// /" / "I ( .r’" f .‘3 ,/ 3. /./. V 4// _ I :f’ 2;" , , A W. J13- . ;" , /.’ INHERITANCE OF FACTORS AFFECTING INFLORESCENCE TYPE AND NUMBER OF FLOWERS ON. THE INFLORESCENCE IN TOMATO Lycopersicon esculentum Mill. Thesis for the Degree of Ph. D. ' MICHIGAN STATE UNIVERSITY JERRY DALE VRIESENGA 1 9 72 ABSTRACT INHERITANCE OF FACTORS AFFECTING INFLORESCENCE TYPE AND NUMBER OF FLOWERS ON THE INFLORESCENCE IN TOMATO Lycopersicon esculentum Mill. BY Jerry Dale Vriesenga Studies on the inheritance of a single flower per truss, compound inflorescence, simple inflorescence and low-flower- number inflorescence were carried out. .The mode of inheri- tance for number of flowers and branching of the simple inflorescence was also investigated. It was found that the character single flower per truss was determined by a recessive gene-from crosses MSU 100 (single flower) x MSU 180 (simple inflorescence) and MSU 100 (single flower) x Pennorange (low-flower-number inflorescence). The Pennorange phenotype--low-flower-number inflorescence-- was conditioned by a recessive gene and the gene was epis- tatic to the gene for single flower. ‘ The mode in inheritance for the single flower character was more complex in crosses with compound inflorescence parents, Apsory and MSU 200. A model was proposed including the recessive single flower gene (a), an inhibitor gene (1) which inhibits the expression of single flower and a restorer gene (3) which negates the effect of the inhibitor. Jerry Dale Vriesenga The single flower gene (a) was suggested to be epistatic to the gene for compound inflorescence. Compound inflorescence was suggested to be conditioned by two genes, intense branching and non-terminal flowering (gtf). The genes were linked with recombination values of 6% in the crosses MSU 100 x Apsory and Apsory x MSU 180, and 11% in the cross MSU 100 x MSU 200. Simple florescence was found to be dominant in all crosses observed. Modifiers affecting the number of flowers on the unbranched inflorescence were studied. rTwo classes, 4-flower and 8-flower were observed in the cross MSU 100 x Pennorange. No dominance was noted between the 4-flower class and the 8—flower class. Two classes were observed in the cross MSU 100 x MSU 200° The classes, lO—flower and 7kflower, segregate to a digenic 9:7 ratio of lO-flower class to 7-flower class. The inheritance of branched inflorescence was suggested to be related to the modifier conditioning the 4-flower and 8-flower classes in the cross MSU 100 x Pennorange. A single gene was suggested to affect branching in the cross MSU 100 x MSU 200 and the gene was independent from the number of flowers on the unbranched inflorescence. It was suggested that intercalary inflorescence (igi)-- indeterminant vegetative growth on the florescence—-was conditioned by one gene. Jerry Dale Vriesenga The gene for jointless pedicel (1) was suggested to be either closely linked or pleiotrOpic with the gene for inter— calary inflorescence (ini). INHERITANCE OF FACTORS AFFECTING INFLORESCENCE TYPE AND NUMBER OF FLOWERS ON THE INFLORESCENCE IN TOMATO Lycopersicon esculentum Mill. BY Jerry Dale Vriesenga A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Horticulture 1972 DEDICATION To Marcy and Our Children ii ACKNOWLEDGMENTS The author wishes to express his appreciation to Dr. S. Honma for his advice and criticism during the prepa— ration of this thesis. Sincere thanks are also extended to, Dr. M. W. Adams, Dr. L. R. Baker, Dr. R. L. Andersen and Dr. H. H. Murakishi for their recommendations in the prepa— ration of this manuscript. Appreciation is also extended to Dr. M. Jordanov for his assistance in initiating the thesis research and to Mr. Amos Lockwood for maintaining the field plots. Special thanks are tendered to my wife, Marcy, for her many hours of work in typing the manuscript and for her immeasurable encouragement and support. iii TABLE CF CONTENTS INTRODUCTION. . . . . . . . . . . . . . . . . . . . . REVIEW OF LITERATURE. . . . . . . . . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . . . . . . . PARENT MATERIAL. . . . . . . . . . . . . . . . . HYBRIDIZATION. . . . . . . . . . . . . . . . . . FIELD TRIAL. . . . . . . . . L . . . . . . . . . Summer 1969 . . . . . . . . . . . . . . . . Summer 1970 . . . . . . . . . . . . . . . . Summer 1971 . . . . . . . . . . . . . . . . DATA 0 O O 9 O O O O O O O O O O O O O O O O O 0 RESULTS AND INTERPRETATION. . . . . . . . . . . . . . Inheritance of Inflorescence Type. . . . . . . . Single Flower per Truss x Low-flower-number Inflorescence. . . . . . . . . . . . . Low—flower-number inflorescence x Simple Inflorescence. . . . . . . . . . . . . Single Flower per Truss x Simple Inflores- cence. . . . . . . . . . . . . . . . . Single Flower per Truss x Compound Inflor- escence. . . . . . . . . . . . . . . . Inheritance of single flower per truss Inheritance of compound inflorescence. Compound Inflorescence x Simple Inflores- cence. . . . . . . . . . . . . . . . . Inheritance of Flower Number on the Unbranched Monochasial Inflorescence . . . . . . . . . Single Flower per Truss x Low—flower-number Inflorescence. . . . . . . . . . . . . Single Flower per Truss x Compound Inflor- escence. . . . . . . . . . . . . . . . iv Page U1 1‘ ll 15 16 16 23 23 23 27 32 33 42 46 50 51 51 56 TABLE OF CONTENTS-~Continued Page Inheritance of Compound Monochasial Inflores- cence. . . . . . . . . . . . . . . . . . . 62 Single Flower per Truss x Low—flower— number Inflorescence. . . . . . . . . 62 Single Flower per Truss x Compound Inflor- escence . . . . . . . . . . . . . . . 66 Single Flower per Truss x Compound Inflor- escence . . . . . . . . . . . . . . . 68 Jointless Pedicel gs. Number of Flowers per Inflorescence. . . . . . . . . . . . . . . 70 Single Flower per Truss (jointed pedicel) x Low—flower—number Inflorescence (jointless pedicel) . . . . . . . . . 70 F3 Populations . . . . . . . . . . . . . . 71 Leafy Indeterminant Inflorescence . . . . . . . 73 Low-flower—number Inflorescence (leafy indeterminant) x Simple Inflorescence 73 Intercalary Inflorescence gs. Jointless Pedicel O O O O O O O O O O O O O O O 74 DISCUSSION 0 0 O O O O O O 0 O O O O 0 O O O O O O O 77 Single Flower per Truss. . . . . . . . . . 77 Low Flower Number per Inflorescence. . . . 78 Compound Inflorescence . . . . . . . . . . 78 Flower Number on the Unbranched Mono- chasial Inflorescence . . . . . . . . 80 Compound Monochasial Inflorescence . . . . 80 Intercalary Inflorescence. . . . . . . . . 81 SUMMARY AND CONCLUSIONS. . . . . . . . . . . . . . . 82 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . 85 LIST OF TABLES TABLE 1. 10. Distribution for inflorescence types in the dif- ferent generations from the cross MSU 100 (P1) X PennOrange (P2) 0 9 o o o o o o o o o o o o o o Chi—square test for goodness of fit to a digenic inheritance for inflorescence type in the cross MSU 100 (P1) x Pennorange (P2) . . . . . . . . . Frequency distribution for number of flowers on the inflorescence for different generations of tomato plants from reciprocal crosses of Penn- orange (P1) x Michigan State Forcing (P2). . . . Chi-square test for goodness of fit to monogenic inheritance for inflorescence type in the cross Pennorange x Michigan State Forcing. . . . . . . Chi-square test for goodness of fit to a mono- genic_inheritance for.single.flower.perutruss. and simple inflorescence in the cross MSU 100 x MSU 180. . . . . . . . . . . . . . . . . . . . . Distribution for inflorescence types in the dif— ferent generations from the cross MSU 100 (P1) x Apsory (P2). . . . . . . . . . . . . . . . . . . Segregation in the F3 generation from F2 selec- tions with the single flower per truss phenotype derived from the cross MSU 100 x Apsory. . . . . Distribution for inflorescence types in the dif— ferent generations from the cross MSU 100 (P2) x MSU 200 (P1) 0 o o o o o o o o 0 °. 0 o o o o o 0 Comparison of observed single flower per truss and compound inflorescence frequencies in the F2 and backcross pOpulations for the crosses MSU 100 x Apsory and MSU 100 x MSU 200 . . . . . . . Segregation of F3 pOpulations from F2 selections of compound inflorescence individuals derived from the cross MSU 100 x MSU 200 . . . . . . . . vi Page 24 26 28 31 32 34 38 39 41 LIST OF TABLES——Continued TABLE 11. 12. 13. 14. 15. 16. 17. 18. 19. PrOposed model for single flower per truss (a) inheritance as affected by I, single flower inhibitor and R, restorer of single flower expression. . . . . . . . . . . . . . . . . . . Segregation of the F2 population of MSU 100 x Apsory for 4 phenotypic classes of inflores- cence types, and dwarf plant habit (g). . . . . Recombination values for intense branching and non—terminal flowering (ntf) as components of ,compound inflorescence.from backcross pOpula- tions of MSU 100 x Apsory and MSU 100 x MSU 200 O 0 O o O 0 O 0 O O O O O O G 6 O O O O O 0 Comparison of observed compound inflorescence frequencies and expected frequencies based on epistatis of single flower character and link- age between intense branching and non-terminal flowering for the crosses MSU 100 x Apsory and MSU 100 x MSU ZOO . . . . . . . . . . . . . . . Distribution of number of flowers on the un- branched.monochasial inflorescence in different generations from the cross MSU 100 (P1) x Penn— orange (P2) . . . . . . . . . . . . . . . . . . Chi-square test for goodness of fit to a no— dominance model controlling the number of flowers on unbranched monochasial inflores- cences for backcross pOpulations from the cross MSU 100 (P1) x Pennorange (P2). . . . . . . . . ~Distribution for flower number on unbranched monochasial inflorescence in segregating pOpu- lations from the cross MSU 100 (P2) x MSU 200 (P1) Calculated percentage values of F1 (dominant) phenotype expressed for each flower number .class and the cumulative mean for the cross MSU 100 x MSU 200 O O O O O O O O O O O O O 0 O O 0 Frequency distributions of individuals segre- gating for number of flowers per inflorescence in pOpulations of one-branched and two-branched compound monochasial inflorescences from MSU 100 (P2) x MSU 200 (P1) . . . . . . . . . . . . vii Page 44 47 49 50 52 55 57 58 6O LIST OF TABLES--C0ntinued TABLE 20. 21. 22. 23. 24. 25. 26. Chi-square test for goodness of fit to a digenic model for number of flowers on the inflorescence in l branch and 2 branched populations from the cross MSU 100 x MSU 200. . . . . . . . . ... . . Distribution of mean flower number per primary monochasium, and frequency (percent) of compound monochasia in populations from the cross MSU lOO (P1) x Pennorange (P2) . . . . . . . . . . . . . Frequency distribution of number of branches per compound monochasial inflorescence and the fre- quency of compound monochasia in the simple in-. florescence segregates for different generations from the cross MSU 100 (P1) x MSU 200 (P1) . . . Frequency distribution of number of branches per compound monochasial inflorescence and the fre- quency of compound monochasia in the simple inflorescence segregates for different genera- tions from the cross MSU 100 (P1) x Apsory (P1). Segregation of F3 pOpulations for flower number and jointless pedicel from jointed pedical (g),. single flower F2 selections from the cross MSU 100 x Pennorange . . . . . . . . . . . . . . . . Segregation for intercalary inflorescence (ini) in the various pOpulations of the cross Penn— orange (P1) x Michigan State Forcing (P2). . . . Chi-square test for goodness of fit to a mono- genic inheritance for intercalary inflorescence for.the F1, F2 and backcross populations from the cross Pennorange (P1) x Michigan State Forcing (P2) . . . . . . . . . . . . . . . . . . viii Page 61 63 66 69 71 74 75 LIST OF FIGURES FIGURE Page 1. Representative types of inflorescence structure making up the compound and simple inflorescence in the tomato. . . . . . . . . . . . . . . . . . 3 The character single flower per truss as it appears on MSU 100 . . . . . . . . . . . . . . . 11 The compound inflorescence phenotype exhibited by a) Apsory and b) MSU 200. . . . . . . . . . . 13 Phenotype of intercalary inflorescence . . . . . 21 Phenotype illustrating the characters a) intense branching and b) non—terminal flowering. . . . . 37 ix INTRODUCTION An inflorescence in this study is considered as a branch or system of branches bearing flowers. Simple inflorescence (g) is observed in most cultivated tomato varieties and displays a wide range of expression. Varia- tion in the number of flowers per unbranched monochasium (Figure lb) and individuals with compound monochasial inflorescences (Figure 1c) were observed in the simple inflorescence class. This study is concerned with the inheritance of variations of unbranched and compound mono— chasial inflorescences as well as with the inheritance of mutants that affect specific inflorescence types (single flower per truss and compound inflorescence). Compound inflorescence (g) exhibits the inflorescence structure of the compound dichasium (Figure 1a). This inflorescence type is associated with intense branching of the inflorescence and high flower number. A study of low flower number per inflorescence is important because of its association with the jointless pedicel character (1). Jointless pedicel has potential value in that a variety possessing this character would have less fruit injury during mechanical harvest. Since fruits bearing .Aomv waumm kumm GOHHMOAMHmMMHU .EDAMMQUOGOE Unsomeoo A0 .Esammnoocoe UOQOGMHQGD An "mdmeflm .Esflmmnoflp Unsomeoo hm "manomeou .oymEou on“ CH mmocwommuoamcH OHmEflm can vasomeoo may as maflxme musuosupm Oocmommuoawcfl wo mommg O>HDmuammmHmmm .H mudmflm H musmfim the jointless pedicel separate from the plant at the surface of the fruit, the possibility of puncture from another fruit's pedicel is eliminated. It was earlier suggested that the jointless character is associated with low flower number per inflorescence and with leafy inflorescence (If). A complete understanding of these associations is desirable in order to determine the value of the jointless pedicel in a breeding program. Numerous investigations have been made on increasing the number of flowers on the first cluster of simple in- florescence bearing tomato varieties by manipulating the environment during early stages of plant growth. Knowledge on the inheritance of flower number and of environmental effects on flower number could increase the yield of future varieties. The investigation of compound inflorescence and the single flower per truss character have little immediate horti- cultural significance. HOwever, studies on the heritability of these characters aid in the understanding of the total genetic make-up of the tomato and may be of future use in breeding. REVIEW OF LITERATURE Inheritance studies on simple and compound inflorescence were first conducted by Crain (4) in 1915. Compound inflor- escence was suggested to be controlled by a single recessive gene (g). F3 segregates in subsequent years showed a devi- ation from the original single gene hypothesis. Segregations from simple and compound inflorescence conducted by Mac Arthur (12) showed a good fit to the single gene model. According to Crain (4), the compound inflorescence character was made up of several distinct types. Butler (3) and Mac Arthur (13) described this character as having appoxi- mately 80 flowers per inflorescence however; later reports by YOung and Mac Arthur (26) and Lewis (11) describe the character as having up to 300 flowers. The classification of inflorescence types suggested by Parkin (20) allows considerable latitude for the expres- sion of both simple and compound inflorescences. The basic structure of the compound inflbrescence is the compound dischasium (Figure 1a). Deviations in the intensity of branching or flower development can result in compound inflorescences with varying flower numbers. The number of flowers per monochasium, and the number of branches on a compound monochasium allow for a wide range of flower number per inflorescence and structural variation within the simple inflorescence. Lewis (11) compared several tomato varieties and observed distinct classes of flower number within unbranched monochasial inflorescences. The presence of compound monochasial inflorescences were also noted. Since the initial studies of simple inflorescence and the compound inflorescence (4,12), other mutants affecting the intensity of branching and inflorescence structure have been reported. Mertens (16) reported on the bifurcate inflorescence character (bi). This character was described as an extremely bifurcate inflorescence but not as greatly bifurcate as the compound inflorescence. The character was thought to be controlled by a single recessive gene. Phenotypic expression of this gene is similar to the branch— ing of compound monochasium as reported by Lewis (11). Both characters are incompletely expressed, but the bi gene yields more bifurcations per inflorescence. The gene cauliflower (93) was reported by Paddock and Alexander (19) to give a multibifurcate inflorescence similar to the compound inflorescence. They differed in that the cauliflower gene failed.to produce mature flowers. An abortive flower cluster mutant was reported by Azzam (1). This mutant, although not fully investigated, was proposed to be controlled by a single recessive gene. A mutant described by Young and Mac Arthur (26) gives a single flower per inflorescence. This character was sug— gested to be closely associated to the macrocalyx (Egg character. A similar phenotype was observed by Myers (18) in the jointless pedicel tomato variety Pennred. A single flower character was observed in a radiated population of Alisa Craig. This mutant was described as a yellow topped plant with a single flower per truss and the flower appeared to be sterile (26). Similar single flower mutants were described by YOung and Mac Arthur (26) and Myers (18), the single flower mutant obtained for this study (25) shows a strong association to the macrocalyx character. Fehleisen (7) described a mutant that provided a single fasciated flower. This recessive character was designated as uniflora (2;). The jointless pedicel (1) described by Butler (2) has been reported to be associated with 1—4 flowers per inflor- escence. Studies by Rick and Sawant (23) and Emery and Munger (6) suggested that the phenotypes low flower number and jointless pedicel were closely associated. The jointless pedicel was previously reported to be associated with other morphological characters in the tomato. The relationship of jointless pedicel to leafy inflorescence has been discussed by Mertens (16), Rick and Sewant (23) and Emery and Munger (6). All of the studies conclude that the jointless pedicel character is closely linked to the leafy inflorescence character or they are pleiotropic. A strong relationship between the number of nodes to the first inflorescence and days to the first inflorescence was reported by Honma, Wittwer and Phatak (8). Elsayed and Foskett (5) reported on the association between jointless pedicel and the number of nodes between inflorescence and the masking of the sp locus by the action of the jointless pedicel character. The close association of the jointless pedicel character to low flower number, leafy inflorescence, number of nodes between clusters and the masking of the sp locus are sug- gested by Emery and Munger (6) to be due to a morphogenic relationship of these characters to the jointless phenotype. MATERIALS AND METHODS PARENT MATERIAL The tomato line MSU 100, a mutant expressing a single flower per truss (Figure 2) was received from Dr. K. VerKerk, Department of Horticulture of the Agricultural University of Wageningen, The Netherlands. This mutant shows similarity to the single flower character illustrated by Young and Mac Arthur (26). MSU 100 was selfed for seven generations prior to being used in this study. The variety Pennorange was used as the low flower number parent. This variety exhibits 1—4 flowers per inflorescence and jointless pedicel. Apsory, a tomato variety received from Dr. M. Jordanov, Maritsa Institute for Vegetable Crops, Plovdiv, Bulgaria, was used as a compound inflorescence parent (Figure 3a). This variety also displays the dwarf plant habit (g). A breeding line of Lycopersicon pimpinellifolium Mill. (MSU 200) was also used as a compound inflorescence parent (Figure 3b). Two cultivars were used as simple inflorescence parents. MSU 180 with 6—18 flowers per inflorescence and Michigan State Forcing with 3—8 flowers per inflorescence. The range 6-18 10 Figure 2. The character single flower per truss as it appears on MSU 100. ll 12 Figure 3. The compound inflorescence phenotype exhibited by a) Apsory and b) MSU 200. Figure 3 14 flowers obserVed on MSU 180 was due to a high frequency of compound monochasial inflorescences. -Michigan State Forcing shows a low frequency of compound monochasial inflorescences. HYBRIDIZATION All parental material was grown in the greenhouse and evaluated for homozygosity, prior to hybridization. Individual plants were used in making each cross. All crosses and selfs were made in the greenhouse. The parents used for hybridization were also selfed for use in this study. Parental plants were maintained by cuttings for making back- crosses. Seed of the parental and segregating populations for the cross Pennorange x Michigan State Forcing were acquired from a previous study by Honma, Wittwer and Phatak (8). The fol— lowing reciprocal crosses were made for this study: MSU 100 x Pennorange MSU 100 x Apsory Pennorange x Michigan State Forcing The following crosses did not include reciprocal crosses: MSU 100 x MSU 200 MSU 100 x MSU 180 Apsory x MSU 180 FIELD TRIAL Seed of the parent, F1, F2, F3 and backcross popula- tions were planted in vermiculite and seedlings were 15 transplanted into flats at cotyledon expansion. When the plants were 5-6 inches tall, they were transplanted in the field. Summer 1969 The parents, F1, F2 and backcross populations for the crosses MSU 100 x Pennorange, MSU 100 x Apsory, and MSU 100 x MSU 200 were grown in the summer of 1969. F2 selections from the following crosses were saved for observation in the F3 generation: a) single flower types from crosses MSU 100 x Pennorange b) single flower, jointless pedicel types from the cross MSU 100 x Pennorange c) compound inflorescence types from crosses MSU 100 x Apsory and MSU 100 x MSU 200 Cuttings were made from the single flower selections and the F3 seed was produced in the greenhouse. Fruits from compound inflorescence individuals were selected in the field and the F3 seed was saved. Summer 1970 The F3 selections made in 1969 and their parents were grown together with the parents, F1 and F2 of the cross MSU 100 (single flower) x MSU 180 (simple inflorescence) in the summer of 1970. This cross was made in order to learn the relationship of the single flower to the simple inflorescence character. Prior to the 1969 season, Pennorange was con- sidered to be a typical simple inflorescence; however, the 16 observed data suggested that Pennorange inflorescence was not representative of the simple inflorescence. Summer 1971 The crosses, Apsory (compound inflorescence) x MSU 180 (simple inflorescence) and Pennorange (1-4 flowers) x Michigan State Forcing (simple inflorescence) were grown in the summer of 1971. The former was grown to investigate the abberant segregation ratios observed for the compound in— florescence phenotypes noted from the crosses MSU 100 (single flower) x Apsory (compound inflorescence) and MSU 100 (single flower) x MSU 200 (compound inflorescence). The cross Pennorange (1—4 flowers) x Michigan State Forcing (simple inflorescence) was observed in order to more fully investi- gate the relationship of the jointless pedicel phenotype to flower production and inflorescence morphology. DATA Data were recorded on an individual plant basis for all observations. Information was recorded on number of flowers and type of inflorescence as it appeared on the first cluster. Classification as to type of inflorescence was not limited to the first inflorescence, i.e., other inflorescences were observed to substantiate the phenotype of the first inflor- escence. In order to test the validity of sampling only the first inflorescence, a random sample of 50 F2 plants from the 17 cross MSU 100 x MSU 200 having phenotypes other than single flower per truss, compound inflorescence or the greater than 34 flowers per inflorescence class, were observed for the 10 consecutive clusters per plant and evaluated for number of flowers and branched inflorescence. The mean number of flowers on the first inflorescence was 12.65 i .361 and did not differ significantly from the 12.96 i .207 mean number of flowers observed on the 500 inflorescences observed on the 50 plants. The frequency of branched inflorescence on the first inflorescence was observed to be .462 and the frequency observed on the 500 inflorescences from the 50 plant sample was .442. The values were compared by the Chi-square test and no significant deviation between frequencies was observed (P = .95-.90) . The above sampling test for the number of flowers per inflorescence per plant suggests that information obtained from the first inflorescence is a valid estimate of pheno- type for the plant. The number of flowers on the first inflorescence were recorded for all plants except those segregates examplifying the phenotype of the compound inflorescence parents (Figure 3a,b). Due to the large number of flowers and the ability to produce flowers indeterminately, the compound inflores— cence types were recorded only as to inflorescence type. The inflorescence types referred to in this study are described by several basic terms. The terminology is after l8 Parkin (20). The following is a list of the terms used and a description of the ianOrescence type they represent: 1) 2) 3) 4) 5) 6) unbranched monochasial inflorescence (unbranched inflorescence) (Figure 1b) - All flowers originate from the primary axis of the inflorescence. compound monochasial inflorescence (branched inflor— escence) (Figure 1c) — The cluster has at least one branch originating from the primary axis of the inflorescence. simple inflorescence — The simple inflorescence is composed of unbranched and compound monochasial in— florescence types. compound dichasial inflbrescence (Figure la) — The inflorescence forks, similar to dichotomous branching, and each axis exhibits branching. compound inflorescence — The inflorescence exhibits the compound dichasial.structure—with'a high degree of branching. Compound dischasial types with a low degree of branching (1-4 branches on each axis) were classified in a high fIOwer number class (greater than 30 flowers or greater than 35 flowers). primary monoChasium - This term is used in relation to estimating the number of flowers per unbranched monochasium. The number of flowers on the branched inflorescence was previously reported to be related to the number of branches 19 and the flower number class of the unbranched monochasium from which the branch originated (11). The unbranched structure from which the branch originates is referred to as the primary monochasium. The number of flowers per primary monochasium is esti- mated from the number of flowers and branches on a compound monochasial inflorescence. Since each branch replaces one flower on the primary monochasium, the total number of flowers (F) plus the number of branches (B) divided by the number of branches plus 1 (1 represents the primary mono- chasial structure) gives the estimated number of flowers for each primary monochasium (EET)' The flower number of the primary monochasial inflorescence is used as an estimate of the unbranched monochasial flower number. This is used to investigate the relationship between inflorescence branch- ing and the number of flowers on an unbranched inflorescence. Compound inflorescences similar to those observed on Apsory and MSU 200 exhibit a non—terminal flowering habit. This character is evidenced by continual production of young flower primordia on the inflorescence. The provisional symbol, Egg, will be used to describe this character. Data were recorded for leafy inflorescences with inde- terminant growth. This character is referred to as inter- calary inflorescence (Figure 4) and is provisionally noted as ini. This information was obtained from the cross Pennorange x Michigan State Forcing. 20 Figure 4. Phenotype of intercalary inflorescence. 21 Figure 4 22 Data were recorded for the segregation of the jointless pedicel character (1) in the crosses MSU 100 x Pennorange and Pennorange x Michigan State Forcing. Data were also taken for the dwarf (g) character observed to segregate in the cross MSU 100 x Apsory. Both the j and g_genes were investigated in order to determine linkage relationships with inflorescence types. Segregating populations where families could be parti- tioned were tested for fit to expected ratios by the Chi— square test. Individual segregating families were tested for homogeneity prior to pooling data for analysis. Population means were compared by use of the "t" test. Where more than two means are compared, Duncan's Multiple Range test was used (24). The scaling test outlined by Mather (l4) and Mather and Jinks (15) was used in the analysis of the cross Penn- orange x Michigan State Forcing in order to test the distribu- tions to an additive-dominance model. »Methods of estimating the number of gene pairs differentiating the parents as described by Mather and Jinks (15) were used as well as methods described by Powers, Lock and Garrett (22) and Powers (21). RESULTS AND INTERPRETATION Inheritance of Inflorescence Type Single Flower per Truss x Low—flower— number Inflorescence The distribution for the parents F1, F2 and backcross pOpulations from the cross MSU 100 (single flower) x Penn- orange (low—flower—number) show three inflorescence types (Table l). The types correspond to the MSU 100 (single flower), Pennorange (low-flower—number) and F1 (simple inflor— escence) phenotypes. Differentiation of the inflorescence types was based on the single flower type showing 1 flower per inflorescence and the low flower number type showing 1—4 flowers per inflorescence. The simple inflorescence type is composed of unbranched and compound monochasial inflorescences. The presence of three inflorescence types, no expression of parental dominance and the frequencies of .181 for the single flower type and .266 for the low—flower-number type in the F2 population suggest a digenic inheritance and the possibility that the parental types are controlled by reces- sive genes. The segregation of two inflorescence types in the backcross populations suggest that the inheritance of the single flower and low-flower-number inflorescence types may not be complex. Segregation in both backcross populations 23 24 Table 1. Distribution for inflorescence types in the dif- ferent generations from the cross MSU 100 (P1) x Pennorange (P2). Number Inflorescence Type of Single Low—flower- Simple Generation Plants flower number inflorescence MSU 100 (P1) 40 40 Pennorange (P2) 40 4 36 (Ple2IF1 27' 3 24 (P2xP1)F1 76 4 72 F1 pooled 103 7 96 (Ple2)F2 290 57 86 147 (P2xP1)F2 284 47 67 170 F2 pooled 574 104 . 153 317 (Ple2)xP1 80 37 5 38 (P2xP1)xP1 129 57 5 67 F1XP1 pooled 209 94 10 105 (Ple2)xP2 121 10 59 52 (P2xP1)xP2 130 ll 62 57 F1XP2 pooled 251 21 121 109 indicates that each parental genotype is homozygous recessive for one of the two pairs of genes concerned, e.g., MSU 100 (gagg) and Pennorange (Aggy). The F2 population was tested for a two gene model of inheritance (Table 2). Chi—square analysis gave a good fit to a 9:4:3 ratio (simple inflorescence type : low-flower— number type : single flower type). The gene for low flower number type was epistatic to the gene for single flower. Chi—square analysis of the backcross populations gave a good fit to a 1:1 ratio of 1 single flower to 1 simple inflores- cence in the backcross to MSU 100 and l low—flower-number to 1 simple inflorescence in the backcross to Pennorange. This suggests the single flower and low flower number inflorescence types are each conditioned by one recessive gene. The expected frequencies for the F2 and backcross to Pennorange were calculated on the basis of overlap values of .100 low-flower—number type in single flower type and .068 simple inflorescence type in the low-flower-number type. The overlap values were derived from the distributions of the MSU 100, Pennorange and F1 pOpulations (Table l). The expected F2 segregation was determined from the adjusted frequencies calculated with the overlap values and a theoretical segregation of .562 : .250 : .188 (simple inflorescence : low—flower—number type : single flower type). The calculations of the adjusted F2 frequencies was accom- plished as follows: 26 .mmmmu Hm3oam mamcwm can HOQEDCIHO3OHMIBOH can mmmhu HOQEDGIHOBOHM I3oa can mHmEflm mo mcwmmmaum>o mnu mo manna mnu co vmumasoamo Tum mosam> vmuommxm .1 oa.uom. OHH.N 1H.Hv m.voa m.eoa .mxm mas vm mom Hm 0» om m.oaa H.¢ma *.mxm om.-om. mom. Iauav men men Hmm Am on om R.oom o.ama m.~ma *.mxm oa.nom. meo.m Amueumv Ram mma «on enm Am m MK Ammummu OHQEHm Hmnfiac HOBOHM madman coaumumnmw owummv |Hm3OHMIBOA mamcfim mo make mocmommuoamGH umnfisz .A~mv mmcmuoccmm x “any ooa sz mnoHo may an Oahu wocmo Immnoamcfl How Oocmuflumscfl oncmmwv m on new no mmmcuoom Mom ummu OHMDOMIASU .m magma 27 single flower type = .188 + (.100 x .250) = .213 simple inflorescence type = .562 - (068 x .562) = .524 low flower number type = .250 + (.068 x .562) ~ (.100 x .250) = .263 Low-flower—number inflorescence x Simple Inflorescence x The inheritance of inflorescence type in the cross Pennorange (low—flower-number type) x Michigan State Forcing (simple inflorescence) was investigated. Differentiation of parental inflorescence types could not be established from data on the number of flowers per inflorescence in the segregating generations(Table 3). Therefore, the data were analyzed quantitatively. Expected F2 and backcross generation means were calcu- lated after the formulae described by Mather and Jinks (15): F2 = §B1 + fB2, B1 = §Pl + TF1 and B2 = §P2 + §F1 (B1 is the mean of the backcross to Pennorange and B2 is the mean of backcross to Michigan State Forcing). The calculated and observed mean flower number values were 4.41 and 4.35 t .152 for the F2, 3.65 and 3.37 i .142 for the backcross to Penn— orange and 4.90 and 5.44 t .206 for the backcross to Michigan State Forcing. The predicted relationships between means and the assumption that the predicted means are based only on the additive and dominance effects of genes were tested by the scaling tests described by Mather (l4) and Mather and Jinks (15). 28 OON. A Oe.m H m H H N O mm Om OH O m Om OmHoom mm 0» om Omm. u OO.m m H H A mH ON NH O OO mefime~mv OOO. A Ov.m H H H N O OH O O H mm AnxAAEmev NOH. A em.m H m A Hm mm mm mm HH ONH OmHoom Hm on om OOH. A me.m H m NH O OH OH m Om melHENAEO mom. A Hm.m H H N O RH m mH O HO mennmmev «OH. A mm.O H H O O OH ON ON ON OH O OmH OmHoom NE NON. A mH.O H H O m HH Om mH O O OO mmlexwmv mmm. A mm.¢ H m O OH OH O mH O m no mmflmmmev OHH. A OR.O N O mm mm Om OH m mNH OmHoom .m OmH. A mm.v H m RH Hm OH O m HR Hmnmewmv HRH. A ms.O H H «H OH HH A m mm HEANEmeO «OH. A NO.m m O OH mH O O me Awmv.m.m.2 mOH. A mm.~ m mm RH O we Inge mmcmnoccmm .m.m A Ommz OH OH «H NH HH OH O O a O m O m m H mucmHm coHnmumcmO mum3on mo HOQEDZ «0 .oz .Ammv mcHouom mumpm cmmHgOHz x AHmV Omcmuoccmm mo mommouo HMOOHQHOOH scum mpcmam oumsou mo ncoflumuocmm use Tummmwp How mocmonwuoamcfi esp no mum3oam mo Mensa: How cowuanwuumwc hocmsvoum .m manna 29 The scaling test is based on the following formulae: A = 2B1 —'P1 - F1 VA = 4V];-1 + V51 + VF; B = 232 — 52 - E1 and vB = 4V§2 + v52 + VEI c = 4E2 - 2E1 - El - 32 vC = 16V§2 + 4v-E-.-1 + V51 + V52 If the values of A, B and C do not deviate significantly from zero, then the additive—dominance model is suggested to be adequate. The values A_= —.57 i .323, B = 1.07 i .471 and C = .23 x .612 were calculated for cross Pennorange x Michigan State Forcing. They do not differ significantly from zero, suggesting that the additive-dominance model is adequate for analysis of variation. The mean number of flowers per inflorescence 4.79 i .116 (F1) and 5.44 i .206 backcross to Michigan State Forcing) did not differ from the Michigan State Forcing parental mean of 5.02 i .192. This indicates that the Michigan State Forcing phenotype (simple inflorescence) was dominant to the Pennorange phenotype (low-flower—number). The degree of dominance was investigated by analysis of variances according to Mather's formula (14): v = %D + in + E1 and vB1 + v = 2D + TH + 2E1 F2 32 where B1 is backcross to Pennorange and B2 is backcross to Michigan State Forcing. The non-heritable variation (E1) was estimated as the mean variance of the Pennorange, Michigan State Forcing and F1 populations. 30 The F2 and backcross generation variances and-non- heritable variance for these generations from the cross Pennorange x Michigan State Forcing are as follows: E1 = 1.27 VB1 = 2.42 VB2 = 4.16 VF2 = 3.16. The high value of V was due to the presence of compound Bz monochesial inflorescences which exhibit flower numbers greater than 8 flowers per inflorescence. By inserting these vaers into the above formulae, values of H (8.58) and D (-.506) were obtained. The negative value of D suggests zero variation due to additive effects and the 8.58 value of H suggests complete genotypic dominance. The number of effective factors was estimated by the _ — 2 *(P2 g P1) ; where ‘P2 is the mean of Michigan State Forcing and P1 is the mean Mather and Jinks (15) formula K1 = of Pennorange. A K1 value of .18 was calculated (i (5.028:58.52L2 = .18) suggesting monogenic inheritance. The monogenic hypothesis is supported by the test described by Powers gt a1. (22) for gene pair estimation. The formulae T'Pl + %'P2 for l—gene pair and-{g'Pl + +%'P2 for 2-gene pair, with P1 recessive (Pennorange) and P2 as the dominant parent (Michigan State Forcing), were used to estimate the expected number of flowers per inflorescence in the F2 popu- lation. Values of 4.40 for a l—gene model and 4.86 for a 31 2-gene model were calculated. The observed F2 mean of 4.35 t .152 was similar to the 1-gene value, indicating monogenic inheritance. The arithmetic mean of the two parents (3.77) suggests separation of the flower number distribution between the 3 and 4 flowers per inflorescence classes for Chi-square analy— sis, the 1-3 and 4-15 flower classes, corresponding to the low flower number type and simple inflorescence type respec- tively. They showed a good fit to a monogenic 3:1 ratio of simple inflorescence to low flower number in the F2 and a 1:1 ratio in the backcross to Pennorange (Table 4). Table 4. Chi-square test for goodness of fit to monogenic inheritance for inflorescence type in the cross Pennorange x Michigan State Forcing. Inflorescence Type Low—flower— number Simple 73 P F2 obs. 46 90 .522 .50-.30 exp.* 51.9 (3:1) Backcross to obs. 66 54 3.292 .10—.05 Pennorange exp.* 75.6 (1:1) * Expected values were calculated on the basis of the overlap of the simple and low—flower-number types. 32 Single Flower per Truss x Simple Inflorescence The cross MSU 100 (single flower) x MSU 180 (simple inflorescence) was studied in order to determine the inheri- tance of the inflorescence type, single flower per truss. No reciprocal cross was made. The F2 pOpulation segregated for single flower and simple inflorescence types corresponding to the MSU 100 and MSU 180 population phenotypes respectively. The F1 popula— tion exhibited simple inflorescence similar to that of the MSU 180 parent. The data were tested for goodness of fit to a ratio of 3 simple inflorescence: 1 single flower (Table 5). The observed segregation did not show a significant deviation from the 3:1 ratio, suggesting single gene inheritance with the single flower per truss character being recessive. Table 5. Chi—square test for goodness of fit to a monogenic inheritance for single flower per truss and simple inflorescence in the cross MSU 100 x MSU 180. Simple inflores— . Generation N Single flower cence ” P MSU 100 37 37 MSU 180 42 42 F1 62 . 62 F2 138 28 110 exp.(3:l)34.5 103.5 1.633 .3o—.2o 33 Single Flower per Truss x Compound W The cross MSU 100 (single flower) x Apsory (compound inflorescence) was made to investigate the inheritance of single flower per truss, compound (g) and simple inflores- cence types. The distribution of inflorescence types for the parents, F1, F2 and backcross populations are presented in Table 6. The F2 segregation for inflorescence types sug~ gests four classes--single flower per truss, compound, simple, and greater than 30 flowers. The simple inflorescence class includes both unbranched and compound monochasial inflorescences (Figure lb,c), which account for the wide variability in number of flowers on the inflorescences observed (3-30 flowers per inflorescence). Compound inflorescences (s) exhibit the compound dichasial inflorescence structure (Figure la). Recombinants that ex- hibit the compound dichasium, intense branching and non- terminal flowering (25;) is defined as the capability to continually produce new flowers on an inflorescence. This character is detected by the presence of new flowers and the production of new flower primordia on old inflorescences. Recombinants with a large number of branches on a com- pound monochasium were observed. They were noted in the class that had greater than 30 flowers per inflorescence. These highly branched segregates exhibited as many as 80 flowers per inflorescence and are easily discerned from compound inflorescence types. Phenotypes with up to 80 flowers 34 mm O «O HO OmHoom ”Exam m m OH mm meAmemmv ON N Om Om NNxANEmeV Hm Om Om OmHoon meflm mm mH me melnmxwmv mm OH Rm meANEmeO ma Hm «Om om OOO OmHoon mm Hm O we mm mHH Nmndmxmmv Om RH ANN mm mom «EAmeHmV mm mm OmHoon .m RH RH Hennmxwmv OH OH Hmnmmxnmv OO OO Awmv suomnm OO OO name OOH Om: mocmommuoamcfl mum3oam mocmommuoamnfl HOBOHM mucmam aowumnwcmw UGDOQEOU om cmnu THQEHm mamnfim mo Hmwmmuo . HOQESZ meme mocmommuoamnH .1~mv muomnm x AHmO OOH nmz mmouo may Eoum mcoflumumcmm DCOHMMMHU map CH momma mocmommuoamcfl How coausnwuumfln .o manna 35 have been previously described as compound types (3,13). Individuals with this phenotype were observed in the F2 and backcross to Apsory populations. The excessive branching may be due to factors affecting the degree of branching, intense branching (Figure 5a), and non-terminal flowering (Figure 5b). The low frequency of segregates with single flower per truss .188 in the F2 and .363 in the backcross to MSU 100 suggests that the inheritance of single flower was not simple. Two F3 lines from eight F2 selections with the single flower phenotype separated and each gave a good fit to a 15:1 ratio of single flower per truss to simple inflorescence (Table 7). A low frequency (.156) of segregates with the compound inflorescence (g) phenotype was observed in the F2 popula- tion. Three-compound inflorescence F2 lines were selected for observation in the F3 generation. One line segregated and showed a good fit to a 3:1 ratio of compound inflorescence to single flower (P = .70-.50), suggesting that a recessive gene conditions the character single flower per truss. The single flower character appears to be epistatic to compound inflorescence (s). This epistatic activity may in part explain the deficiency of gs recombinants in the F2 pOpula- tion. Compound inflorescence was expected to segregate with a frequency of .25 in the F2 pOpulation. Based on the epistatic relationship between the genes~for single flower and compound 36 Figure 5. Phenotypes illustrating the characters a) intense branching and b) non-terminal flowering. 37 Figure 5 38 Table 7. Segregation in the F3 generation from F2 selections with the single flower per truss phenotype derived from the cross MSU 100 x Apsory. Simple Line N Single flower Inflorescence X2 P 69-22 67 67 69-23 71 71 69-24 65 65 69-25 89 89 69-26 73 73 69-27 69 69 69-21 74 69 5 exp.(15:1) 69.4 4.6 .036 .90-.7o 69-36 70 66 4 exp.(15:l) 65.6 4.4 .039 .90-.70 inflorescence, the expected frequency in the F2 is .1875. The observed value of .156 representing 75 individuals in a population of 480 did not differ from the expected 90 indi— viduals (P = .10-.05). The expected frequency of compound types in the back— cross to Apsory is .5. The observed frequency of .433 which represents 33 compound types in a population of 81 individuals did not deviate significantly from the expected (P = .10—.05). Similar results to those observed in the cross MSU 100 x Apsory were obtained in the cross MSU 100 (single flower) x MSU 200 (compound inflorescence). 39 The distribution of inflorescence types for the parents F1, F2 and backcross populations are presented in Table 8. Four classes of inflorescenCe types were Observed in the F2 (single flower, simple, compound and greater than 34 flowers). The backcross to MSU 100 segregated for simple inflorescence and single flower and the backcross to MSU 200 segregated for greater than 34 flowers, simple and compound inflorescence types. The simple inflorescence types include both unbranched and compound monochasial inflorescences. Table 8. Distribution for inflorescence types in the differ- ent generations from the cross MSU 100 (P2) x MSU 200 (P1). Inflorescence Type Number Simple Greater Compound of Single inflor- than 34 inflor— Generation plants flower escence flowers escence MSU 200 (P1) 40 4O MSU 100 (P2) 40 40 F1 34 34 F2 499 87 326 16 70 BC to P1 84 44 6 34 BC to P2 38 13 25 The segregation for four inflorescence classes in the (MSU 100 x MSU 200) F2 (Table 8 ) and (MSU 100 x Apsory) F2 (Table: 6) were tested for association by Chi-square analysis 40 and no difference in segregation was suggested (P = .70-.50). Similarity in segregation of the backcross to the single flower parent (P = .90-.80) and the backcross to the compound inflorescence parent (P = .70-.50) were also observed. Frequencies for single flower per truss and compound inflor- escence types are presented in Table 9. Table 9. Comparison of observed single flower per truss and compound inflorescence frequencies in the F2 and backcross pOpulations for the crosses MSU 100 x Apsory and MSU 100 x MSU 200. Single Compound flower inflorescence (MSU 100 x Apsory) F2 .188 .156 (MSU 100 x MSU 200) F2 .174 .140 (MSU 100 x Apsory) x MSU 100 .363 (MSU 100 x MSU 200) x MSU 100 .352 (MSU 100 x Apsory) x Apsory .433 (MSU 100 x MSU 200) x MSU 200 .405 The frequencies of single flower type .174 in the F2 and .352 in the backcross to MSU 100 suggest again (as in the MSU 100 x Apsory cross), that the inheritance for single flower is not simple. While F3 lines from single flower (MSU 100 x Apsory) F2 selections segregated for simple inflorescence, none of the 7 F3 lines from single flower (MSU 100 x MSU 200) F2 selections segregated. 41 The frequency of compound inflorescence recombinants, .140 in the (MSU 100 x MSU 200) F2 pOpulation, was lower than expected for a single recessive gene (g) inheritance. F3 lines from compound inflorescence F2 selections exhibited segregation for single flower types in l of 4 lines grown (Table 10). Line 69-37 gave a good fit to a 3:1 ratio of compound inflorescence to single flower. The single flower gene is suggested to be epistatic to the gene for compound inflorescence. Table 10. Segregation of F3 populations from F2 selections of compound inflorescence individuals derived from the cross MSU 100 x MSU 200. Inflorescence Type Line N Compound Single Flower X? P 69-38 68 68 69-39 75 75 69-40 73 73 69-37 66 54 12 exp.(3:l) 49.5 16.5 1.636 .30-.20 The segregation of compound inflorescence (g) progeny observed in the (MSU 100 x MSU 200) F2 pOpulation was tested to an expected segregation frequency of .1875. .The expected value is based on epistasis between single flower per truss and compound inflorescence. The observed frequency .140 42 representing 70 compound types in 499 F2 individuals was tested to the expected frequency of .1875 representing 93.5 compound types in 499 F2 individuals. The observed compound inflorescence frequency showed a poor fit to the expected frequency (P = less than .01), suggesting the two gene, epistatic hypothesis to be inadequate. Compound inflorescence types are expected to occur with a frequency of .5 in the backcross to MSU 200. The observed frequency of .405 which represents 34 compound types in a population of 84 individuals does not differ from the ex— pected .5 or 42 compound types in a population of 84 (P = .10—.05). Inheritance of single flower per truss. A hypothetical model for the inheritance of the character single flower per truss is based on the following observations: 1) Low frequencies of single flower types were observed in (MSU 100 x MSU 200) F2 (.174), (MSU 100 x Apsory) F2 (.188), (MSU 100 x MSU 200) x MSU 100 (.352) and (MSU 100 x Apsory) x MSU 100 (.363) (Table 9). 2) F3 generations from compound inflorescence (MSU 100 x MSU 200) F2 and (MSU 100 x Apsory) F2 selections segregated 3:1 (compound yg. single flower), sug- gesting a recessive gene controlling the character single flower and that gene to be epistatic to the gene for compound inflorescence. 43 3) Segregation for simple inflorescence types in the F3 generations of single flower types from (MSU 100 x Apsory) F2 selections. The frequency of single flower per truss segregates ob- served in the backcross (MSU 100 x Apsory) x MSU 100 gave a poor fit (P = .05—.Ol) to a 1:1 segregation; while the back- cross (MSU 100 x MSU 200) x MSU 100 showed no significant deviation from an expected 1:1 ratio (P = .10—.05). Since results from the previous crosses, MSU 100 x Penn— orange and MSU 100 x MSU 180, suggested the character single flower per truss was controlled by a single recessive gene, the low frequency of single flower types in single flower x compound inflorescence crosses probably is a result of factors transmitted by the compound inflorescence parents. The repres— sive effect on single flower expression in the backcross to MSU 100 suggests the factor or factors to be dominant. The 3:1 segregation of compound gs, single flower for the F3 lines from the F2 compound inflorescence selections suggests no inhibitor of single flower expression. Yet, the 15:1 segregation for single yg, simple inflorescence in F3 lines from single flower (MSU 100 x Apsory) F2 selections suggest segregation of inhibition factors. A model is pr0posed to explain the observed single flower segregations (Table 11) for the single flower x com— pound inflorescence crosses. The model is based on three major genes-—§ for single flower per truss, ;_- inhibitor of 44 Table 11. Proposed model for single flower per truss (a) inheritance as affected by I, single flower inhibitor and g, restorer of single flower expression. MSU 100 x Apsory or MSU 200 eeiiy; AAIIE. E. .43 II & Female gametes and frequency BC to MSU 100 a i_£ a.i R a 1,; a_l 3.... male .125 .125 .125 .125 .... gamete a‘i g single single single Female gametes and frequency F2 eir 3.1.5 elr ELEM” male .125 .125 .125 .125 .... gametes and ' frequency a i_£ .125 single“single single _ _.§ .125 single single single single _‘I g .125 single single _.l.B .125 single single single single 45 single flower expression and R.- restorer of single flower expression. ;_inhibits the expression of g; in the absence of R. In the presence of R, the single flower character segregates as a single gene. The genotype of MSU 100 is §_i,£_and the genotype of compound inflorescence lines is A_I_§. The calculated frequencies of single flower types, based on the model (Table 11), in the backcross and F2 pOpu- lations are .375 and .203 respectively. These values were used to calculate expected segregations for single flower in segregating generations from the crosses MSU 100 x MSU 200 and MSU 100 x Apsory. No significant deviations were ob- served in the (MSU 100 x MSU 200) F2 (P = .20-.10), (MSU 100 x MSU 200) x MSU 100 (P = .80—.70), (MSU 100 x Apsory) F2 (P = .50-.30) and (MSU 100 x Apsory) x MSU 100 (P = .90-.80) suggesting a good fit to the proposed model. From this model, the folloWing would be expected in the F3 generations of single flower lines selected from F2 popu- lations of either MSU 100 x MSU 200 or MSU 100 x Apsory crosses: 4 with 13:3 ratios; 2 with 3:1 ratios and 10 with no segregation3 however, the 2 (MSU 100 x Apsory) F3 lines that showed segregation, each segregated with a 15:1 ratio of single flower to simple inflorescence. This suggests that other factors may be involved in the expression of the single flower character. 46 Inheritance of compound inflorescence. The inheritance of the compound inflorescence (g) as noted in the crosses MSU 100 x MSU 200 and MSU 100 x Apsory was quite similar to that previously reported (12). The apparent epistasis be- tween the single flower gene and compound inflorescence gene may account for the observed deficiency of compound inflor- escence types in these F2 populations. The epistasis however, did not compensate for the deficiency of the gs types in the (MSU 100 x MSU 200) F2 pOpulation. The (MSU 100 x Apsory) F2, backcross to Apsory and backcross to MSU 200 did not deviate significantly from expected segregation at the 5% level; however, all these populations showed P values of .10-.05. The low P values were due to low number of compound inflorescence types. Previous investigations concerning the compound inflor- escence reported on deviations in expected segregation (4), variation in degree of compounding (4), and descriptions of the phenotype ranging from 80 flowers (3,13) to 200-300 flowers (11,26) per compound inflorescence. Based on these observations, the low frequency of compound inflorescence segregates may be the result of a more complex inheritance. The presence of recombinants with high flower number and moderate branching or non-terminal flowering may be con- sidered a form of compound inflorescence. The low frequency of recombinants in the greater than 30 flower class (greater than 34 in the MSU 100 x MSU 200 cross)(hereafter to be 47 referred to as 30 flower class) and the deficiency of com- pound inflorescence types suggest that this class may be associated with compound inflorescence. The possible association of the compound and 30 flower _c1ass was investigated by observing the dwarf gene (g) known to be linked to the s locus (3). The dwarf character was transmitted by the Apsory parent. The segregation for dwarf (g), single flower, compound and simple inflorescence and the 30 flower class are presented in Table 12. Assuming that the 30 flower class is associated with compound inflorescence, the frequency of dwarf types within the compound inflorescence class and the 30 flower class should indicate linkage. There were 48% dwarf, compound inflorescence types, 47.6% dwarf, 30 flower types and 12.3% dwarf simple inflorescence recombinants (Table 12). Table 12. Segregation of the F2 pOpulation of MSU 100 x Apsory for 4 phenotypic classes of inflorescence types, and dwarf plant habit (g). If Number Simple Greater Compound of Single inflor- than 30 inflores— plants flower escence flowers cence Total obs 480 90 294 21 75 Dwarf Obs 102 20 36 10 36 Percent dwarf 22.2 .12.3 47.6 48.0 48 The 22.2% dwarf types in the single flower class exceeded the frequency observed for simple inflorescence types. The in- creased frequency does not suggest linkage of single flower to dwarf but probably is a result of the epistatic relation— ship of single flower and compound inflorescence. The high frequency of combinations of dwarf with compound and dwarf with 30 flower types which was observed in the (MSU 100 x Apsory) F2 suggest that dwarf is probably linked to both compound inflorescence and the 30 flower class. The compound inflorescence phenotypes of the Apsory and MSU 200 parents exhibit very intense branching and non- terminal flowering. The 30 flower class plants shows many branches per inflorescence but few exhibit non—terminal flowering. Based on these observations, the intense branch- ing and non-terminal flowering (23;) are suggested to be linked and compose the compound inflorescence phenotype exhibited by the parents. The intense branching phenotype exhibits high flower number and the compound dichasial struc- ture suggesting that the intense branching gene may be the .g gene. Recombination values between genes for intense branch- ing and non-terminal flowering were estimated from the back— crosses (MSU 100 x MSU 200) MSU 200 and (MSU 100 x Apsory) Apsory. Recombination values 10.89% for (MSU 100 x MSU 200) MSU 200 and 6.17% for (MSU 100 x Apsory) Apsory were calcu- -lated according to the methods outlined by Immer (9) (Table 13). 49 OHO. A O0.0H an O N ON OON ems x IOON Om: x OOH Omzv ONO. H NH.O HO H O ON shaman x Asuomnm x OOH Omzv .m.m A usmoumm ++ WHO + + HMO mmuOHOmN mCHAUCMHQ mcH£UGMHQ OmcmucH omcmpaH .OON Om: x OOH Om: 6cm muomna x OOH am: no mcoHumaomom mmouoxomn scum wocmomouoamcw onoomsoo mo mpcmcomeoo mm Hmucv mcHHm3on HmoHEHmplooo 6cm mcHnoomun mmcmucH How mmsHm> CONDOGHQEOOOM .MH maflme 50 The recombination values were used to estimate the expected segregation frequency of compound inflorescence types in the F2 and backcross populations of MSU 100 x Apsory and MSU 100 x MSU 200 (Table 14). All calculated values were similar to the observed values. Table 14. Comparison of observed compound inflorescence fre— quencies and expected frequencies based on epistatis of single flower character and linkage between intense branching and non-terminal flower- ing for the crosses MSU 100 x Apsory and MSU 100 x MSU 200. Backcross to Compound Parent F2 Cross obs. exp. obs. exp. MSU 100 x Apsory .433 '.469 .156 .165 MSU 100 x MSU 200 .405 .455 .140 .149 Compound Inflorescence x Simple Inflorescence The cross Apsory (compound inflorescence) x MSU 180 (single inflorescence) was made in order to support the pro- posed mode of inheritance of the compound inflorescence character (g) noted in the MSU 100 x Apsory and MSU 100 x MSU 200 crosses. The F2 distribution showed segregation for simple and compound inflorescence. Segregates showing high flower number and many branches or the non-terminal flowering habit were also observed. 51 The F2 segregation 74 simple : 14 compound inflorescences gave a poor fit (P = .05-.02) to an expected 3:1 ratio, and suggested that more than one factor of linkage affected the inheritance of the compound inflorescence. Recombinants with many branches and non-terminal flowering were observed. The intense branching and £5; loci were suggested to be linked and comprise the compound inflorescence phenotype. .This link— age is similar to that proposed in the MSU 100 x Apsory and MSU 100 x MSU 200 crosses. A recombination value for intense branching and gt; of 5.97 i .017 was calculated from the F2 data. This value is similar to that observed in the single flower x compound inflorescence crosses (Table 13). The chi—square analysis suggests the observed compound inflorescence segregation shows a good fit (P = .20-.10) to the expected segregation based on linkage. Inheritance of Flower Number on the Unbranched Monochasial Inflorescence Single Flower per Truss xfigow3flower number Inflorescence In the cross MSU 100 (single flower) x Pennorange (low- flower—number), the simple inflorescence types showed segregation for number of flowers per inflorescence. The segregation was observed in the distribution for number of flowers on the unbranched inflorescence (Table 15). A high proportion of individuals in the 8-9 flower range were noted 52 vmm. H mo.m H H H N h mm mm ¢ N mOH mm 0» 0m OOH. H m~.n m mm mm ma OH m O o MOH Hm on on mmo. H mH.m H H om mm Hm mm mm MH O mmm Nm OMH. H 0H.o m m om mm mm m m Om Hm .m.m H COOS HH OH m m O m m O m momma COHpmHocmw mHOBOHm mo HOQEDZ pmzoomun Ids Hmuoa . 3.: Omcmuoocmm x HHmv OOH sz mmouu may scum moowumumcmm ucmummmwo CH wocmomm IuonoH HOHmmnoocos.6w£ocmnnoo may do mumzoam mo Hogans mo GOHquHHumHQ .mH OHQMB 53 in the backcross to MSU 100 while a high number of indi- viduals were present in the 5-6 flower range in the backcross to Pennorange. The mean number of flowers for the F1 and F2 popula- tions were 6.10 i .127 and 6.19 t .085, respectively. The mean flower number of the backcross to MSU 100 was 7.29 = .177 and the backcross to Pennorange was 5.06 i .234. The observed F1 mean (6.10) and F2 mean (6.19) did not differ. Using the formula described by Mather (14), F2 = 231 + 2B2, where B2 is the mean of the backcross to MSU 100 and B2 is the mean of the backcross to Pennorange, the expected F2 pOpulation mean of 6.17 flowers per unbranches inflorescence was calculated. The formula can be written as O = £31 + §B2 - F2 and the deviation from 0 can be tested: t = dSVIatlon dev where sag; =J/TTVEi + 2V3; + V52 . No Significant difference between the observed and calculated F2 means was observed. The similarity of the F1 and F2 means and the F2 mean being the midpoint between the two backcross means suggest no dominance effect in the inheritance of flower number classes within the unbranched monochasial inflorescence phenotype. A single gene model is prOposed and the gene is symbol- ized as g. The genotype of Pennorange is 92 and MSU 100 is 99. The flower number expressed by the gg and 99 genotypes were estimated by assuming the F1 mean to be the midpoint in a no dominance situation. The flower number values were calculated as follows: 54 gg 5.06 - (6.10 - 5.06) 4.02 flowers .QQ 8.48 flowers 7.29 + (7.29 - 6.10) The phenotype gg is expressed as 4—flower class, gg_as 8—flower class, and Cg as 6-flower class. Assuming that the backcross pOpulations are composed of 2 6-flower class (99) plus % parental type (QQ or £9), the model was tested by comparing the skewed portion of the back- cross pOpulation to the corresponding classes in the F1. The skewed portions, 3, 4 and 5 flowers in the backcross to MSU 100 and 7, 8, 9 and 10 flowers in the backcross to Pennorange (Table 15), were assumed to arise from the effect of the 93 genotype. The expected vaer for the backcross to MSU 100 was calculated as follows: frequency of F1 total 6-flower class expected individ- pOpulation with x (unbranched) in the = uals with 3, 4 and 3, 4 and 5 flowers BC to MSU 100 5 flowers in the 2 BC to MSU 100 34/94 x 103/2 = 18.6 The same formula was used to calculate the expected number of individuals in the 7, 8, 9 and 10 flower class in the backcross to Pennorange. The total number of unbranched inflorescences was estimated on the basis of the overlap of the Pennorange and F1 populations. The observed values show no deviation from the expected (Table 16) indicating a good fit to the no dominance hypothesis for inheritance of flower number on the simple inflorescence. 55 m0.loa. H.0m m.mH .me who.m mm NH mOH Nm ou Om O.OO O.OH .nxm om.|om. mOm. mm ma moa Hm ow mm m NXV .wmmma Hm. .mmmma Mm. mmoomommHOHmGH COHDMHOGOU + Hmpqmumm HOHpumm UOQOCMHQCD mo HOQESZ .Ava mmomuoacmm x HHmV OOH sz mmouo map EOHM maoHumHomom mmouoxomn How moocmommuoamoH HmHmmsoocoE pmsocmuncs no mumBOHm mo Hogans on» maHHHouucoo HOUOE OOGMGHEOUIOG m on “Hm mo hungooom How ummv THOSUMIHQU .ma OHQmB 56 Single Flower per Truss x Compound Inflorescence Simple inflorescence types in the F1, F2 and backcross populations of the cross MSU 100 (single flower) x MSU 200 (compound inflorescence) showed variability with regard to the number of flowers on the unbranched inflorescence (Table 17). Mean flower number values for the F1 and back- cross to MSU 200 were 10.65 i .41 and 10.43 i .47, respec- tively. Means of 8.27 i .66 for the backcross to MSU 100 and 8.65 i .21 for the F2 were Observed. The similarity of the F1 and backcross to MSU 200 means suggest that the high flower number individuals (8-16 flowers) make up the dominant class. The high flower number class is referred to as the lO-flower class. The F2 distribution for number of flower on the un- hranched monochasium does not indicate any distinction between the lO-flower class and the low flower class (3-7 flowers). The low flower number class is described as the 7-flower class. The unbranched populations were analyzed using a formula similar to that prOposed by Powers (21) and illustrated by Leonard, Mann and Powers (10). The formula %§'x 100 was used to estimate the number of genes involved. The frequency in the F2 for a given class is divided by the parental fre— quency for the same class, the result multiplied x 100 for conversion to percent to give an estimate of the prOportion of parent types in the F2 pOpulation. 57 OO. A NN.O H N H N H O N H H H OH NO on OO OO. A O0.0H H H H O O N O O HN .O on OO HN. A OO.O H O NH OH AH ON HN ON O O N N NOH NO HO. A O0.0H H H H O O O OH HO .O.O A cams OH OH OH OH NH HH OH O O R O O O m Omens OOHOOHOOOO mHOBOHm wo HOQEOZ UOQUOOHQ TOO Hmuoe .AHmV oom mm: x Hva 00H sz mmouo TAD Eoum mCOHumHomom mcHummmumOm CH OOCOOOOHOHHCH HOHmmHHoocoE vogue—mung: so 3855 .8393 How GOHDDQHHHOHQ .OH 9395 58 Since MSU 100 and MSU 200 are assumed to be of the recessive phenotypes single flower and compound inflorescence respectively, the F1 and F2 unbranched inflorescence popula- tions were tested by the formula F2/F1 x 100 (Table 18). This formula gives an estimate of the frequency (in percent) of F1 phenotypes (dominant class) in the F2 population. Table 18. Calculated percentage values of F1 (dominant) phenotype expressed for each.flower number class and the cumulative mean for the cross MSU 100 x MSU 200. Calculated Flower Percentage for , Number Frequency Each Class Cumulative Class F2 F1 (F2/F1 x 100) Mean 12-16 .134 .214 62.6 11 .099 .214 46.3 54.5 10 .119 .357 33.3 47.1 9 .162 .214 75.7 54.2 The comparisons of individual classes show some varia— bility (33.3%—75.7%) which may be due to sampling size of the-F1 and F2 pOpulations. The Cumulative mean for the 4 individual comparisons was 54.2%, suggesting that the F2 popu- lation is composed of 54.2% F1 types. This frequency (54.2%) of dominant flower number class lO-flower class suggests digenic inheritance for unbranched inflorescence flower I 59 number class with the dominant class comprising 9/16 (56.2%) of the total pOpulation. The mean of 4 individual values (54.2%) did not differ from the expected 56.2% for a digenic model (P = .90-.70). The digenic inheritance assumes that unbranches inflor- escence classes are independent of the parental phenotypes and that branching can occur with equal frequency in the lO—flower class and 7—flower class. The latter assumption can be noted by observing the segregation of compound mono- chasial inflorescences with one and two branches (Table 19). The relationship between number of flowers per unbranched inflorescence and the number of flowers on one-branches com- pound monochasia has been previously reported (11). Specific flower number classes of unbranched inflorescences give a predictable number of flowers when the inflorescence branches. The low frequency of one—branched inflorescences in the 12-flower class of the (MSU 100 x MSU 200) F2 pOpulation and the 13 flower class being the low flower number limit of the F1 and backcross to MSU 200 distributions (Table 19) (dominant class) suggest the lZ-flower class to be the point of sepa- ration between the two classes. Using this as the point of separation, a segregation ratio of 22 (8-12 flowers-~low flower class) : 36 (13—23 flowers--dominant class) is noted. The observed segregation was not significantly different from an expected 7:9 ratio (Table 20), suggesting the higher flower class to be dominant. 60 o m O O N d N N ON Nm N .m H H H m Hm 0# 0m+Hm moaocmnnlo3a ONAH ONONN HNOON OHOOH OH OH OH OH OH OOOOHO eoHumumcmO ”H03OHM MO Hmpegz M0 0 OZ N H H H O NO on om N N N H O N H OH .O on em H ,O O N O O OH O O O O O OO NO H N N O O H O OH .O noomunlmco MN NNQHN ONémH mafind 0H ma ma NH NH Ha OH m£m mgcmam cOHumanmw mHOBOHw mo HOQEDZ mo .02 .AHOO OON Om: x ANOO OOH 50: scum OOOOOOOOHDHOOH HOHmmnoocoe pcoomsoo wogocmunlozu 6cm Umnocmunloco mo mcoHumHomom CH mocmommuoamcw Ham mu03oam mo HOQEDO How mchmmOHmmm MHOOUH>HUCH mo mcoHquHHumHU hocmovoum, .mH OHQMB 61 Table 20. Chi-square test for goodness of fit to a digenic model for number of flowers on the inflorescence in l branch and 2 branched populations from the cross MSU 100 x MSU 200. Number of Low Flower Dominant Distribution Plants Class Class If P l branch 58 22 36 exp.(9:7)25.4 32.6 .808 .50-.30 2 branches 27 10 17 exp.(9:7)ll.8 15.2 .465 .50-.30 The absence of segregates in the l7-flower class in the two—branched compound monochasial inflorescence of the (MSU 100 x MSU 200) F2 distribution and the l7-flower class being the lowest observed individual in the F1 and backcross to MSU 200 distribution, suggests l7 flowers to be the separa- tion point between the two flower number classes. .A ratio of 10 (13-16 flowers) : 17 (18 and greater flowers per inflorescence) was observed. The 18 and greater number of flowers correspond to the distributions of the F1 and backcross to MSU 200 and is considered the dominant class. Sixty-two and nine-tenths percent of the two-branched inflor— escences are of the dominant type. The observed segregation gives a good fit to the digenic 7:9 ratio of 13—16 flowers to the dominant class (Table 20). The segregation for flower number on both one-branched and two-branched compound monochasial inflorescences show 62 agreement to the digenic inheritance as suggested in the analysis of flower number on the unbranched monochasial inflorescences. Other crosses. No segregation for number of flowers on the unbranched monochasial inflorescence was noted in the crosses Pennorange x Michigan State Forcing and MSU 100 x Apsory. Inheritance of Compound Monochasial Inflorescence Single Flower per Truss x Low-flower- number Inflorescence The frequency of compound monochasial inflorescences- observed in the F1, F2 and backcrosses for the cross MSU 100 (single flower) x Pennorange (low—flower-number) are pre- sented in Table 21. The frequency of branching in the F2 (8.7%) is approximately 2 times that observed in the backcross to Pennorange (3.6%). Assuming that the backcross to Penn- orange is composed of2'F1 frequency + i-P2 frequency, where P2 is Pennorange, the 3.6% of the branched inflorescences in the backcross to Pennorange are prObably due to a factor or factors transmitted by the F1. The degree of branching con- tribUted by the Pennorange parent is suggested to be small. .The compound monochasial inflorescence frequencies of the F2 (11.9%) and the backcross to MSU 100 (10.4%) are greater than that observed in the FL, suggesting a factor or factors for increasing branching frequency may have been transmitted by the MSU 100 parent. .HO>OH Rm way an OOHumH>OO ucmoHMHcmHm OuOOHOGH mumpuOH HGOHOMMHQ .1 63 03 OOO. H m0.0 H m O.m_ OOH NO 0» om Q mOO. H O0.0 H m O.m MOH Hm om mHm. H Om.m NH O OH m.HH ONm Nm m O0.0 H mm.O O m m 0.0H OHH Hm 0» om O.m.m H cmmz OHIO O mum Omsocmum mmocwomOHOHmcH OGOHumumcmm Eonmguocoz MHMEHHO Hon muw30Hm ucmuumm OOQUGMHQGD 6cm Ongocmum mo HmnEoz .HNmV mmcmuocomm x HHmv OOH Om: mmouo map Eouw OGOHOOHOOOO :H OHmmnoocoe vasomeoo mo Hucmoummv mocmsmmum can .EDHOOSUOCOE mumEHHm H0O HOQEDO HO3OHM ones «0 GOHuonHHpmHQ .HN OHQOB 64 Assuming the relationship reported by Lewis (11) between the number of flowers per unbranched monochasium and the frequency of branched inflorescence, the gene regulating flower number class (Q) and the frequency of branched inflor- escence may be considered to be the same. .The 2g genotype would have little effect on branching, the g9 genotype would produce branched inflorescences at a frequency of 8.7% and the §Q_genotype would have an increased branching frequency which was estimated from the F2 and backcross to MSU 100 to be 30.4% and 12.2% respectively. .The estimates were based on monogenic segregation frequencies as follows: ‘backcross to MSU .104 — i-5 x .087) _ 122 .5 ‘ ° F2 = 1‘92 + %_ C_C + '4" E = .119, E = .119-(.25x.(2)g)-(.5x.087)___ 100=%£g+&-$=.104,QQ= .304. The difference between estimated effect of g§_may be due to population size of branched types. Primary monochasial flower number values were used to estimate the unbranched flower number class from which branch- ing occurred. The distribution of primary monochasial flower number (Table 21) exhibited three c1asses—-3-5 flowers, 6 flowers and 7-10 flowers. The mean number of flowers per primary monochasium for the F; (4.77 flowers) and the backcross to Pennorange (4.75 flowers) do not differ significantly and suggest branching t0“ occur predominantly of 3-5 flower individuals. The mean number of flowers for the F2 (5.97 flowers) and the backcross to MSU 100 (6.33 flowers) differed significantly from the F1 mean. The higher mean values indicate that the F2 and 65 backcross to MSU 100 tend to exhibit branching on higher flower number types. If the F1 genotype (Q9) onIy effects the 3-5 and 6 flower classes and the 99 genotype effects only the 6 and 7-10 flower classes, then the effect of 99 on branched inflor— escence frequency can be estimated relative to the 99 (F1) branching frequency. The 6Fflower class segregates are partitioned into either the 3—5 or 7-10 flower classes on the basis of the F1 segregation. The two 6 flower types observed in the backcross to MSU 100 were partitioned with one individual to each class. An adjusted ratio of 4 (3-5 flowers, 9g) : 8 (7-10 flowers, 99) was obtained. With a monogenic segregation of l‘gg : 1' pg, the 9Q effect was 2 times (2%% = 2) the effect of the F1 on branching. An adjusted F2 segregation of 20 (3-5 flowers, 9;) : 19 (7—10 flowers, 99), was obtained. With a monogenic segregation of l g9 : 2 99 ; l gg and 99 giving no effect on branching, the effect of 99 was estimated as 1.9 times (2%L%2:-_%fi= 1.9) the effect of the F1. Two times effect of the F1 gives an estimated 9Q effect on branching of .174, (2 x .087), which approximates the mean of the values esti— mated above from the F2 (.304) and backcross to MSU 100 (.122). The F1 phenotype showed unbranched inflorescences with a range of 3 to 11 flowers. Compound monochasial inflores- cence was suggested to arise from unbranched inflorescences 66 with 3—6 flowers. The absence of branching on individuals with a higher number of flowers (7-10) suggest that the production of branched inflorescences is not based on the number of flowers but may be related to the specific genotype. Single Flower per Truss x Compound Inflorescence The frequency distribution of the number of branches on each of the compound monochasial inflorescences for the cross MSU 100 (single flower) x MSU 200 (compound inflorescence) is shown in Table 22. Individuals with greater than 4 branches per inflorescence were associated with either the compound dichasial structure or non-terminal flowering and a high number of flowers. Table 22. .Frequency distribution of number of branches per compound monochasial inflorescence and the fre- quency of compound monochasia in the simple inflor- escence segregates for different generations from the cross MSU 100 (P2) R MSU 200 (P1). Number of Simple In- florescence Total Generation Types 1 2 3' 4 )4 Frequency F1 34 17 2 l .588 F2 264 58 27 10 4 23 .462 BC to P1 50 13 6 3 2 5 .580 BC to P2 25 5 2 2 .360 67 The frequency of branched inflorescences for the F1 population was observed to be .588. The similarity in the frequency of inflorescence branching in the F1 and backcross to MSU 200 pOpulations suggest that the high frequency was due to a dominant gene. Based on a dominant gene expression giving .588 fre- quency of compound monochasial inflorescences, the frequency of branched inflorescences attributable to the MSU 100 parent is estimated from the backcross to MSU 100. Assuming the backcross to MSU 100 is composed of the frequencies trans- mitted by MSU 100 and the F1 populations with a prOportion i-MSU 100 + 2-F1. The estimated frequency of compound mono- .360-(.5x.58§L) chasial types produced by MSU 100 is .132 (P2: This value is in close agreement with the frequency of branch- ing attributed to the MSU 100 that was estimated from the backcross (MSU 100 x Pennorange) x MSU 100 (.122). Using the dominant expression as 58.8% branched inflores- cences and the recessive as 13.2% branched inflorescences, the expected F2 segregation of branched inflorescences was calculated. The segregation in the F2 population was tested for goodness of fit to an expected monogenic segregation and a good fit was obtained (P = .90—.80). The relationship between compound monochasial inflor- escences and the flower number class of unbranched mono- chasium was investigated. In the F2 and backcross to MSU 100, segregation for number of flowers per unbranched monochasium 68 was noted (Table-l7). Two flower number classes were hypothe- sized--a 7-flower class and lO-flower class (F; type). They segregated with a ratio of 9:7, lO-flower class to 7-flower class. According to Lewis (11), the high flower number class (10 flowers) should exhibit a greater frequency of branched inflorescences than the 7—flower class. The distributions of one-branch and two—branched com- pound monochasial inflorescence types (Table 19) were pre- viously partitioned into two classes. The 8-12 and 13-16 flower classes correspond to 7-flower class individuals, and the 13—23 and 17 and greater flower classes correspond to 10— flower class individuals. The segregation observed in one and two—branched inflorescences did not differ significantly from the digenic segregation observed in the unbranched monochasial pOpulation (Table 20). The similarity of the segregation ratios for the one and twonranched types to the unbranched inflorescence population suggests that the 7-flower class and the lO-flower class have the same capacity to produce branched.inflorescences. Single Flower per Truss x Cgmpound Mrficence The distribution of number of branches per branched inflorescence and the frequency of branching for segregating generations from the cross MSU 100 (single flower) x Apsory (compound inflorescence) are presented in Table 23. 69 Table 23. Frequency distribution of number of branches per compound monochasial inflorescence and the fre— quency of compound monochasia in the simple inflorescence segregates for different genera- tions from the cross MSU 100 (P1) x Apsory (P2). Number of Simple In— Number of Branches Total florescence Greater Fre- Generation Types 1 2 3 4 than 4 quency F1 33 5 2 .212 F2 315 59 20 6 1 30 .368 BC to P2 48 16 3 l 4 .500 BC to P1 51 15 3 3 .420 Individuals with greater than 4 branches per inflorescence may be either compound monochasial or compound dichasial inflorescences. Compound dichasial types may arise from indi- viduals expressing the intense branching phenotype (Figure 5a). The branching frequencies include all branched individuals. The frequency of compound monochasial inflorescences for the backcross pOpulations (42.0% for the backcross to MSU 100 and 50.0% for the backcross to Apsory) exceeded the F; (21.2%) and the F2 pOpulation frequency (36.8%). The back- cross values suggest that each parent contributed to in— creased branching. The F1 value (21.2%) is assumed to be the dominant ex- pression. If the backcross branched inflorescence frequen- cies are due to recessive genes, each parent (MSU 100 and 7O Apsory) must carry independent genes for increased compound monochasial inflorescence. The branching effect of MSU 100 estimated in the cross MSU 100 x Pennorange was 12.2%. The effect of MSU 100 on branching in the cross MSU 100 x Apsory appears to be much greater. The increased branching effect is probably due to an interaction of loci. The low branching frequency (21.2%) of the F1 population suggests that the interaction may be due to recessive genotypes. Jointless Pedicel gs. Number of Flowers per Inflorescence Single Flower per Truss (jointed pedicel) x pr—flower—number Inflorescence (joint- less pedicel) The close association of the jointless pedicel (1) character and a low number of flowers per inflorescence was suggested by Rick and Sawant (23) and Emery and Munger (6). Data obtained on jointless pedicel and flower per inflor— escence from the cross MSU 100 (single flower, jointed pedi- cel) x Pennorange (low—flower—number inflorescence type, jointless pedicel) also suggest a strong association between the two characters. Recombination values of 1.99 t .002 and 1.80 i .005 were observed in the backcross to Pennorange and F1 pOpulations respectively. F2 linkage values were calcu— lated according to the methods described by Immer (9). 71 Information regarding the relationship of the jointless pedicel to the character single flower (3) and to the low— flower-number type was obtained from the segregation of F3 populations derived from F2 selections of jointless and jointed pedicel single flower plants. F3 Populations Fifteen F2 individuals with jointed pedicel (g) and single flower (a) were observed. Three of the selections segregated for flower number and for the jointless pedicel character (Table 24). Assuming the recessive genotype (b) governing the-Pennorange phenotype (low-flower-number type) to be epistatic to the single flower genotype (Table 2), segregation was expected. Table 24. Segregation of F3 populations for flower number and jointless pedicel from jointed pedical (g), single flower F2 selections from the cross MSU 100 x Pennorange. Number of Single Low-flower- Jointless Line Plants Flower number Pedicel (1) 69—8 54 42 12 15 69-10 68 58 10 13 69-15 74 60 l4 l9 Four lines from jointless pedicel (j) and single flower (a) F2 individuals were also grown. Two lines showed a 72 distribution for flower number similar to Pennorange and two lines bred true for single flower. Two single flower and jointless pedicel F2 selections, reverted to the Pennorange phenotype when grown in the greenhouse. The relationship between jointless pedicel and the low- flower—number class was observed in the F3 generations. -AlI l~1 segregates for number of flowers in lines 69-8, 69—10 and 69-15 (Table 24) were jointless; therefore, pleiotropy or tight linkage was suggested. .g Assuming linkage or pleiotropy between jointless and low flower number, jointless, single flower selections are expected to segregate similar to the Pennorange parent. However, jointless single flower selections were observed to breed true for the single flower character in 2 of 6 F3 selections. F3 lines segregating for jointless pedicel only were also observed. The fact that jointless segregates from F3 populations of jointed pedicel F2 selections both breed true for single flower and segregate for number of flowers suggest that a gene or genes other than the single flower gene, a, is neces- sary for jointless types to exhibit the character single flower per truss. 73 LeafyfiIndeterminant Inflorescence Low—flower-number Inflorescence (leafy indeterminant) x simple Inflorescence While the jointless pedicel, leafy inflorescence-rela- tionship has been well-investigated (16,23), the relationship of jointless pedicel and leafy-indetenminant inflorescence has not. Leafy indeterminant inflorescence is associated with non—terminal growing of the inflorescence. Such growth is characterized by the production of flowers—followed by leafy growth then flowers followed by growth, etc. This pat- tern continues indefinitely. Parkin (20), in his discussion on inflorescence evolution, describes this character as intercalary inflorescence, provisionally referred to aS-ini“ in this study (Figure 4). The frequency of intercalary inflorescence in the parents, F1. F2 and backcross pOpulations of the cross Pennorange x Michigan State Forcing is presented in Table 25. Assuming the .155 frequency of Michigan State Forcing to be due to a dominant factor, since both the F1 and backcross to Michigan State Forcing exhibit similar frequencies. The Pennorange expression (.875) was suggested to be due to a recessive factor. The expected intercalary inflorescence frequencies for the F2 and backcross to Pennorange pOpulations were estimated from the formula iPl + §P2 for the F1 and 2P1 + 2P2 for the backcross to Pennorange where P1 is the fre- quency observed in Pennorange and P2 is the frequency observed 74 Table 25. Segregation for intercalary inflorescence (ini) in the various pOpulations of the cross Pennorange (P1) x Michigan State Forcing (P2). Number of in; Generation Plants in; Normal (+) Frequency Pennorange 48 42 6 .875 M.S.F. (P2) 45 7 38 .155 l F1 7 12 59 .169 "“1 F2 136 50 86 .367 BC to P2 98 16 82 .163 . E) BC to P1 120 62 58 .517 w in Michigan State Forcing.‘ The expected frequencies for the F1 and backcross to Michigan State Forcing equal that ob— served in the Michigan State Forcing population. The expected frequencies were used to determine the theoretical segrega— tion of different generations for chi-square analysis. The observed intercalary inflorescence segregation suggests a good fit to a monogenic inheritance (Table 26), with low frequency of occurrence the result of a dominant gene. Intercalary Inflorescence Xi- Jointless Pedicel The relationship between intercalary inflorescence and thejointless character was also investigated. .The frequency of ini types with the jointless (j) and jointed (g) characters 75 Table 26. Chi—square test for goodness of fit to a monogenic inheritance for intercalary inflorescence for the F1, F2 and backcross populations from the cross Pennorange (P1) x Michigan.State Forcing (P2). Number Intercalary Normal of Types Types Generation Plants Obs. .Exp- Obs. Exp. X? P F1 71 12 11.0 59 60 .1075 .80-.70 F2 136 50 53.7 86 82.3 .4213 .70-.50 BC to P2 98 16 15.2 82 82.8 .0499 .90-.80 BC to P1 120 62 61.7 58 58.3 .0029 .98-.95 were recorded from the F2 and backcross to Pennorange. Frequencies of .939 and .935 for jointless progeny in the F2 and backcross to Pennorange respectively were not signifi- cantly different from the .875 frequency observed in the Pennorange pOpulation. The frequency of in; g progeny ob- served in the F2 (.181) did not differ from the frequency observed in the Michigan State Forcing parent (.155), but the .061 in; i frequency observed in the backcross to Penn— orange was significantly different from the Michigan State Forcing. These data suggest that the expression of intercalary inflorescence may be a function of the jointless and jointed pedicel characters. The apparent single gene segregation for intercalary inflorescence may be segregation for jointless and jointed pedicel. 76 If intercalary inflorescence is linked to jointless pedicel, recombinants of jointless and low in; frequency and jointed and high in; frequency would be apparent in the F2 and backcross to Pennorange population frequencies. Jointed high in; recombinants are expected to cause the in; fre- quency to be greater than that exhibited by the Michigan State Forcing parent and jointless-low-igi recombinants are expected to cause the igi_frequency to be lower than that exhibited by Pennorange (.875). However, the only significant deviation was the low.jointed pedicel in; frequency (.061) observed in the backcross to Pennorange. Since no effects of linkage were observed, pleiotropy or close linkage between intercalary inflorescence (ini) and jointless pedicel (j) is suggested. DISCUSSION The inheritance of the inflorescence types, single flower per truss (a), compound (g), simple and low-flower— number was determined. Inheritance studies were conducted for flower number on the unbranched simple inflorescence and the occurrence of the compound monochasial (branched inflor- escence) characters. Single Flower per Truss The character single flower per truss was suggested to be controlled by a single recessive gene from crosses to a simple inflorescence line (MSU 180), and a low-flower-number per inflorescence line (Pennorange) (Table 2). The crosses of single flower (MSU 100) to compound inflorescence lines (MSU 200 and Apsory) showed segregations that suggested a more complex inheritance. A model (Table 11) composed of the single flower per truss gene (a), an inhibitor (I) gene which inhibits the expression of a_and a restorer gene (3) which restores the expression of a in the presence of I was proposed. The segregation for single flower in the F2 and backcross to single flower parent for both crosses of single flower x compound inflorescence showed a good fit to the 77 78 expected base on this model. The prOposed model could not explain the 15:1 ratio of single flower to simple inflores— cence noted in the F3 generations from single flower F2 selections. Low Flower Number per Inflorescence The low flower number phenotype of the Pennorange variety is suggested to be controlled by a single-recessive gene (Tables 2,5). The 9 'F1 c1ass' : 4 low flower class : 3 single flower per truss segregation observed in the (MSU 100 x Pennorange) F2 (Table 2), suggests the gene for low flower class to be epistatic to the gene single flower per truss. Compound Inflorescence Compound inflorescence was previously reported to be controlled by a single recessive gene (g) (5,14). Deviations from monogenic inheritance (4) were reported. In the present study segregation for compound inflorescence types deviated from the expected in (MSU 100 x MSU 200) F2 and (Apsory x MSU 180) F2 populations. Segregation of the (MSU 100 x Apsory) F2, backcross to Apsory and (MSU 100 x MSU 200) x MSU 200 pOpulations did not differ from the previously reported segregation. Individual plants with high flower numbers (30-80 flowers) and many branches on the inflorescence were observed 79 in crosses involving compound inflorescence. An association between the high flower number segregates and compound inflor— escence (g) is suggested from the following evidences: a) The deficiency of gs recombinants in the F2 pOpula- tions. b) Both high flower number and compound inflorescence were suggested to be linked to the dwarf plant habit (d) (Table 12). c) Previous reports suggested that 200-300 flowers per inflorescence is not a prerequisite for compound inflorescence (3,13). The phenotype of compound inflorescence expressed by the MSU 200 and Apsory parents (200-300 flowers) is the re- sult of intense branching of the inflorescence and a non- terminal flowering habit. Recombinants with either intense branching or non—terminal flowering were Observed in segre- gating populations, suggesting the MSU 200 and Apsory to carry genes for intense branching and non—terminal flowering. The intense branching character exhibits high flower numbers, suggesting that this may also be conditioned by the g gene. Linkage values between intense branching and non-terminal flowering were calculated from the backcrosses (MSU 100 x Apsory) Apsory and (MSU 100 XTMSU 200) MSU 200, and the F2 of Apsory x MSU 180. Recombination values ranged from 6% to 11%. The observed segregation frequencies of compound inflor- escence show close agreement to the expected values based on the linkage model (Table 14). 80 Flower Number on the Unbranched Mono- chasial Inflorescence The F1 phenotype (simple inflorescence) was suggested to be the dominant class. The mean number of flowers on the simple inflorescence progenies from the backcross to Pennorange and the backcross to MSU 100 suggested segrega— tion of modifiers which affected the number of flowers on the unbranched monochasial inflorescence. The modifier or modifiers contributed to Pennorange is suggested to give a phenotype of 4-flowers per unbranched infloresence, and the modifier or modifiers transmitted by MSU 100 were suggested to give a phenotype of 8-flowers per unbranched inflorescence. The two phenotypes exhibited no dominance and the F1 phenotype showed a mean of 6 flowers. Segregation for flower number class within the simple inflorescence was observed in the cross MSU 100 x MSU 200. The classes, lO-flowers and 7-fIOwers, exhibited a digenic 9:7 ratio of lO—flowers to 7-flowers. Compound Monochasial Inflorescence The occurrence of compound monochasial inflorescences in the different generations from the cross MSU 100 x Penn- orange was investigated (Table ZI). The factor(s) transmitted by Pennorange and MSU 100 was suggested to contribute less than 3.6 percent and from 12.2 to 30.4%, respectively. The F1 showed 8.7 percent compound monochasial types and their occurrence was suggested to be related to the genotype con— ditioning number of flowers per inflorescence. 81 The estimated frequency of compound monochasial inflor- escences transmitted by the MSU 100 parent in the cross MSU 100 x MSU 200 was 13.2 percent and.is similar to the value estimated in the cross MSU 100 x Pennorange. The fre- quency of the F1 was 59 percent and the backcross to MSU 200 was 58 percent suggesting that the higher frequency is due to a dominant factor. The observed frequencies of the F1, F2 and backcross generations gave a good fit to a monogenic hypothesis. The frequency of compound monochasial inflorescences noted in the cross MSU 100 x Apsory suggest that the MSU 100 parent contributed an effect greater than was previously noted in the MSU 100 x Pennorange and MSU 100 x MSU 200 crosses. The high frequencies in the backcross pOpulations (42% and 50%) and the low frequencies of the F1 (21.2%) and F2 (36.8%) may be due to an interaction or recombination of several genes. Intercalary Inflorescence The character intercalary inflorescence, indeterminant vegetative growth in the inflorescence, was suggested to be conditioned by a single recessive gene, in; (Table 26). Studies on the relationship of the jointless pedicel (j) to intercalary inflorescence (ini) suggest the jointless gene to be either closely linked to the gene for intercalary inflorescence or the genes are pleiotrOpic. 2 SUMMARY AND CONCLUSIONS 1. The results of crosses between single flower per truss (MSU 100), compound inflorescence (Apsory and MSU 200) simple inflorescence (MSU 180 and Michigan State Forcing) and low—flower-number per inflorescence (Pennorange) were evaluated to determine the mode of inheritance for these characters. The simple inflorescence phenotype was evaluated for inheritance of number cf flowers on the unbranched.inflor- escence and for the occurrence of compound monochasial inflor— escences. 2. Single flower per truss was suggested to be condi- tioned by a single recessive gene, designated as a, from the crosses MSU 100 x Pennorange and MSU 100 x MSU 180. Single flower inheritance in the crosses MSU 100 x Apsory and MSU 100 x MSU 200 was suggested to be more complex. A three gene model was proposed to explain the probable mode of inheritance: a - conditions single flower per truss IH - inhibitor of a (single flower) expression Iw — restores a (single flower) expression in the presence of I. 82 6 83 3. Low-flower—number inflorescence, as found in the Pennorange parent, was suggested to be conditioned by a single recessive gene, from crosses MSU 100 x Pennorange and Pennorange x Michigan State Forcing. The low flower gene was suggested to be epistatic to the gene for single flower per truss. 4. Compound inflorescence, as expressed by the parents, Apsory and MSU 200, was suggested to be conditioned by two linked genes, intense branching and non-terminal flowering (93;), with recombination values between genes for intense branching and non-terminal flowering of 6 percent in the crosses MSU 100 x Apsory and Apsory x MSU 180 and 11 percent in MSU 100 x MSU 200. 5. Simple inflorescence phenotypes exhibited segregation for number of flowers on the unbranched inflorescence. Modifiers were prOposed to condition the 4-flower class and 8-flower class observed in the cross MSU 100 x Pennorange and the 7 and 10—flower classes observed in the MSU 100 x MSU 200 cross. The modifiers conditioning the 4 and 8-flower classes were suggested to show no dominance while the digenic 9:7 ratio of lO-flower class to 7-flower class suggests dominance of the lO-flower class. 6. The occurrence of compound monochasial inflores- cences in the cross MSU 100 x Pennorange was suggested to be conditioned by a single recessive gene, designated as g. The gene was suggested to behave as follows: 84 cc - conditions low frequency of compound monochasial types (3.6 percent or less) 9g - conditions 8.7 percent compound monochasial types QC - conditions 12-30 percent compound monochasial types. 7. Compound monochasial inflorescence, as noted in the cross MSU 100 x MSU 200, was thought to be inherited as a single gene. 1 8. Intercalary inflorescence was suggested to be condi— c tioned by a recessive gene, 22;, which is either closely linked or pleiotrOpic with the jointless pedicel gene, 1. B IBLIOGRAPHY 10. 11. 12. BIBLIOGRAPHY Azzam, H. 1962. Abortive flower-cluster mutant in tomatoes. J. Agr. Univ. Puerto Rico. 46 : 69-72. Butler, L. 1936. Inherited characters in the tomato. II. Jointless pedicel. J. Hered. 27 : 25-26. Butler, L. 1952. The linkage map of the tomato. J. Hered. 43 : 25—35. Crain, M. B. 1915. Heredity of types of inflorescence and fruits in the tomato. J. Genet. 5 : 1-11. E1 Sayed, M. N. K. and R. L. Foskett. 1963. Masking of the sp gene in the jointless tomato. Proc. Amer. Soc. Hort. Sci. 83 : 705—709. Emery, G. C. and H. Munger. 1970. Alteration of growth and flowering in tomatoes by the jointless genotype. J. Hered. 62 : 51-53. Fehleisen, S. 1967. Uniflora and conjunctaflora : two new mutants in tomato. T. G. C. Rep. 17 : 26-28. Honma, S., S. H. Wittwer and S. C. Phatak. 1963. Flowering and earliness in the tomato. J. Hered. 54 : 212—218. Immer, F. R. 1930. Formulae and tables for calculating linkage intensities. Genetics 15 : 81—98. Leonard, W. H., H. 0. Mann and L. Powers. 1957. Partitioning method of genetic analysis applied to plant height inheritance in barley. Colo. Agr. Exp. Sta. Tech. Bull. No. 60. Lewis, D. 1953. Some factors affecting flower produc- tion in the tomato. J. Hort. Sci. 28 : 207—220. Mac Arthur, J. W. 1928. Linkage studies with the tomato. Genetics 13 : 410-420. 85 13. 14. 15. 16. 17. l8. 19. 20. 21. 22. 23. 24. 25. 26. 86 Mac Arthur, J. W. 1934. Linkage groups in the tomato. J. Genet. 29 : 123-133. Mather, K. 1949. Biometrical Genetics. Dover Publica- tions, New York. 158 pp. Mather, K. and J. L. Jinks. 1971. Biometrical Genetics (2nd ed.) Cornell Univ. Press., Ithaca, New York. 382 pp. Mertens, T. R., and A. B. Burdick. 1954. The morphology, anatomy, and genetics of a stem fasciation in Lycopersicon esculentum. -Amer. J. Bot. 41 : 726-732. Muller, C. H. 1940. A revision of the genus Lycopersigon. U.S.D.A. Misc. Pub. 382. Myers, C. E. 1944. Stemless tomatoes are unlikely to gouge each other in basket. Penn. Agr. Exp. Sta. : Ann. Rep. 56. Suppl. 1 to Bull. 446. Paddock, E. F. and L. J. Alexander. 1952. Cauliflower, a new recessive mutation in tomato. Ohio J. Sci. 52(6) : 327—334. Parkin, John. 1914. The evolution of the inflorescence. J. Linn. Soc. 42 : 511-563. Powers, L. 1955. Components of variance method of genetic analysis applied to weight per fruit of tomato hybrid and parental pOpulations. U.S.D.A. Tech. Bull. No. 1131. Powers, L., L. F. Locke and J. C. Garrett. 1950. Partitioning method of genetic analysis applied to quantitative characters of tomato crosses. U. S.D.A. Tech. Bull. No. 998. Rick, C. M. and A. C. Sawant. 1955. Factor interaction affecting phenotypic expression of the jointless character in tomato. Proc. Amer. Soc. Hort. Sci. 66 : 354-360. Steele, R. G. D. and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill Inc., New York. 481 pp. ' Verkerk, K. 1962. Mutagenic action of neutron radiation on tomatoes. XVIth Int. Hort. Congr. II : 124-127. Young, P. A. and J. W..Mac Arthur. 1947. Horticultural characters of tomatoes. Texas Agr. Exp. Sta. Bull. 698. MICHIGAN STATE UNIV. LIBRRRIES l”WI!”WIWIWIIWI'HH‘VIlWHIMIWI 31293103229500