely for assisting in the planting and harvesting of my studies. Finally, I owe many thanks to my parents, Clarence and Mary Ellen, and also to Art, Madeline, Lucille, and Grandma Samoy for all of their prayers, understanding and support. And to Barry, thanks for your confidence in me and your continuous encouragement. iii LIST OF TABLES . LIST OF FIGURES TABLE OF GENERAL INTRODUCTION . . List of References CHAPTER.1: Interrelationship of plant architecture yield components in the pinto bean ideotype Abstract Introduction Materials and Methods . Results and Discussion Conclusion List of References . CHAPTER 2: Allozyme evaluation genotypes Abstract Introduction Materials and Methods . Results and Discussion Conclusion List of References SUMMARY . . List of References APPENDIX A . CONTENTS iv 13 14 16 22 39 42 45 46 48 54 69 71 75 78 79 APPENDIXB...... List of References CHAPTER 1 Table 1.1 Table 1.2 Table 1.3 Table 1.4 Table 1.5 CHAPTER 2 Table 2.1 Table 2.2 Table 2.3 LIST OF TABLES Characteristics of small-seeded (20 g/ 100 seeds) navy' and.:medium-seeded. (40 9/100 seeds) pinto bean lines used as parents. . Correlation coefficients (r) for architecture rating with three architectural traits in F2 and 1", generations (E. Lansing, MI 1990 and 1991). . . . . . . . . . . . . Broad-sense (H), narrow-sense (h?), and realized heritability (HR) estimates for architecture rating and branch angle from four dry bean crosses. . . . . . . . . . . Mean response of branch angle, hypocotyl diameter, and plant height to directional selection for high and low architecture ratings. F35 families grown at E. Lansing, MI 1991. . . . . . . . . . . . . . . . . . Mean response to directional selection for high and low seed weights. N84004 x P89405 F35 families grown at E. Lansing, MI 1991. Characteristics of small-seeded (20 9/100 seeds) navy‘ and.‘medium-seeded. (40 g/1OO seeds) pinto bean parental lines . . . . . Allelic segregatigpggit the DiTg-igo locus from crosses of géap-l ' / Dia -2 ' genotypes with Diap-l ' [Diap-zws'1o genotypes. E. Lansing, MI, 1990 . . . . . . . . . . . . Allelic segregation at the Diap-z locus of DJIap-195-95[Diary-21m"ws genotypes for Low (1, 2, and 3), Intermediate (4, 5, and 6), and High (7, 8, and 9) architectural categories of four F2 populations. E. Lansing, MI 1990. . vi 17 24 25 31 35 50 61 62 Table 2.4 APPENDIX A Table A.1 APPENDIX B Table 8.1 Table 8.2 Table 8.3 Table 8.4 Joint.F} segregation of Diap-Z alleles with plant growth habit . . . . . . . . . . . . Means and ranges for plant architecture rating and branch angle in the F2 and F generations (East Lansing, MI 1990 an 1991). . . . . . . . . . . . . . . . . . . Advanced navy bean lines and cultivars evaluated for DIA genotype . . . . . . . . Advanced. pinto Ibean lines and. cultivars evaluated for DIA genotype . . . . . . . . Advanced great northern bean lines and cultivars evaluated for DIA genotype . . . Mean architecture rating of Diap_2105,1os homozygous, Diap-2‘05-100 heterozygous, and Diap-Z °°-‘°° homozygous genotypic classes for eight l“2 populations. East Lansing, Michigan 1990. . . . . . . . . . . . . . . vii 65 79 80 82 85 87 CHAPTERl Figure 1.1 CHAPTER 2 Figure 2.1 LIST OF FIGURES Comparison of the navy architype, N84004, and the pinto ideotype, P89405. P894055 displays modified pod traits, allowing for the recovery of medium seed size. . . . . Figure 2.1. Diaphorase zymograms from leaf (L) and. root (R) tissue extracts. Migration distances (Rf) as calculated by Sprecher (1988b) are given on the left. Origin is located at the arrow. Diap-l100 homotetramer located at Rf=0.72; Diap-I"s homotetramer, Rf=0.65; Diap-Z105 homotetramer, Rf=0.10; Diap-2"m, Rf=-O.20. (S - Slow; S/U - Slow/Unique; U - Unique; U/F - Unique/Fast; F - Fast). . . . . . . viii 41 56 GENERAL INTRODUCTION Traditionally, breeding for enhanced yield potential emphasizes the selection.oijield.per se, or the incorporation of genes conferring resistance to diseases and/or insects (Mock and Pearce, 1975). Ideotype breeding differs from traditional breeding in that goals are described for specific morphological traits, resulting in a defined model plant (Rasmussen, 1987). The model plants, or 'ideotypes', are expected to perform in a predictable manner within a defined environment. Donald (1968) proposed that genetic advances for yield could be made when selecting for yield-enhancing morphological traits where the phenotypic goal for each trait is predetermined. While Donald's ideotype emphasized the selection of modified morphological traits, Blixt and Vbse (1984) noted that morphological phenotypes are the result of a biochemical or physiological message expressed in the phenotype. They contended that the genotype should be considered in defining an ideotype because allelic interactions within the genotype and interactions between alleles and the environment produce the resulting phenotype. Rasmusson (1987) expanded the ideotype definition even further to include physiological, biochemical, anatomical and phenological traits. He assumed 2 that such traits can be genetically manipulated and assembled into one genotype to enhance yield. Ideotypes, designed for yield enhancement, have been defined for many crops including'rice (Jennings, 1964), barley (Rasmusson, 1987), wheat (Donald, 1968), peas (Snoad, 1985), corn (Mock and Pearce, 1975) and dry beans (Adams, 1973) . Adams' (1973) original ideotype was later modified and described as an 'architype' (Adams, 1982) because of the emphasis placed on specific architectural characteristics of the plant. The proposed architype featured a moderate number of basal branches, upper internodes longer and more numerous than basal internodes, thick hypocotyl, tall stature, narrow profile, high values of first-order yield components (pod number, seed.weight, and seeds per pod), leaf area index near four at flowering, and an indeterminant type II growth habit. The three main dry bean plant habits grown in North America are types I, II, and III. Type I beans are determinate whereas, type II and type III beans are indeterminate. Beans with type III growth habit have relatively weak and semi-prostrate branches and concentrated pod load in the basal portion of the plant. Unlike type III growth habit, both the main stem and branches are strong and upright in type II beans (Singh, 1982). Kelly et a1. (1987) reported that type II upright growth habit provided greater yield stability in Michigan than either type I or type III growth habits. The erect growth habit of 3 type II beans was also shown.to decrease the severity of white mold (Sclerotinia sclerotiorum) through disease avoidance (Blad et al., 1978). The microclimate beneath the canopy of the type II plants was dryer, warmer, and less favorable for white mold development than the microclimate beneath the canopy of prostrate materials. The success of ideotype breeding in the small-seeded (20 g/1OO seeds) navy bean market class (Kelly et al., 1984) prompted the architectural improvement of the prostrate, type III, medium-seeded (40 9/100 seeds) pinto bean market class. The low frequency of genotypes possessing medium seed size in combination with type II growth habit, suggests a genetic barrier limiting recombination of the two traits. Within the 16,799 accessions catalogued at the CIAT (Centro Internacional de Agricultura Tropical; Cali, Colombia) germplasm bank, only 0.5% possess erect, type II architecture in combination with medium seed size (Kelly and Adams, 1987). In artificial crosses, genetic recombination between the architectural traits of the navy and black bean gene pools and the desired seed size, shape, and color of the pinto bean gene pool was also limited. Using a recurrent selection breeding scheme, Kelly and Adams (1987) combined erect plant architecture with medium seed size to develop a pinto bean ideotype. However, erect plants with medium seed size were not recovered until cycle three (C3) of recurrent selection. Previous selections based 4 on desired architecture had small seed size, whereas selections based on medium seed size were architecturally undesirable. The difficulty of combining erect architecture with medium seed size was attributed to a genetic barrier caused by a negative linkage between the two traits. Acquaah (1987) investigated plants originating from five distinct cycles of a phenotypic recurrent selection breeding scheme developed by Kelly and Adams (1987). Parental correlation analysis indicated.a highly significant, negative correlation between architecture rating and seed.weight. The negative association was weakened after one cycle of intermating and presumably broken in.C§ when medium-seeded, upright plants were recovered. Using factor analysis, he also confirmed that architectural traits and seed size traits represent two divergent gene pools. In the analysis, architectural traits had positive loadings, and seed size traits were negatively loaded, providing further evidence for a negative linkage. Acquaah (1987) concluded that phenotypic recurrent selection was effective in recombining plant architecture and medium seed size in the pinto bean ideotype. In the recurrent selection breeding scheme, medium seed size was gradually recovered over cycles (Acquaah, 1987) . Erect plant architecture, on the other hand, was established in early cycles and stabilized in C3. The complex of traits governing plant architecture was recovered almost as a unit. Furthermore, plant architecture, although. a quantitative 5 trait, can be scored qualitatively for selection and that score is stable across environments. Acquaah et al. (1991) used stepwise multiple regression to determine traits significantly contributing to plant architecture. Hypocotyl diameter, plant height, branch angle, and pods on the main stem (especially in the mid section) most frequently contributed to architecture rating. The traits were moderately heritable (Acquaah et al., 1989), therefore, erect type II growth habit can be achieved when selection is based on the architectural traits. Based on growth habit, morphological characteristics, and regions of adaptation, Singh (1989) identified six gene pools within the Middle American center of domestication (Mesoamerican) and four gene pools within the South American center of domestication (Andean). The Andean snap or stringless bean has two additional groups of variation. Singh et al. (1991a) further defined six races composed of one or more of the previously determined gene pools. Multivariate statistical analysis based on morphological, agronomic, and molecular data, determined three races within each of two major gene pools: Middle American (Mesoamerican) and Andean South American. The Middle American gene pool consists of races Mesoamerica, Durango, and Jalisco. Races Chile, Nueva Granada, and Peru compose the Andean South American center of domestication (Singh et al., 1991a). 6 One subgroup of race Mesoamerica within the Mesoamerican gene pool is characterized by indeterminate, erect, type II landraces (Singh et al. , 1991a) identified previously by Adams (1973) as sources for improved plant architecture. The same landraces also cluster on the basis of a common diaphorase (DIA) allele, Diap-Z‘os. (Locus and allele designation is as described by Koenig and Gepts, 1989.) This group corresponds to cluster E of Singh et al. (1991b). Sprecher (1988b) surveyed advanced, architecturally erect bean lines for DIA and found a high frequency of the 'Unique' banding pattern or zymotype among ideotypes. In P. vulgaris, Weeden (1984) reports three distinct zymotypes for DIA, a tetrameric enzyme coded by two tightly linked loci (Sprecher, 1988a) . Sprecher (1988a) determined the presence of seven alleles, including two nulls, at the two DIA loci, Diap-l and Diap-Z. Diap-l displays four alleles and Diap-Z, three alleles. The five banded pattern visualized in genotypes homozygous at both loci results from the formation of two intra-locus homotetramers and three inter-locus heterotetramers (Harris and Hopkinson, 1976; Pasteur et al., 1988). The Unique zymotype associated with type II architecture is produced by the DIA genotype Diap-l’s/Diap-z‘os. Diaphorase is a widely distributed flavoprotein (Dixon and Webb, 1979) involved in the pyruvate dehydrogenase complex and the 2-oxoglutarate dehydrogenase complex (Webb, 1984) . The enzyme is reported as Enzyme Commission (EC) 1.6.4.3 7 (Brown et al., 1978; Kiang and German, 1983; Taggart et al., 1990), EC 1.6.99.3 (Brown, 1983; Pedersen and Simonsen, 1987),and EC 1.8.1.4 (Westphal and Wricke, 1989). Enzyme Commission 1.6.4.3 was later reclassified as EC 1.8.1.4 (Webb, 1984) . All enzymes catalyze oxidoreduction reactions but differ as to their respective acceptor compounds (Webb, 1984) . Diaphorase may have evolutionary significance because of the close association between the D.iap--2105 allele and type II growth habit in beans. The Diap-z105 allele is found only in cultivated erect plants with type II growth habit. Singh et al. (1991b) assayed a total of 227 landraces representing a geographical distribution from Mexico to Argentina and Chile. Variability for nine polymorphic loci provided information on gene flow, from wild to cultivated P. vulgaris. In wild P. vulgaris materials, four alleles, Diap-blmz, Diap-l'm, Diap-l", and Diap-l’s' are present at the Diap-l locus, whereas cultivated materials display only two alleles, Diap-l100 and Diap-l’s. At the Diap-z locus however, wild materials possess fewer alleles than do the cultivars. Alleles Diap-Z100 and Diap—Z105 are present at the Diap-z locus in the cultivated P. vulgaris. Wild materials possess the Diap-z100 allele but lack the Diap-Z105 allele (Singh et al., 1991b) . Allozyme analysis of 83 wild P. vulgaris accessions representing a region from Mesoamerica to Argentina also lack the Diap-Z‘os allele (Koenig 8 and Gepts, 1989). The Diap-Zms allele was not present in the sampled wild materials and may be the result of a mutation during or shortly after the domestication of type II genotypes (Singh et al., 1991b). Plant domestication is an evolutionary process whereby wild plants are changed genetically by human selection, usually rendering them less fit for survival in the wild (Pickersgill, 1986). The genetic variability of the domesticated populations is due, in part, to the number of individuals involved in the formation of the domesticated population (Ladizinsky, 1985). Populations domesticated from a small sample of wild materials represent a limited fraction of the variability present in the wild population (Pickersgill, 1986). This phenomenon of reduced variability is also known as 'founder effect'. Founder effect refers to the establishment of a new population by a small number of individuals representing only a portion of the total variability (Mayr, 1942). The genetic composition of a population is limited. to» the alleles introduced by' the founders (Schwaegerle and Schaal, 1979). The significance of founder effect on reduced variability depends not only upon the number of founding individuals but also upon: 1.) the rate of gene exchange between the wild and the cultivated populations and 2.) the appearance of mutations during or after the founding event (Ladizinsky, 1985). Restricted gene exchange either by geographical isolation or 9 reproductive isolation is a crucial requirement of speciation (Mayr, 1942) . In common bean, self-pollination decreases the chance of hybridization between cultivated materials and their wild relatives and allows the founder plants to establish separate populations (Ladizinsky, 1985) . Although founder effect typically results in reduced variability, the appearance of mutations during speciation may produce novel variability in the crop plant. Introduction of crop plants to new environments allows for the selection of the mutations and the establishment of geographic races (Ladizinsky, 1985). In the first chapter of this thesis, the genetic relationship between erect plant architecture and seed size is investigated. In addition, the heritability of the architectural complex is examined. The objective of the second chapter is to determine the genetic relationship between erect plant architecture and the Diary-2‘05 allele. 10 LIST OF REFERENCES Acquaah, G. 1987. Genetic and multivariate statistical evaluation of a phenotypic recurrent selection program for recombining erect architecture and large seed size in Phaseolus vulgaris L. Ph.D. Dissertation, Michigan State University, East Lansing, MI. 317 pages. Acquaah, G., M.W. Adams, and J.D. Kelly. 1989. Broad-sense heritability estimates of several architectural traits in dry beans (Phaseolus vulgaris L.). Annu Rept Bean Improv Coop 32:75-76. Acquaah, G., M.W. Adams, and J.D. Kelly. 1991. Identification of effective indicators of erect plant architecture in dry bean (Phaseolus vulgaris L.). Crop Sci 31:261-264. Adams, M.W. 1973. Plant architecture and physiological efficiency in the field bean. In: Potentials of field beans and other food legumes in Latin America. CIAT, Cali, Columbia. pp 266-278. Adams, M.W. 1982. Plant architecture and yield breeding. Iowa State J Res 56:225-254. Blad, 8.L., J.R. Steadman, and A. Weiss. 1978. Canopy structure and irrigation influence white mold disease and microclimate of dry edible beans. Phytopathology 68:1431-1437. Blixt, S. and P.B. Vose. 1984. Breeding towards an ideotype -- aiming at a moving target? In: Crop breeding, a contemporary basis. Pergamon Press, Oxford. pp 414-426. Brown, A.H.D. 1983. Barley. In: Tanksley, S.D. and Orton, T.J. (eds) Isozymes in plant genetics and breeding, Part B. Elsevier, Amsterdam. pp 55-78. Brown, A.H.D., E. Nevo, D. Zohary, and O. Dagan. 1978. Genetic variation in natural populations of wild barley (Hordeum spontaneum). Genetica 49:97-108. Dixon, M. and E.C. Webb. 1979. Enzymes, third edition. Academic Press, New York. Donald, C.M. 1968. The breeding of crop ideotypes. Euphytica 17:385-403. 11 Harris, H. and D.A. Hopkinson. 1976. Handbook of human electrophoresis. North-Holland Publishing Co., Amsterdam. Jennings, P.R. 1964. Plant type as a rice breeding objective. Crop Sci 4:13-15. Kelly, J.D. and M.W. Adams. 1987. Phenotypic recurrent selection in ideotype breeding of pinto beans. Euphytica 36:69-80. Kelly, J.D., M.W. Adams, and G.V. Varner. 1987. Yield stability of determinant and indeterminate dry bean cultivars. Theor Appl Genet 74:516-521. Kelly, J.D., M.W. Adams, A.W. Saettler, G.L. Hosfield and A. Ghaderi. 1984. Registration of C-20 navy bean. Crop Sci 24:822. Kiang, Y.T. and M.B. German. 1983. Soybeans. In: Tanksley, S.D. and Orton, T.J. (eds) Isozymes in plant genetics and breeding, Part B. Elsevier, Amsterdam. pp 295-328. Koenig, R. and P. Gepts. 1989. Allozyme diversity in wild Phaseelus vulgaris: further evidence for two major centers of genetic diversity. Theor Appl Genet 78:809- 817. Ladizinsky, G. 1985. Founder effect in crop-plant evolution. Econ Bot 39:191-199. Mayr, E. 1942. Systematics and the origin of species. Columbia Univ Press, New York. Mock, J.J and R.B. Pearce. 1975. An ideotype of maize. Euphytica 24:613-623. Pasteur, N., G. Pasteur, F. Bonhomme, J. Catalan, and J. ' Britten-Davidian. 1988. Practical isozyme genetics. Ellis Horwood Limited Publishers, Chichester. Pedersen, S. and V. Simonsen. 1987. Tissue specific and developmental expression of isozymes in barley (Herdeum vulgare L.). Hereditas 106:59-66. Pickersgill, B. 1986. Domestication and its taxonomic consequences. Acta Hort 182:319-327. Rasmusson, D.C. 1987. An evaluation of ideotype breeding. Crop Sci 27:1140-1146. 12 Schwaegerle, K.E. and 8.A. Schaal. 1979. Genetic variability and founder effect in the pitcher plant Sarracenia purpurea L. Evol 33:1210-1218. Singh, S.P. 1982. A key for identification of different growth habits of Phaseolus vulgaris L. Annu Rept Bean Improv Coop 25:92-95. Singh, S.P. 1989. Patterns of variation in cultivated common bean (Phaseolus vulgaris, Fabaceae). Econ Bot 43:39-57. Singh, S.P., P. Gepts, and D. Debouck. 1991a. Races of common bean (Phaseolus vulgaris, Fabaceae). Econ Bot 45:379-396.. Singh, S.P., R. Nodari, and P. Gepts. 1991b. Genetic diversity in cultivated common bean: I. Allozymes. Crop Sci 31:19-23. Snead, B. 1985. The need for improved pea-crop ideotypes. In: Hebblethwaite, P.D., Heath, M.C., and Dawkins, T.C.K. (eds) The pea crop: a basis for improvement. Butterwerths Press, London. Sprecher, S.L. 1988a. New isozyme variants at two NADH diaphorase loci in dry beans: correlations to gene pools and commercial classes. Annu Rept Bean Improv Coop 31:92. Sprecher, S.L. 1988b. Allozyme differentiation between gene pools in common bean (Phaseolus vulgaris L.), with special reference to Malawian germplasm. Ph.D. Dissertation, Michigan State University, East Lansing, MI. 207 pages. Taggart, J.B., S.P. McNally, and P.M. Sharp. 1990. Genetic variability and differentiation among founder populations of the pitcher plant (Sarracenia purpurea L.) in Ireland. Hered 64:177-183. Webb, E.C. 1984. Enzyme nomenclature. Academic Press, Inc., Orlando. Weeden, N.F. 1984. Distinguishing among white seeded bean cultivars by means of allozyme genotypes. Euphytica 33:199-208. Westphal, L. and G. Wricke. 1989. Genetic analysis of DIA, GOT and PGI isozyme loci in Daucus carota L. ssp. sativa. Plant breed 102:51-57. CHAPTER ONE INTERRELATIONSHIP OF PLANT ARCHITECTURE AND YIELD COMPONENTS IN THE PINTO BEAN IDEOTYPE ABSTRACT Linkage relationships between architectural traits, plant architecture, and seed size were analyzed in F} populations derived from crosses between type I, II, and III Phaseelus vulgaris lines from small-seeded (20 g/100 seeds) and ‘medium-seeded. (40 g/1OO seeds) common. bean lines. The architectural trait branch angle was the most significant contributor to plant architecture. Upright plant architecture was estimated to be moderately high to highly heritable (h2== 0.60-0.85). Seed size and upright.plant.architecture‘were not linked. However, a negative linkage would appear to exist if selection for a medium-seeded pinto ideotype were based visually on the pod traits of the navy ideotype. Our results provide evidence that, due to yield component compensation, the pinto ideotype differs from the navy architype for such pod traits as number of pods per plant or number of seeds per pod. 13 14 INTRODUCTION Ideotype breeding in dry beans (Phaseolus vulgaris) has been used successfully to improve plant architecture and yield stability (Kelly et al., 1984; Kelly et al., 1987). The dry bean ideotype, or 'architype' (Adams, 1982) , emphasizes a moderate number of basal branches, thick hypocotyl, tall stature, narrow profile, and indeterminant type II growth habit (Adams, 1973). Singh (1982) defined three dry bean growth habits, type I, determinant upright; type II, indeterminate upright; and type III, indeterminate prostrate. Type II upright growth habit provides greater yield stability (Kelly et al., 1987) and can reduce the incidence of disease (Blad et al., 1978). Acquaah et al. (1991) identified hypocotyl diameter, plant height, branch angle, and pods on the main stem, especially in the midsection, as substantial contributors to upright plant architecture. Architectural improvement based on the architype concept was successful in the small-seeded (20 g/100 seeds) navy bean market class (Kelly et al., 1984) which motivated the architectural improvement of the medium-seeded (40 g/ 100 seeds) pinto bean market class. Pinto beans differ from the navy architype in having prostrate, indeterminant type III growth habit while producing 70% of the pods at the basal nodes (Kelly and Adams, 1987). The majority of pods are in contact with the soil surface. Architectural improvement was limited in the pinto class by negative linkages between seed 15 size and upright plant architecture (Kelly and Adams, 1987; Acquaah, 1987). Kelly and Adams (1987) used a phenotypic recurrent selection breeding scheme to combine the architectural characteristics of the small-seeded navy architype with the seed size, shape and color of the prostrate type III pinto bean. The complex of traits interacting to produce the architype were recovered almost as a unit. However, the frequency of recovery was low in the initial cycles of recurrent selection. In the initial cycles, the architectural selections had small seed size (20-25 g/100 seeds) whereas selections based on large seed size (>30 g/100 seeds) were architecturally undesirable. Characteristics such as seed color and.mottling recombined freely in the first segregating generations, but architecture and seed size did not show evidence of recombination until cycle 3 (C3). When the seed size traits of the pinto bean class were combined with the architectural traits of the navy architype, the resulting pinto ideotype had larger but fewer pods than the navy architype (Acquaah et al., 1991). This investigation was conducted to (1) determine the genetic relationship between the complex traits of seed size and upright plant architecture and (2) to define a pinto bean ideotype for the humid midwestern United States. 16 MATERIALS AND METHODS 1‘ .- e ' - .tio s ’. be wee; s-eo s' - an- no i- .‘1 architm Plant materials A partial diallel was created with six dry bean lines representing small (20 g/100 seeds) and medium (40 g/100 seeds) seed-size classes and types I, II, and III growth habits (Table 1.1) . The small-seeded navy bean parental lines were Seafarer, N84004 and Michelite, whereas Ouray, P89405, and UI-114 were the medium-seeded pinto bean parents. N84004 and P89405 are MSU breeding lines developed as sources of upright, type II plant architecture in navy and pinto bean germplasm. Crosses were made between seed sizes but within growth habits (three crosses), between growth habits but within seed sizes (six crosses), and between both seed sizes and growth habits (six crosses) for a total of 15 crosses. Approximately 320 seeds per F2 population were planted at a 20 cm spacing within rows and a 51 cm spacing between rows in East Lansing, Michigan, 1990. Where emergence was poor, a type I kidney bean, Isabella, was hand planted two weeks later to ensure uniform interplant competition. Parental lines were included as checks on each side of the F};populations. lfi plants from the six crosses segregating for both seed size and.plant.type were backcrossed to each.parent to provide narrow-sense heritability estimates. The six crosses 17 Table 1.1. Characteristics of small-seeded (20 g/100 seeds) navy and medium-seeded (40 9/100 seeds) pinto bean lines used as parents. Bean Commercial Seed Growth Architecture Line Class Size Habit Rating Ul-l 14 Pinto Medium Ill 1 (prostrate) Michelite Navy Small Ill 2 Seafarer Navy Small I 4 Ouray Pinto Medium I P89405 Pinto Medium ll N84004 Navy Small ll 18 included: Seafarer x P89405, Seafarer x UI-114, N84004 x Ouray, N84004 x 01-114, Michelite x Ouray, and Michelite x P89405. Agronomic evaluation At maturity, approximately 210 plants per F2 population, 30 plants per parental genotype and 35 F1 plants per backcross were individually rated for plant architecture on a scale from 1 to 9. A rating of 1 depicts prostrate growth habit, whereas a rating of 9 represents the upright navy architype. Ratings 2 through 8 were scored according to the degree of deviation from the two extremes. In order to quantify the architectural ratings, all plants were evaluated for agronomic traits that contribute to erect plant architecture (Acquaah et al., 1991). The architectural traits included branch angle, hypocotyl diameter and plant height. Branch angle was measured as the inclination of the branch to the central axis, hypocotyl diameter was measured just above the soil surface and plant height was measured as the length of the main stem excluding the vine. Seed weight was recorded on the weight of 50 seeds. l9 Heritability estimates Broad-sense heritabilities for plant architecture were calculated using the following formula: 2 2 Z 2 H=a,2 - 1/3 (aP1 + up, + 0F.) [1] 2 “$2 However, the geometric mean of the parents and the F3 more closely estimated environmental variance and equation [2] was used to estimate broad-sense heritability for the architectural traits. H=o’.z - [(02.1) (oz...) (02.1) 1‘” [21 Z arz For both [1] and [2] 02,2, 02,1, 02,2, and 02,, are variances of F2, parent 1, parent 2 and F1 generations, respectively (Fehr, 1987). Narrow-sense heritabilities were estimated by parent- offspring regression of F2:3 progeny means on an F} parent. The linear regression model proposed by Lush (1940) is: Y,=a+in+ei [3] where Y}:= performance of offspring of the ith parent a = mean performance of all parents evaluated b = linear regression coefficient )9 = performance of ith parent ei = experimental error 20 Narrow-sense heritability was also estimated according to the Warner backcross method (Warner, 1952) . The equation used to estimate heritability is: h2 = 2(02r2) ' (0231 + 0232) [4] 2 “:2 where 02,2 is the variance among F2 plants and 0231 and 0232 are .the variances among plants from the backcrosses of the F1 to parent 1 and parent 2, respectively. Realized heritability, or the ratio of response to selection differential, was estimated as the difference in mean performance of high and low F3 progeny divided by the difference in the mean of the F2 parents (Fehr, 1987): 2 .. - h " xhioh, r3 ' xlou, r3 [5] finish, r2 ' ilow, r2 The predicted genetic advance (6,) due to selection for plant architecture is estimated by the population variance and the heritability of architecture (Lush, 1945). c, = iopun [6] where i = selection intensity in standard deviation units a," = phenotypic variance H = heritability of the quantitative trait ec ' n se ect'o or i ve sus owa ch'te ture rat Four pepulat ions derived from the crosses Seafarer XP89405, N84004 x Ouray, N84004 x UI-114, and 21 Michelite x P89405 were selected for inclusion in a directional selection experiment. The cross N84004 x UI-114 is typical of crosses initially'made by Kelly and Adams (1987) to combine the plant architecture of the navy architype with the seed size of the prostrate, type III pinto. The Michelite x P89405 cross is the reciprocal, a prostrate, type III navy crossed to the pinto ideotype. The remaining two populations, Seafarer x P89405 and N84004 x Ouray, are type I beans crossed to an ideotype. For each population, ten F} plants, or approximately 5% of the plants analyzed for architectural traits, were selected for both high and low architecture ratings for a total of 20 F35 families. For the N84004 x UI-114 and Michelite x P89405 families, high architecture ratings were 6 or greater and low architecture ratings were 3 or less. The Seafarer x P89405 and N84004 x Ouray populations had high architecture ratings of 7 or greater and low architecture ratings of 5 or less. The only restriction for inclusion in the study was that the F} plants produced 60 or more seeds. Twenty seeds derived from a single F} plant were spaced planted (20 cm spacing) in a row and replicated three times in a completely randomized block design at East Lansing, Michigan, 1991. Data were collected on architecture rating and seed weight. Five plants with uniform interplant competition, were randomly selected from each entry and evaluated for plant height, branch angle, and hypocotyl diameter. 22 - .-. .91. —- r 91 o riot v-_~ -w s--- .-’-i The population derived from the cross N84004 x P89405 was studied to evaluate the effect of selecting for large or small seed size. .Adams (1967) defined yield.as the product of three components; X, the average number of pods; Y, the average number of seeds per pod; and z, the average weight per seed. Five percent of the total number of F} plants analyzed for architectural traits, were randomly selected for both large (>15 g/50 seeds) and small (<12 g/50 seeds) seed size for a total of 20 F55 families. The only restriction for inclusion in this investigation was that the F3 plants produced 60 or more seeds. Twenty seeds derived from an.F}jplant were planted at a 20 cm spacing within the row and replicated three times in a randomized complete block design at East Lansing, Michigan, 1991. Each family row was rated for plant architecture. Five plants with uniform interplant competition were randomly selected from.each entry and evaluated for the total number of pods, the total number of seeds, and the weight of 100 seeds. RESULTS AND DISCUSSION ab ' i la c 't tu Twelve F2 populations derived from a partial diallel were rated for plant architecture on a scale of 1 to 9. In addition, architectural traits (branch angle, hypocotyl diameter, and plant height) were measured to substantiate the 23 architecture rating. In the F2 generation, branch angle most frequently correlated to architecture rating. Three exceptions were crosses between type I and type III plants (Table 1.2). The crosses did not involve the architectural plant types (N84004 and P89405) therefore, the exceptions support the importance of branch angle, as a trait generally associated with narrow plant profile, and a key component of type II growth habit. The remaining architectural traits, hypocotyl diameter and plant height, were less important contributors to architecture rating in these particular populations. Similar trends were observed in the plant architecture directional selection studies but, branch angle constituted an even greater portion of the architecture rating. The higher association is attributed to the lack of intermediate architectural ratings which may confound branch angle with other architectural traits. Bread-sense heritability estimates for plant architecture were exceptionally high indicating the environment has little effect on the expression of plant architecture (Table 1.3). A large portion of the genetic variation is heritable and can be selected as evident by the narrow-sense estimates. Using parent-offspring regression of F2:3 progeny means on an F2 parent, architecture rating was estimated to be moderately high to highly heritable. The realized heritability estimates indicate that greater architectural differences occurred in 24 Table 1.2. Correlation coefficients (r) for architecture rating with three architectural traits in F2 and 1“, generations (E. Lansing, MI 1990 and 1991). Trait . Branch Hypocotyl Plant Population Crosst Gen. Anglet Diameter Height Seafarer x Michelite l,N x Ill,N F, .06 -.24' ' ' -.53' ' ' Seafarer x Ul-114 I,N x "L? F, -.20" .02 .15“ ‘ Michelite x Ouray lll,N x I,P F, -.12 -.12 -.31* r I Ouray x Ul-114 I,P x lll,P F, .13 .03 .04 Seafarer x N84004 I,N x ll,N F, -.44' * ' .22 ' ’ * .44' * * Ouray x P89405 I,P x ll,P F, -.39* * ' -.17' .03 N84004 x Michelite ll,N x lll,N F, -.45*" -.06 . .03 P89405 x Ul-114 Il,P x lll,P F, -.35”* -.33" -.24”* Seafarer x P89405 I,N x II,P F, -.34“* .20" 34“” F, -.79' r r -.01 .01 N84004 x Ouray ll,N x I,P F, -.29*" .03 .04 F, -.76'" -.14 -.17 N84004 x UI-114 ll,N x III,P F, -.18” .10 .03 F, -.71"* -.14 .18 Michelite x P89405 III,N x ”P F, -.25' ” -.03 -.05 F, -.88* " ' -.26* -.32' '.".*” Significant at the .05, .01 and .001 levels, respectively. 1 l-determinate bush, ll-indeterminate upright, Ill-indeterminate prostrate. P and N designate pinto and navy commercial classes, respectively. 1 Actual means and ranges for plant architecture rating and branch angle are reported in Table A.1. 25 Table 1.3. Broad-sense (H), narrow-sense (hz), and realized heritability (HR) estimates for architecture rating and branch angle from four dry bean crosses. Branch Angle Architecture Rating Seafarer x P89405 Cross H h2 H, _l N84004 x Ouray N84004 x Ul-114 Michelite x P89405 Seafarer x P89405 N84004 x Ouray N84004 x Ul-l 14 Michelite x P89405 26 the F3 generation than in the F3 generation. In a study by Acquaah (1987) plant. architecture :remained. stable lacross environments as heritability estimates were obtained from generations grown in two different seasons. Plant architecture was also relatively environmentally stable when grown in East Lansing, Michigan and Chimaltenango, Guatemala. The cross N84004 x UI-114 is typical of crosses made to combine plant architecture with medium seed size, therefore, gain from selection was estimated for the population N84004 x 01-114. using a 10% selection intensity, average architecture rating of the progenies is predicted to be 31.3% greater than the mean of the unselected F, generation. Selection for plant architecture can be effectively conducted in early generations as indicated by the heritability estimates and the Gs. Erect plant architecture was recovered in the original cycle of recurrent selection and enhanced in subsequent cycles (Kelly and Adams, 1987). Narrow-sense heritabilities calculated from the Warner backcross method were inconsistent. Due to time constraints, BCflfi plants from the backcrosses were evaluated rather than 8C11"2 plants. BC1F2 plants would have provided more accurate and reliable estimates of the nonheritable components of variance, and therefore, more valid narrow-sense heritability estimates. For this reason, Warner backcross heritability estimates will not be presented. 27 Plant architecture appears to have a relatively simple mode of inheritance although it is classified as a quantitative trait because of its complex interactive morphological features. The fact that progress for upright plant architecture was made in the F2 generation implies that major genes govern the trait. The minimum number of genes controlling plant architecture was estimated according to the Castle-Wright equation (Wright, 1934) which only detects genes for which the parents differ. The estimated number of genes governing plant architecture was 2.2 for the Seafarer x P89405 population, 3.9 for the N84004 x Ouray population, 4.7 for the Michelite x P89405 population and 6.3 for the N84004 x UI-114 population” The Seafarer x P89405 and N84004 x Ouray populations were derived from crosses between type I and type II growth habits which have more similar architectural characteristics than type II and type III growth habits (Singh, 1982). The estimated number of genes governing plant architecture provides evidence that type I and type II growth habits are also genetically more similar. Frazier et al. (1958) investigated upright and sprawling, prostrate bush growth habits of snap bean. They found that three major recessive genes may be responsible for the upright bush habit. In other studies, a few major genes have likewise accounted for a substantial portion of the variation of quantitative traits. Sullivan and Bliss (1983) suggested that one or a few genes were responsible for enhanced levels of 28 phaseolin in near-homozygous inbred-backcross lines of common bean. .Also in common bean, Vallejos and Chase (1991) identified a gene, Ssz-l, that exhibited additive gene action and accounted for 30-50% of the seed size differences between parents which had a two-fold difference in seed weight. In oats, as few as two or three major genes may have a large effect on oat mosaic virus resistance, a quantitative trait with relatively simple inheritance (Uhr and Murphy, 1992). However, large populations are necessary to detect a significant departure from the normal distribution characteristic of traits influenced by many polygenes (Theday and Thompson, 1976). The architectural complex, in addition to being controlled by few genes, has a moderately high heritability estimate, and can be easily and effectively selected in early generations. These attributes characterize plant architecture as a suitable trait for a backcross breeding scheme. IHewever, in. a backcross breeding scheme, the intensity of plant architecture may be effected as desirable alleles for plant architecture are lost in each backcross. As a result, large populations would have to be developed in order to ensure that favorable alleles are maintained in the population and the subsequent recovery of architecturally desirable plants. Phenotypic recurrent selection allowed for the recovery of the major architectural genes in the initial cycle whereas minor genes were accumulated in later cycles. In addition, 29 medium seed size was recovered without affecting the intensity of plant architecture. Seed size is an additive trait that may be governed by as many as ten effective factors in bread crosses between wild and cultivated P. vulgaris (Motto et al. , 1978) . The cyclic breeding scheme was necessary to gradually accumulate all of the seed size genes as well as minor architectural genes. The major contributor to plant architecture in both the F, and the F3 generations was branch angle and therefore, one would expect high to moderately high branch angle heritabilities. Broad-sense heritabilities were of similar magnitude as those estimated by Acquaah et al. (1989) and provide evidence of genetic variation for branch angle (Table 1.3). Branch angle heritabilities calculated from parent- effspring regression were low relative to the estimated plant architecture heritabilities (Table 1.3) . Branch angle may have lower heritability estimates because there is less initial variability. The narrow-sense heritability estimate in the N84004 x Ouray population was considerably lower than the estimates for the remaining three populations. The low estimate (h2 = 0.07) may indicate that the N84004 x Ouray population is more responsive to environmental differences. It is possible, however, to differentiate between high and low architecture ratings as indicated by the realized heritability estimate. The narrow-sense heritability estimates suggest 30 that a component of the genetic variance for branch angle is due to additive genetic variance. The narrow-sense heritability estimates for branch angle may be biased because branch angle was not directly selected in the F, generation nor was it randomly sampled in the F, generation. Furthermore, environmental differences can effect the magnitude of the regression coefficients when parents and offspring are not.grown.in the same season (Bartley and Weber, 1952). Variable environmental conditions decrease heritability estimates, whereas, uniform conditions increase heritability estimates (Falconer, 1960). For all crosses realized heritability estimates for branch angle exceeded parent-offspring heritability estimates and were comparable or somewhat higher than the broad-sense estimates (Table 1.3). For the N84004 x UI-114 population, the realized heritability estimate exceeded the theoretical limit indicating that greater branch angle differences occurred in the F3 generation than in the F, generation. Although realized heritability estimates may be biased by environmental effects, inbreeding, maternal effects (Falconer, 1960), or limited sampling, the estimates do signify that narrow branch angle selected in the F, generation is expressed in the F3 generation. FR, families selected for desirable plant architecture had significantly narrower profiles than did those selected for undesirable plant architecture (Table 1.4) . This supports 31 Table 1.4. Mean response of branch angle, hypocotyl diameter, and plant height to directional selection for high and low architecture ratings. F,:3 families grown at E. Lansing, MI _ NS = Not significant at the 0.05 level. 1991 . r—-_ Architecture Rafing Trait F,:, Family High Low Branch Angle (degrees) Seafarer x P89405 30.4 33.0 .001 N84004 x Ouray 27.0 33.0 .004 N84004 x Ul-1 14 32.6 38.4 .001 Michelite x P89405 29.8 35.5 .03 Trait Average 30.0 35.0 .001 Hypocotyl Diameter (mm) Seafarer x P89405 7.7 7.6 NS N84004 x Ouray 7.2 7.4 NS N84004 x Ul-114 6.8 7.2 NS Michelite x P89405 7.9 7.9 NS Trait Average 7.4 7.5 NS Plant Height (cm) Seafarer x P89405 39.5 38.3 NS N84004 x Ouray 34.0 35.3 NS N84004 x Ul-114 34.0 36.2 NS Michelite x P89405 40.4 42.9 NS Trait Average 32 the realized heritability estimates and indicates that selection for branch angle can be effectively conducted. In fact, Acquaah (1987) showed that desired branch angle was recovered after C1 of recurrent selection and was improved in subsequent cycles. It seems evident, however, that progress for erect plant architecture can be made when selection is based on narrow plant profile of which branch angle is the major contributing factor. F,5 family means for hypocotyl diameter and plant height did not vary due to directional selection based exclusively on architectural rating (Table 1.4.). Both traits increased in magnitude from the original cycles to the advanced cycles of recurrent selection (Acquaah, 1987). Even though significant differences were not observed in response to directional selection, the traits may be emphasized and modified in subsequent generations. Selection for either trait would influence the other because they are highly correlated (correlation coefficients range from r=0.30 to r=0.71 in the F,:3 families) . Relationship between upgight giant erchiteetuze end seed sleight Six F, populations; Seafarer x P89405, Seafarer x UI-114, N84004 x Ouray, N84004 x UI-114, Michelite x Ouray, and Michelite x P89405; were evaluated for plant architecture and seed weight. In five of the six populations, architecture 33 rating and seed weight were nonsignificantly correlated. The exception was the population Seafarer x P89405. The significant but low positive correlation between seed weight and architecture (r=0.157, P<.02) indicates that upright erect plant architecture associates with increased seed weight. This association is not unexpected because the donor of medium seed size is also the improved architectural type. Overall, the F, data provide evidence that architecture and seed weight are not negatively linked. In the architecture directional selection studies, architecture rating and seed weight were nonsignificantly correlated in the N84004 x 01-114 F,:3 families and in the N84004 x Ouray F,:3 families. The Michelite x P89405 F,:3 families had a significant but low negative correlation (r=-0.26, P<.05) between architecture rating and seed weight. However, the mean seed weight of F,:3 families originally selected for high architecture rating (27.9 g/100 seeds) was not significantly lower than the mean seed weight of F,:3 families selected for low architecture rating (29.9 g/100 seeds). Seed weight and architecture were significantly correlated (r=0.41, P<.001) in the Seafarer x P89405 F,:3 families. A significant positive correlation also existed in the F, generation . 34 te 'o s ' ' com o ents A directional selection study was conducted with the population N84004 x P89405 to investigate why medium-seeded, upright beans were not recovered in the initial cycles of Kelly and Adams' (1987) recurrent selection. N84004 and P89405 possess type II growth habit so progeny segregated only for yield components X, average number of pods; Y, average number of seeds per pod; and 2, average weight per seed (Adams, 1967). F35 families selected for high seed weight had a significantly higher mean seed weight than families selected for low seed weight (Table 1.5). Regression of the F,:3 progeny means on an F, parent provided narrow-sense heritability estimates for seed weight. The estimates were r=0.70 (P<.001) for Seafarer x P89405, r=0.70 (P<.001) for N84004 x Ouray, r=0.69 (P<.001) for N84004 x UI-114, and r=0.54 (P<.01) for Michelite x P89405. The high to moderately high seed weight heritabilities agree with other heritability estimates (Nienhuis and Singh, 1988; Motto et al., 1978). Directional selection for high versus low seed weight did not result in significant differences in seed yield (Table 1.5) . When compared to the navy ideotype of equal yield potential, the pinto ideotype essentially represents a two-fold increase in the 2 component for seed size. Therefore, one or both of the remaining yield components must be compensating to maintain yield. 35 Table 1.5. Mean response to directional selection for high and low seed weights. N84004 x P89405 Fm: families grown at E. Lansing, MI 1991. Seed Weight T . 'a't High Low Seed Weight (9/100 seeds) 30.9 25.2 Yield (grams) 36.9 35.1 Number of Pods 29.5 30.7 Number of Seeds per Pod 4.1 4.6 Number of Seeds 120.1 140.7 NS=Not significant at the 0.05 level. 36 The average number of pods, yield component X, did not significantly differ due to directional selection. However, yield component Y, the average number of seeds per pod, was significantly different in response to directional selection (Table 1.5) . The average number of seeds per pod had a significant negative correlation with seed weight (r=-0.45, P<=.001), signifying that as seed weight is increased, the average number of seeds per pod decreases. In the N84004 x P89405 population, the yield component seeds per pod compensated for lower seed weight to maintain yield. Directional selection for high or low seed weight also resulted in significant differences in the mean number of seeds (Table 1.5). Correlations between seed number and the yield components were all significant. The correlation coefficients for seed number with the average number of seeds per pod and with the average weight per seed were r=0.47 (P<.001) and r=-0.30 (P<.001), respectively. Pod number was highly correlated to seed number at r=0.82 (P<.001), therefore, number of pods could also compensate for yield. Indeed, the pinto ideotype developed through recurrent selection (Kelly and Adams, 1987) had larger and fewer pods than did the navy ideotype (Acquaah et al., 1991). Yield component compensation is also present in barley where selection for an increased number of heads results in lower kernel weight or fewer kernels per head (Benbelkacem et al., 1984). 37 Due to yield component compensation, the pinto ideotype differs from the architype defined for the navy bean (Adams, 1982) . Selection based on the pinto ideotype should be successful in recovering medium-seeded, type II plants. Our results provide evidence that medium seed size and plant architecture are inherited independently and can be combined based on the pinto ideotype. e ' o ' e e A strong association between seed size and growth habit exists in common bean. In fact, naturally occurring medium-seeded.materials with type II architecture are rare as evident from evaluation of accessions at the CIAT germplasm bank (Kelly and Adams, 1987). Singh et al. (1991) used the association between growth habit and seed weight in combination with other morphological and molecular traits to classify races within P. vulgaris. The lack of recombination between type II growth habit and seed size may be attributed to the lack of adequate breeding systems to combine the desirable traits of two diverse germplasm pools. Recurrent selection however, allowed for the gradual recombination of traits from the architectural germplasm pool with the seed-size germplasm pool. The use of recurrent selection in self-pollinated crops is limited by the time and expense necessary to make a large number of crosses (Avey et al., 1982). Recurrent selection has been successfully used for 38 increased phaseolin in common bean (Delaney and Bliss, 1991), early heading in winter wheat (Avey et al., 1982), kernel weight in spring wheat (Busch and.Kofoid, 1982), and seed size in soybean (Tinius et al., 1991). The architectural germplasm pool of the common bean is characterized by acute lower branches, stiff stem, fruiting on basal and upper nodes, small seeds, and 6-8 seeds per pod. In contrast to the architectural types, the seed-size germplasm has prostrate branches, weak stem, fruiting concentrated in basal nodes, medium sized seeds, and 4-5 seeds.per pod (Singh, 1989; Singh et al., 1991). The structural characteristics- related to yield in both germplasm pools belong to three developmental patterns: size (weight),‘ number, and architectural factors (Denis and Adams, 1978) . The pinto ideotype resulting from recurrent selection was balanced for all three factors where source was balanced against sink. There are many different combinations which would constitute a genotypic balance for such morphological and physiological characteristics (Denis and Adams, 1978). A cyclic breeding scheme such as recurrent selection was necessary to rearrange the balanced constituents of the parental germplasm into a new order of balance. The pinto ideotype represents the most balanced combination of the desired morphological and physiological components of the architectural and seed-size germplasm pools. 39 CONCLUSION Plant architecture is a complex trait that appears to have a relatively simple mode of inheritance. Desirable plant architecture was recovered in-toto in the initial cycles of recurrent selection (Kelly and Adams, 1987) and our data show that plant architecture is a highly heritable trait. The data also indicate that architecture and seed weight are not negatively linked, in contrast to the evidence of Kelly and Adams (1987). A negative linkage would appear to exist if selection were based visually for the pod traits of the navy architype. The navy architype pod traits, which include a large number of pods per plant and many seeds per pod, may visually appear to favor increased yield. When in fact, selection for increased number of seeds per pod or pods per m3, results in reduced seed weight (Nienhuis and Singh, 1986). Selection on this basis would therefore result in the development of small-seeded architypes. Early generation (81) F, recurrent selection was an appropriate method to combine the architectural traits of the navy germplasm with the seed size, shape and color of the pinto germplasm. Backcross breeding, as an alternative to phenotypic recurrent selection, would not have been an efficient method to introduce plant architecture into pinto bean germplasm. A cyclic breeding scheme such as recurrent selection was necessary to increase the frequency of favorable alleles 40 governing seed traits. The seed color and mettle traits of the pinto commercial class were recovered as early as C1, whereas, the major genes for seed size were not recovered until later cycles (Kelly and Adams, 1987). The additional cycles were necessary to maintain upright plant architecture while increasing seed size. Coyne (1968) showed that seed size is an additive trait that may be governed by as many as ten effective factors in broad crosses with large differences in seed size (Motto et al., 1978). The number of cycles necessary to combine erect plant architecture and medium seed size in a balanced relationship may have been reduced if selection were based on the pod traits of the pinto bean ideotype in earlier recurrent selection cycles. Selection for medium-seeded, upright plants should be based on the pinto ideotype which is physically different from the navy architype (Figure 1.1) . Due to yield component compensation, the pinto ideotype proposed for the humid midwestern United States exhibits fewer seeds per pod and fewer pods per plant than the navy architype. Both traits can be visually selected, allowing for the rapid recovery of medium seed size in combination with desired type II plant architecture. 41 Figure 1.1. Comparison of the navy architype, N84004, and the pinto ideotype, P89405. P89405 displays modified pod traits, allowing for the recovery of medium seed size. 42 LIST OF REFERENCES Acquaah, G. 1987. Genetic and multivariate statistical evaluation of a phenotypic recurrent selection program for recombining erect architecture and large seed size in Phaseolus vulgaris L. Ph.D. Dissertation, Michigan State University, East Lansing, MI. 317 pages. Acquaah, G., M.W. Adams, and J.D. Kelly. 1989. Broad-sense heritability estimates of several architectural traits in dry beans (Phaseolus vulgaris L.). Annu Rept Bean Improv Coop 32:75-76. Acquaah, G., M.W. Adams, and J.D. Kelly. 1991. Identification of effective indicators of erect plant architecture in dry bean (Phaseolus vulgaris L.). Crop Sci 31:261-264. Adams, M.W. 1967. Basis of yield component compensation in crop plants with special reference to the field bean Phaseolus vulgaris. Crop Sci 7:505-510. Adams, M.W. 1973. Plant architecture and physiological efficiency in the field bean. In: Potentials of field beans and other food legumes in Latin America. CIAT, Cali, Columbia. pp 266-278. Adams, M.W. 1982. Plant architecture and yield breeding. Iowa State J Res 56:225-254. Avey, D.P., M.W. Ohm, F.L. Patterson, and W.E. Nyquist. 1982. Three cycles of recurrent selection for early heading in winter wheat. Crop Sci 22:908-911. Bartley, B.G. and C.R. Weber. 1952. Heritable and non- heritable relationships and variability of agronomic character in successive generations of soybean crosses. Agron J 44:487-492. Benbelkacem, A., M.S. Mekni, and D.C. Rasmussen. 1984. Breeding for high tiller number and yield in barley. Crop Sci 24:968-972. Blad, 8.L., J.R. Steadman, and A. Weiss. 1978. Canopy structure and irrigation influence white mold disease and microclimate of dry edible beans. Phytopathology 68:1431-1437. Busch, R.H, and K. Kofoid. 1982. Recurrent selection for kernel weight in spring wheat. Crop Sci 22:568-572. 43 Coyne, D.P. 1968. Correlation, heritability and selection of yield components in field beans, Phaseolus vulgaris L. Proc Am Soc Hort Sci 93:388-396. Delaney, D.E., and F.A. Bliss. 1991. Selection for increased percentage phaseolin in common bean. 2. Changes in frequency of seed protein alleles with 81 family recurrent selection. TAG 81:306-311. Denis, J.C., and M.W. Adams. 1978. A factor analysis of plant variables related to yield in dry beans. Crop Sci 18:74-78. Falconer, D.S. 1960. Introduction to quantitative genetics. MacLehose and Co, Ltd, Great Britain. Fehr, W.R. 1987. Principles of cultivar development, Vol 1; theory and technique. McGraw-Hill, Inc; New York. Frazier, W.A., J.R. Baggett, and W.A. Sistrunk. 1958. Transfer of certain Blue Lake pole bean pod characters to bush beans. Proc Amer Soc Hort Sci 71:416-421. Kelly, J.D. and M.W. Adams. 1987. Phenotypic recurrent selection in ideotype breeding of pinto beans. Euphytica 36:69-80. Kelly, J.D., M.W. Adams, A.W. Saettler, G.L. Hosfield and A. Ghaderi. 1984. Registration of C-20 navy bean. Crop Sci 24:822. Kelly, J.D., M.W. Adams, and G.V. Varner. 1987. Yield stability of determinant and indeterminate dry bean cultivars. Theor Appl Genet 74:516-521. Lush, J.L. 1940. Intra-sire correlations or regressions of offspring on dam as a method of estimating heritability of characteristics. Proc Am Soc An Prod :293-301. Lush, J.L. 1945. Animal breeding plans. Iowa State Univ Press, Ames. Motto, M., G.P. Soressi, and F. Salamini. 1978. Seed size inheritance in a cross between wild and cultivated common beans (Phaseolus vulgaris L.). Genetica 49:31- 36. Nienhuis, J. and S.P. Singh. 1986. Combining ability analyses and relationships among yield components, and architectural traits in dry beans. Crop Sci 26:21-27. 44 Nienhuis, J. and S.P. Singh. 1988. Genetics of seed yield and its components in common bean (Phaseolus vulgaris L.) of Middle-American origin. II. Genetic variance, heritability and expected response from selection. Plant Breed 101:155-163. Singh, S.P. 1982. A key for identification of different growth habits of Phaseolus vulgaris L. Annu Rept Bean Improv Coop 25:92-95. Singh, S.P. 1989. Patterns of variation in cultivated common bean (Phaseelus vulgaris, Fabaceae). Econ Bot 43:39-57. Singh, S.P., P. Gepts, and D.Debouck. 1991. Races of common bean (Phaseolus vulgaris, Fabaceae). Econ Bot 45:379-396. Sullivan, J.G. and F.A. Bliss. 1983. Genetic control of quantitative variation in phaseolin seed protein of common bean. J Amer Soc Hort Sci 108:782-787. Theday, J.M. and J.N. Thompson, Jr. 1976. The number of segregating genes implied by continuous variation. Genetica 46:335-344. Tinius, C.N., J.W. Burton, and T.E. Carter, Jr. 1991. Recurrent selection for seed size in soybean: I. Response to selection in replicate populations. Crop Sci 31:1137-1141. Uhr, D.V. and J.P. Murphy. 1992. Heritability of oat mosaic resistance. Crop Sci 32:328-331. Vallejos, C.E. and C.D. Chase. 1991. Linkage between isozyme markers and a locus affecting seed size in Phaseolus vulgaris L. TAG 81:413-419. Warner, J.N. 1952. A method for estimating heritability. Agron J 44:427-430. Wright, S. 1934. The results of crosses between inbred strains of guinea pigs differing in number of digits. Genetics 19:537-551. CHAPTER TWO ALLOZYME EVALUATION OF UPRIGHT PLANT GENOTYPES ABSTRACT Advanced navy, pinto and great northern breeding lines and cultivars with upright type II plant architecture were analyzed for an allozyme variant of the enzyme diaphorase (DIA) . The Diap-z105 allele, characteristic of the Unique DIA zymotype, was present in 71% of genotypes with type II plant architecture. The genetic relationship between upright plant architecture and the Diap-z105 allele was further investigated in F, populations derived from crosses between navy and pinto bean parental lines differing for type I, II, and III growth 2105 allele associated with habit and DIA genotype. The Diap- upright plant architecture was not linked to the architectural complex or the architectural traits of branch angle, hypocotyl diameter, and plant height. Directional selection for Diap-Z105 did not result in significant architectural differences. Due to random drift, the Diap-z105 allele, initially associated with type II architecture through founder effect, may be maintained in such genotypes without providing greater fitness or without being associated with a locus or linked loci governing upright plant architecture. 45 46 INTRODUCTION Allozyme variants have been used extensively to: assess genetic variability (Brown et al., 1978; Kesseli and Michelmore, 1986; Weeden, 1983), provide information on evolutionary and genetic relationships (Bassiri and Adams, 1978a; Bassiri and Adams, 1978b; Koenig and Gepts, 1989b; Singh et al., 1991b; Wall and Wall, 1975), determine linkage groups (Benches and Quires, 1989; Koenig and Gepts, 1989a; Muehlbauer et al., 1989; Smed et al., 1989), and create linkage maps (Tanksley and Rick, 1980; Weeden and Marx, 1984) . Allozyme variants of six isozyme systems separate the common bean (Phaseolus vulgaris) into small-seeded Mesoamerican and large-seeded Andean gene pools (Sprecher, 1988b; Koenig and Gepts, 1989b). In addition, allozyme variants also provided indications of gene flow from wild to cultivated P. vulgaris and supported the separation of five subgroups within the Mesoamerican gene pool and four subgroups within the Andean gene pool (Singh et al., 1991b). Sprecher (1988a) reported that an.allozyme variant of the enzyme diaphorase (DIA) is associated with tropical black upright genotypes of race Mesoamerica of the Mesoamerican gene pool (Singh et al., 1991a). She determined that DIA is a. tetrameric enzyme coded by two tightly linked loci, Diap-l and Diap-z, with four and three alleles, respectively. Allelic variability at the Diap-l locus separated Pu vulgaris into Mesoamerican (Diap-l”) and Andean (Diap-lm) gene pools. 47 (Allozyme loci and alleles are designated as in Koenig and Gepts, 1989b.) Allelic variability at the Diap-z locus differentiated within the Mesoamerican gene pool. The allele, Diap-zm, commonly exists in both Andean and Mesoamerican materials however, the Map-2“” allele is present only in race Mesoamerica (Sprecher, 1988a; Koenig and Gepts, 1989b; Singh et al., 1991b). Furthermore, both Diap-2‘°° and Diap-2‘°5 alleles are present in cultivated P. vulgaris whereas, wild materials possess only the Diap-Z10° allele (Koenig and Gepts, 1989b; Singh et al., 1991b). The Diap-Zw'5 allele which is associated with tropical black/upright genotypes (Sprecher, 1988a) possibly resulted from a mutation during or after domestication of such genotypes (Singh et al., 1991b). The genotype Diap-1”/Diap-2‘°5 produces the 'Unique' zymotype and may have evolutionary significance because of its close association with race Mesoamerica of P. vulgaris (Sprecher, 1988b) . Included within this race are the tropical black genotypes (Singh et al., 1991a) used for the architectural improvement of dry beans (Kelly and Adams, 1987). In an isozyme survey of upright type II beans, Sprecher (1988b) reported that 75% were homozygous for the Map-2‘” allele. A linkage with upright plant architecture is suggested because the Diap-z105 allele was maintained through indirect selection. 48 The objectives of this investigation were (1) to determine the genetic relationship between the Diap-z105 allele and upright, erect plant architecture and (2) if the Diap-z105 allele is not linked to any component of the architectural complex, determine why the allele is associated with type II plant architecture. MATERIALS AND METHODS Gen- ' e at'onsifo betwee - - ._1, . ch’te - on. tie _ 105 1e Plant materials Advanced Michigan State University (MSU) breeding lines and cultivars representing navy, pinto, and great northern commercial classes were surveyed for DIA to determine the frequency of the Diap-z105 allele among advanced Mesoamerican genotypes. The lines evaluated in this investigation exhibited mainly type II plant architecture but some type I and type III lines were also included (Kelly personal communication). Two plants were analyzed for each pinto and great northern line, whereas, four plants were evaluated per navy line. Eight F, populations segregating at the Diap-z locus were evaluated to study the genetic relationship between upright plant architecture and the Diap-Z105 allele. The bean lines used to generate the populations represent both navy and pinto 49 bean commercial classes and differed in type I, II, and III growth habit and DIA zymotype (Table 2.1). The populations studied were crosses between zymotypes and included: Seafarer x N84004 (Type I, Slow x Type II, Unique) Seafarer x P89405 (Type I, Slow x Type II, Unique) N84004 x Michelite (Type II, Unique x Type III, Slow) N84004 x Ouray (Type II, Unique x Type I, Slow) N84004 x UI-114 (Type II, Unique x Type III, Slow) Michelite x P89405 (Type II, Slow x Type II, Unique) Ouray x P89405 (Type I, Slow x Type II, Unique) P89405 x UI-114 (Type II, Unique x Type III, Slow) Approximately 320 seeds per F, population were planted at a 20 cm spacing within rows and a 51 cm spacing between rows at East Lansing, Michigan, 1990. Where emergence was poor a type I kidney bean, Isabella, was hand planted two weeks later to ensure uniform interplant competition. Parental lines were included as checks on each side of the F,jpopulations. Isozyme analysis Young, fully expanded leaves sampled from 16-42 day old field grown plants were analyzed for DIA. All steps were carried out on ice or at 4°C unless otherwise specified. Two leaf disks, 1.0 cm in diameter, were placed in 1.5 ml eppendorf tubes. Potassium phosphate grinding buffer (0.35 ul) modified from Weeden (1984) was added to each eppendorf. Specifically, 0.25% Triton X-100, a detergent .25va w ease: stead m __ =95 >52 oo—NiQmmn . See: also e __ 528.2 8:... 832 _ 0 _ s geese _ 32m Seems m _ c5522 oEE >830 _ geese 32m 2166.5 e _ =mEm >>mz eoeoeoom w 8.N-Qm.~Q W 32m seems N z. :35 >62 2:225. _ one—Niqmmq 32m erase _ =. 523.2 85.. 3.-.: i oa>eoE>N 35950 956m «56.1 3% «no.0 on: 389350 329320 2308203 539.0 poem _a_o._oEEoo ccom W I J ooa\m e: advocate—swoon one See .mooon ooa\o o3 cocoouiaacam mo mowuuauouocuecu .uocaa Heucouom econ oucam .uooom .H. m manna 51 added to enhance the release of enzymes from membrane bound organelles (Kephart, 1990) , was added rather than Weeden's (1984) recommended 0.5% concentration. The 0.08 M potassium phosphate grinding buffer also contained 20% sucrose (w/v) , 5% PVP-40 and 14 mM 2-mercaptoethanol. The Triton X-100 and 2-mercaptoethanel were added just prior to use. To release enzymes, tissue was macerated using a 'Con-Torque' mechanical grinder. Following homogenization, extracts were centrifuged in a refrigerated microfuge (2-4°C) at 12,000 rpm for 5 minutes to remove cellular debris and produce zymograms with sharper resolution and less streaking (Wendel and Weeden, 1989) . The purified homogenate was absorbed into 3 mm x 10 mm chromatography paper wicks and stored at -15°C for 24 to 48 hours prior to use. The electrophoresis buffers used were modified from Sprecher and Vallejos' (1989) lithium borate/tris citrate (Li-Bo) discontinuous system. Diaphorase was resolved on a 10% (w/v) starch gel prepared 18 to 36 hours in advance using Buffer A (29 mM lithium hydroxide and 192 mM boric acid, pH 8.1) and Buffer B (6 mM citrate and 51 mM lithium hydroxide, pH 8.4) in a 1:9 ratio. Twenty-two grams of starch in combination with 20 ml of Buffer A and 200 ml of Buffer 8 produced a gel of adequate thickness to obtain three slices. The buffer reservoirs were filled with Buffer A. Wicks from 18 segregating samples along with two parental samples were inserted into a vertical slice 4 cm from the base 52 of the gel. Electrophoresis was carried out at 45 mA for ten minutes to load the proteins into the gel. The wicks were then removed and electrophoresis resumed at 38 mA for 4.5-5.5 hours at 2-4°C. The DIA assay contained in 50 ml: 100 mM Tris-HCl pH 8.5, 14 mg beta-NADH, 20 mg MTT (3-[4,5-Dimethylthiazol-2-yl]- 2,5 diphenyltetrazolium bromide) and 1.0 mg 2,6-dichlorophenel indophenol (Weeden, 1984). Both anodal and cathodal sections of a gel slice 1.5 mm thick was placed in.a plastic tray along with the staining solution. The gel was incubated in the dark at 37°C and scored after two hours. The second most abaxial slice of the three slice gel produced the most scorable and repeatable results. Agronomic evaluation All plants analyzed for DIA were rated at maturity for plant architecture on a 1 to 9 scale and evaluated for traits contributing to plant architecture as described in Chapter 1. The weight of 50 seeds was also determined. I 0! se -c ., o 9 r' .- um 9 21§2_,105 a] J g, e Four populations derived from the crosses N84004 x UI-114, Michelite x P89405, Ouray x P89405 and P89405 x UI-114 were selected for inclusion in a directional selection.experiment~ 'The cross N84004 x 01-114 is typical of 53 crosses initially made by Kelly and Adams (1987) to combine the plant architecture of the navy architype with the seed size of the prostrate, type III pinto. The Michelite x P89405 cross is the reciprocal, a prostrate, type III navy crossed to the pinto ideotype. The Ouray x P89405 and P89405 x UI-114 were included based on the results of the F, data. For each population, eight F, plants, or approximately 5% of the plants analyzed for DIA, were selected for both the Diap-z105 and the Diap-z100 alleles for a total of 16 F,:3 families. The only restriction for inclusion in the study was that the F, plants produced 60 or more seeds. O Wenty seeds derived from a single F, plant were spaced planted (20 cm spacing) in a row and replicated three times in a completely randomized block design at East Lansing, Michigan, 1991. Plant architecture and loo-seed weight were determined for each family row. flee ef enthzacnese :eaction as a non-linked locus Twenty Fa: families from the cross N84004 x UI-114 were 'nitially selected for high (10 F,:3 families) and low (10 F,:3 families) architecture rating (Chapter 1). The F,“ progenies were screened in the greenhouse for resistance to the alpha race of anthracnose (Colletotrichum lindemuthianum) . Dominant genes at the A locus in N84004 confers resistance to race alpha (Cardenas et al., 1964). In a normally segregating population approximately 56.3% of the N84004 x UI-114 F,“ 54 progenies would be expected to be resistant to race alpha. However, selection for plant architecture may maintain a greater frequency of non-linked loci from the architectural germplasm pool. Detection of a significantly higher frequency of the A locus in architecturally desirable lines may be used as an indication of the representative contribution of the architectural germplasm pool in the final progeny. Five to ten.F,“ plants per N84004 x UI-114 family were inoculated. ten. days after emergence *with, race alpha of C. lindemuthianum and placed in a mist chamber for 48 hours. The plants were evaluated for disease reaction after uniform symptoms developed on UI-114, the susceptible parental check. N84004, the navy architype, is resistant to race alpha. Reaction to anthracnose was classified as resistant, intermediate, or susceptible. RESULTS AND DISCUSSION 0 s z s Diaphorase isozymes are most clearly expressed in root tissue, however, in this investigation, leaf tissue was used for the DIA analysis. ILeaf and root tissues produce zymograms with slightly different expression of isozymes, but it is possible to distinguish between isozyme variants. Zymograms exhibited by random samples of root tissue were in concordance with their corresponding leaf tissue extract. 55 The three DIA zymograms observed with greatest frequency were the 'Slow,' the 'Unique,‘ and the heterozygote between the Slow and Unique banding patterns (Figure 2.1). Fast and Unique/ Fast heterozygous zymograms were additionally observed in breeding lines surveyed for DIA. Zymograms produced by root tissue sampled from genotypes homozygous at both loci display five dark staining bands. Genotypes heterozygous only at the Diap-Z locus (Slow x Unigue heterozygote) produce 15 bands which is in agreement with Shaw's (1964) formula to predict the number of isozymes formed by multimeric enzymes. Leaf tissue zymograms do not produce banding patterns characteristic of a tetrameric enzyme coded by two alleles. Homotetramers produced by the Diap-Z locus appear after staining whereas homotetramers produced by the Diap-l locus cannot be visualized (Figure 2.1) . Genotypes heterozygous at the Diap-Z locus exhibit additional bands intermediate to the Diap-Z‘os and the Diap-2'°° allozymes. Characterization of the DIA genotype was based on the intensely staining bands with relative migration distances (Rf, as calculated by Sprecher, 1988b) of -0.20, 0.05, 0.10, and 0.30. The interpretation of shoot zymograms was verified in some cases by the clearer reaction from the root tissue extracts. In addition, leaf extracts from known root DIA genotypes were included as checks in each gel. Although leaf and root zymograms differ, they represent the same DIA genotype as was verified by root and shoot protein extracts 3mg 1 m 33:33.5 1 is 1053.5 1 5 105233on 1 3m .5on n my .omernum .eemummE “Senna .uoaeuumuoeoz skimmed “moénum .uofinuuouoaon 87¢er unfouum up pounced uofiouuouoaon Santana .3093 ecu um eeueoea ma :fimfluo .uuoa ecu :o eu>Hm one AnmmeV Honouumm an pennaeoamo mm Aumv mooenumwe :ofiunuoflx .uuomuuxo manna» .5 been can .3 and." Sony menuwogu emeuocmnfln .H.N Tasman ION.i 56 mmlnta :_83 Tex 8d mt 9. cm em 838 x 9:222 edA one at 3 we 3. 3 I: x 4832 «SA 33 at we on 5 550 x 4832 SA 8; mt we we we 8:225. x 39.»: BSA 86 e: an we me 838 x 3.38m «SA 5d «2 em «a we 48.32 x Esteem an N. can : more? more? 872: . E Nx .88. 8:238 m. .omma .Hz .oeemeeq .m me_o__< wins .uuguocom S..8.NIQT.HQ\8.BHIQT.HQ an“; mochuocoo Nimer\8.8HiQefiQ mo mouuouu Beau wooed Nimofln on» an cowunmoumom 0.30:2 .m.m canon. 38..» an .3 use «o $05.39 2023303 e -ll|- I ll - . . ll- ll|l llll llllrllL 3X and - 8.- . r . ....F -- .- . m- 1 . , -N .. -. - - as: L mod A mmé amp hm mm 5 3238.25 L edA 5o 8 e .2 e 26.. 838. x 8:225. L m.oA 8.6 m: m e e :9... L To A 8.2 5: oe me we 828.228... L sex 36 3 e e e 2,3 S I: x 4832 L 2 ToA 8.4 e m P o :9: L 6 _ 88A 43.. B. an 2 ma 82858:. L L eoA 3.6 on «.— em 8 26.. 3 I: x Semen. L , eoA 8.6 m. e m e :9... L L To A «e. e 8: on em on 82858:. L L 3A Se 2 o _ o 25.. 838. x >950 L E. N. Eu” : moved. more? 878. 3888 8338 m. L L E «x .88 888262 ,_ mo_o=< NONE .omm." H: 59.3an .m .mcofieoaecom ~.m been no mofiuomouno 353323.. 8 85 .e .t :3: one .3 use 6 .3 3380535 .8 one .N .3 33 you moguoeoe Steeewimefiexmefiwimewn mo meooa Nimewn on» no cofiunmoumom OMHTHH< .m.~ manna. 63 Linkage between the Diap-z105 allele and loci determining the quantitative trait (QTL) , plant architecture, was also investigated. Analysis of variance of plant architecture rating for different Diap-z genotypic classes (Diap-zws'ms, 2105,100 2100,100) Diap- and Diap- was significant for only one population, P89405 x UI-114 (Table 8.4) . Orthogonal contrasts as described by Osborn et al. (1987) were conducted to compare the mean architectural rating of the Diap-Zws'105 homozygous genotypic class with the mean of the Diapaz'm"100 homozygous genotypic class and to compare the mean architectural rating of the two Diap-z homozygous classes combined, with the mean of the Map-21°53” heterozygous class. Both contrasts were significant in all but one of the F2 populations (Table 3.4) . However, the genotypic class with the highest average plant architecture rating was not consistent throughout all eight F2 populations. In only two populations, P89405 x 01-114, and Ouray x P89405, the DRIP-2m"105 homozygous genotypic class had the highest average architecture rating, whereas, in four populations the Diap-zma'100 homozygous class had the highest average architecture rating. Therefore, because a consistent trend was not present, the contrasts, although significant, did not indicate a linkage between the Diastp-z105 allele and a locus or group of linked loci governing plant architecture. Directional selection for the Diap-Z‘os allele versus the Diap--2100 allele did not result in significant architectural 64 differences. In all four populations, the average architectural ratings for the Diap-lw/Diap-Z105 F2:3 families was not significantly higher than the average architecture rating of the Diap-1”/D.iap-2‘°° F25 families. The results of our directional selection study confirm that type II plant architecture and the Diem-2105 allele segregate independently. The crosses N84004 x Ouray, Seafarer x N84004, Ouray x P89405, and Seafarer x P89405 were further evaluated for linkage associations between the Diap-st allele and indeterminate versus determinant growth habit. In the four l"2 populations, allelic segregation at the Diap-z locus in combination with growth habit indicated that Diap-z105 was independent of growth habit (Table 2.4). Our results indicate that Diap-z105 is not linked to plant architecture or any of the three architectural traits evaluated. The association between the Dialp-Z105 allele and type II plant architecture may be due to founder effect and subsequent selection for a larger portion of race Mesoamerica parentage needed to refine upright plant architecture. The major constraint to founder effect is that a relatively small number of individuals, or colonizers, are involved in the establishment of a new population (Mayr, 1942). The genetic variability of the resulting population is limited to the alleles introduced by the colonizing individuals (Schwaegerle and Schaal, 1979) . A mutation occurring simultaneously with 65 .ofiuuu casamocm: aunumumuaun umuomdxm mg» on ufiu no mmmcuoomc no m .sufiafinmnoum # .muwnmc nusoum mumcwaumumc can mumswaumumpcw auficmwm a can H .>Hm>wuommmmu .SFNIQoflQ you msomxuoaon can .msovauouwumn .8.N..Q3d you msomhuoaon «D .m .m h -- - ----m...co_§=...o...JL 8.? 2.. m8 an m? 3 8m 8 mm. m. 898 8.8.. _ L m.oA mm; of m um 8 8 m mm 88?. x 22.6 9? 92m N: 9 mm 2 8 S 8 .625 x 2882 _ 8A 3% mt m 8 a 3 m mm 88?. 5258.0. L 8A Sam at m 8 a K o. .5 «882 x .2325 L «a. Nx .33. 03 .3 Q: 5.. Qm uoahn msomhnoumuonumxo .3onum .msqficauo .umamum .mcwa non adamamm usom » .Ammmd .svcfimv mvmuumoum macawaumumoswnHHH umcfi> ocoa saws unawums muncfiaumumocwunHH mocw> uuonm saw: unmwums ouncfiahmuopcwunHH “monosmun xmma sags swan mumsfiaumumcnnH “swan mumswaumumc unmwumsumu h mod cod cod moH mos mod mofl mod mod mod mod cod ooa con cod ooH ooH cos ooH 883 can mca.ooa cod moa mca.ooa Numuflm Illlllllwudwaad .mnauocom «He now cmumsam>m mun>wuaso use mmcwa coma >>mc omocn>6¢ mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm «84.8 MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH QH QH nH 2H QH DH QH DH CH CH MH MH MH MH #MNNHIMHMAN DammtnmmtnmmmmmmDDDDDDDDmmD £565666655868655555555685 DDmammmmmmmmmmDDDDDDDDmt/ib DDWDUJUJUJUJUJIDUJU)UJUJ’DDDDDDDDUIUJD sum\ 3 Mesoauan: wHoomz mHmomz «Hmomz monomz momomz nNmomz mmeomz OHwomz Nocomz eoovmz omawcmm ucwumm Hflmnouflz CCMHGHS oooaummam nonmau Maoomsa umumumom “mu—3 ammomz mmmomz ommomz nbmomz ovmomz waddlfiflwfl .H.m OHQMB 81 .mm>uoa>~ msowwuoumuonum\a .3oamum .OSUHCDHD .umnmum .mcwa you moamamm Room » .Ammma .nmcfimv mumuumoua mumsfiaumumpcauHHH «0:«> OCOH suds unofiums muncflaumumccwunHH umcfi> uuosm saw: unqfluds mumcwaumumpcfiuMHH “mucocmun xems saw: swan mumcwauoumpunH “swan oumcfieumump ucmwumsncH h cod mm mumumum HHH muHHmnOHZ mod mm Danna“: nHH cocmauw OOH mm mumumum nHH poo3um0uu ooa mm mumumum nHH nwamuucmu moa mm Dunno“: nHH umcoosom moa.ooa mm Dumuaum nHH omio mod mm Dunno“: mHH Hmaam> cmsm moa mm Dunno": mHH manuaoz moa mm Dunno“: mHH Hucm>< moH mm Dunno": MHH unumnuuoz Mummflm Mummflm Humummmuu #mmNHIMQMHN «aflqlquwu manuaad .26.»:008 H.m manna 82 .mnauosan msovhuoumumcumxa .Boamum .msvacaua .umnmnm .mmmH .umnumumw m .mcwa you mmHQamm 038 # ..~mma .nacfim. oucuumoun muucfieumumoafiuHHH «02.; unocm nu“: anode—Q: mumcwamumocwun: «monocnun 8.803 and: £95 oumcwahmumcunH .P mod ooa moH ooH moH.ooH mod 883 can 883 ooH ooa mod mod mod mod moa.ooa mod mod mod mos moa mod mod mod ooa Mumuflu illlllllwmauaflm .mm>uocmu «Ho you coumsHm>m wun>wuaso can mmcwa coon oucwm poocn>c¢ mm mm mm mm ooa.mm mm 883 cod cod mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm unmmwu MHH MHH MHH MHH MHH MHH mHH MHH MHH MHH mHH MHH MHH mHH MHH MHH MHH MHH MHH mHH MHH MHH mHH MHH DH samualmqmdu 555555555555556fififiéfififisas mDDDDDDDDmDDDmemmmDmmamD HMQNMGHNN numomm Hhmomm Ohmomm mmmomm mmmomm hmmomm mmmomm momomm ¢mmomm nwmomm memomm Hemomm mmmomm wmmomm bmmomm memomm meomm mnmomm nmmomm Nmmomm homomm nmfimmm eHemwm mocmmm hnuzo mudqlmmmm .N.m OHQMB .mmma .umnowudm m .mQHHOEHN maoahnoumumcumxa .3oamum .msvwcsub .umamum .mcwa you moadamm 039 H .Ammma .nmcwmv munuumoum mumcwfiuoumpcwuHHH “0:82 phone cu“: Havana: ouncfiauouopcflum: “monocmun 8303 :33 swan ouncwfiumumpunun .P ooH mm mum HHH enumHo ooa.moa ma mun HHH candmua 883 mm mum mHH .omozc mamomH mod mm a": 8HH omunc moa mm a": «HH muumfim mod mm on: mHH soaflmm oca mm mum mHH madame moa.ooa mm sum mHH moaamm ooH ooa.mm mum mHH «oaamm mod mm a": 8HH madame mod mm a“: 8HH madame 3 ooa mm mum mHH Hoaamm 3 mod mm on: mHH eamomm moa.ooa oofl.mm m\aua uHH nmmomm cod ooa mum «Ha «mmomm 883 883 mum nHH ammomm can ooa mum mHH mmmomm cod ooa mum mHH mmmomm 883 883 mum mHH smmomm mod mm a": «HH nmmomm mod mm a”: MHH uwmomm mod mm a": «HH ammomm cod ooH mum uHH owmomm 883 mm mum mHH mhmcmm moa mm a”: 8HH memomm moa.oca ooa.mm m": mHH psmomm cod cos mum mHH «emomm «umuwm «umqu «umumqmwu tmmxmlmqmam mqflqlqmwm uuauad< .Ac.ucoo. ~.m manna 84 .mmma .umnomudm m .mmhuoahu msooanoumumsum\o .30Hmum .osvwcbuo .umnmnm .OCHH you moamamm 038 » .Ammma .nmcfimc mumuumoud mumcweumumecfiuHHH “05$ unocm 553 ”5359.. mumcfigoum—uawueHH «mecca—mun 8303 and; swan mumcwamuoounH .P cod mm mum HHH «HHiHa cod mm mmum HHH xnecfim cod mm mum HHH oHHmnuo Mumuflu «ummflm “mmummaxu Causalmq84m unflqlqmmu wuawaa< .Ae.ucoov «.m 83589 .mmhuoahu anomauoumumnumxa .3onmum .osvwcoub .umnmnm .mme .umcomumm m .mcaa you moaaanm 038 « Ammma .nvcwmv munuumoun .mumcwsuouochuHHH mocw> uuonm saws Havana: mumcwaumumocquHH 2 mos mm moa.ooa mm mod mm moa mm moH mm mod mm mos mm mod mm mod mm mod ma moa mm 5 mod mm 8 mod mm mod mm 883 mm ooa mm 883 mm mod mm mod mm cod mm mos mm mod mm mod mm mod mm mod mm «umuwm «Human nuawaac H D D MHH MHH mHH MHH QHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH MHH :3Stat):ca:>Dcnaam:::>D:3:>D:3:>D:::>Dcn D=D=>D€DU)D=305mCDSJDEDEJD:DSJD:3=>D:D=> as 888183.... 4.) .30 you 00.82325 muu>wuHso use was: soon :umnuuo: unmum 0005264 wchmU NmHomw mmHomU HmHomU mnHomU mnHomU hnHOmU mnHomU mnHomU HnHOmU OMHomw mNHomw mNHomU bNHomU mNHoaU mNHomU emHomw 50Homw mOHOmU mOHomU vOHomU MOHOQU NOHomU HOHomw onwnad mg “4 Emma .mmauocmo . n .m 0839 .mmma .umnomuam m .mm>uoa>N maomanououozuwxo .30Hmum .mavflcoub .ummmum .mCHH Han mmamamm 039 H Awmma .nwcwmv oumuumoum .mumcaaumHOGCauHHH umcfi> uuonm spas Havana: oumcwaumumGCHHMHH # ooH mm mmum HHH magnum ooH mm mmum HHH aoaam> cod no mmum HHH Haumm ooH mm mmum HHH unwfiauwum cod mm mmum HHH mmuHa ooa ooa mum mHH vamamw 5 mod mm a“: «HH namamw 8 mod mm Duo mHH mamamu moH.ooH mm m": mHH Hamamu moa.ooa mm m”: mHH oamamu moa.ooa mm mum\a mHH momamu mod mm on: uHH .momamu ooa mm mum mHH pomamw moa mm on: mHH momamw mod mm a“; mHH momamo no” no pun mHH «omamu mod mm on: mHH nomamu moa mm can, aHH momamu mod mm a“: «HH Homamo cod cod mum mHH mmaomw ooa ooa mum mHH omaomu ooa mm mum mHH Huaomo NHQMflq «umwflu ”mmxumanu #umualuqmam mudqlqmwm moawaac ..c.uaoov n.m manna MICHIGAN smTE UNIV. LIBRQRIES MW‘IIWIHINIINIWIH\IWIIHHIHMWWI 31293010263402