. . I a I . . : . . . . . . . . r ~51...an... «use ”.311 .“nfiJ—aazridnfild: in. law... v5.3...rnfiufiihosxrgllu: us . grgcidfinfidawn . , x r . u . 3. . It; . . x 1 5?». : ... I. .1 _. . Star... E .r I"... :3. 1.. .x: ....‘:vr.cr... 3.113;... . . . 3...! .. s 55.6... . L. Lu». .9 .. 5. .. ... A . L i L.. .fr.v....... L5. , . . . .v 53/. u .1 A ..A ... . ..,.....u.,.ys,.ww# "‘ , , t . .31; r1... . . , ‘1’. 1:...u-flr. , , , . . ... 1 r\ . . . . . f. .x... r r7 . v. . . 1:99. I Y .19... r _ .scsojf » 4 . . 4. (I; f. .‘IZLLLII L311 .32. A .. .2113: 5.1.5.. . E: f 1 :v I... w. m, . v‘mw r .......,x.........w“w.. ;, _ ., . 11.. . ids. 5.... p J / r a s _ (1"- ? ' D . .. . , . a??? $2.4. V , . V , V 5., .. amxnlrrufi. to; . . V’flwv _ a A . .- .. c f a. £15}. 1., -. .. .. .2; 2 . V 7... :3. . .(. ..11r~ v . ..,. 4.6;: , ..r qgflu-‘1‘ .1 ,. a .w. -nw-uumwu .<. . . 6.8.... 2.. 3.! .C . A. t , p... tutu-VJ. .. I. .5 6..) rt... aW/nflwxm (‘ coin ~11 .r. . I .av..\.2 .r. 5.... . . . .. .. . . . , .2233 ” rerHhmzsd, vi... . .. , . . . z , a... fixiznr : 17.2. n; . ,. . , , . . , . $4.1: Jar ., . v a . r) in” n. .1...» 5.. r :1» fr mq‘g-‘,..j. 1" l tuna. n Woo-mam . tn}; . VI: .1- r. Ewavvvri. u 7. . 2.. ”($71... (I: I , IL-I. nah? :3: w......w......uz. .......m..w....?,fhh.r.i... {£1} 1W3}?! a» .V VI: ”Fifivtgf . . x... LC: 2:... .ll.)..:§ veg ihiiwffiuwx i: .§¥&Prrfirt§w«u¥ufig ctf‘}.i. m} .. tb.(n‘1s}u§u\.fn .r «LMM- n mgmnywymmnwrmwwwmmum $2.335; 186 . _ Umv rrsrty ‘ 147.336. 5-. ‘t‘?’ 31” -' In "an. This is to certify that the thesis entitled THE INHERITANCE OF SEVERAL CHARACTERISTICS IN CELERY AND TOMATO presented by John C . Bouwkamp has been accepted towards fulfillment of the requirements for Ph-D degree in Horticulture Major professor Date Feb. 13, 1969 0-169 ABSTRACT THE INHERITANCE OF SEVERAL CHARACTERISTICS IN CELERY AND TOMATO By John C. Bouwkamp Studies were carried out on celery to determine the inheritance of vernalization response, pinnae number, and leaf shape. It was found that vernalization response was due to a single dominant major gene, Yr (which resulted in easy bolting). Cytoplasmic effects on the response to vernali- zation were also noted. Pinnae number was effected by two genes which segregated in the F2 giving a 13:3 ratio. Leaf shape was determined by a single gene pair, d3, with shallow toothed leaves dominant to deeply toothed leaves. Studies on tomatoes were conducted to determine the mode of inheritance and linkage relationships of sepal length, sepal shape, fruit shape, beaked fruit, and pedi- cel length. It was found that fruit shape was determined by two genes, designated orf and Egrf with additive effects. These two genes were located 40 recombination units apart. Beaked fruit was conditioned by a single dominant gene designated Bk—2. This gene was located approximately mid— way between the loci of the two fruit shape genes, or.f and John C. Bouwkamp f ggf . Sepal shape was determined by a single gene pair designated en for ensiform sepals. The dominant allele causes acuminate sepals. The gene en was located at 3“ recombination units from Bk—2 and 16 units from the near- est fruit shape gene. Sepal length was determined by a single gene pair with incomplete dominance and designated ii; §§ §£ conditioned long sepals; §_ ii, short; and §§ ii: intermediate. S5 was located 20 recombination units from 22, 39.“ units from Bk-2, and 24 units from the nearest fruit shape gene. Pedicel length was found to be quanti— tatively inherited. Correlation analysis substantiated the gene order postulated from the recombination data- THE INHERITANCE OF SEVERAL CHARACTERISTICS IN CELERY AND TOMATO By John C. Bouwkamp A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Horticulture 1969 DEDICATION To my father, mother, and Ellen \ ii ACKNOWLEDGMENTS The writer wishes to express his sincere apprecia— tion to Dr. Shigemi Honma for his assistance and criticism in preparation of this manuscript. Appreciation is also tendered to Dr. S. H: Wittwer, Dr. K. T. Payne, Dr- J. D. Downes, and Dr, L. R. Baker for their helpful suggestions. The writer also wishes to express his appreciation to Mr. Jerry Vreisenga and Mr: Amos Lockwocd for their help in maintenance of field plots. The writer especially wishes to express appreciation to his wife, Ellen, for her continuing encouragement and her assistance in preparation and proof reading the manuscript. iii TABLE OF CONTENTS Page ACKNOWLEDGMENTS . . . . . . . . . . . . iii LIST OF TABLES. . . . . . . . . . . . . Vi LIST OF FIGURES . . . . . . . . . . . . V111 PART I: THE INHERITANCE OF VERNALIZATION RESPONSE, PINNAE NUMBER, AND LEAF SHAPE IN CELERY (Apium graveolens L. var. dulce) INTRODUCTION 2 REVIEW OF LITERATURE. A Vernalization. A Leaf Shape. 7 MATERIALS AND METHODS 9 RESULTS AND DISCUSSION . . . . . . . . . . 14 Bolting. . . . . . . . . . . . . . 1A Petiole Color. . . . . . . . . . . . l9 Pinnae Number. . . . . . . . . . . . 20 Leaf Shape. . . . . . . . . . . . . 28 SUMMARY . . . . . . . . . . . . . . . 40 PART II: THE INHERITANCE OF FRUIT SHAPE, BEAKED FRUIT, SEPAL LENGTH, SEPAL SHAPE, AND PEDICEL LENGTH IN TOMATO (Lycopersicon esculentum Mill) INTRODUCTION . . . . . . . . . . . . . A3 REVIEW OF LITERATURE. . . . . . . . . . . 45 MATERIALS AND METHODS . . . . . . . . . . 48 iv RESULTS AND DISCUSSION Fruit Shape Beaked Fruit Sepal Length Sepal Shape . Pedicel Length SUMMARY CONCLUSIONS. BIBLIOGRAPHY Table 10. LIST OF TABLES Mean bolting percentages for various durations of cold treatment for celery plants from F2 populations from reciprocal crosses of Golden Plume x MSU l30 Variance analysis of bolting percentages in F populations from reciprocal crosses of Géiden Plume x MSU 130 Frequency (in per cent) of parental and F populations from reciprocal crosses of Golden Plume x MSU 130 for pinnae number and a Chi—square test for goodness of fit to a 13:3 ratio Chi—square test of bolting and non—bolting groups for heterogeneity of pinnae number in pooled F2 Chi-square test for goodness of fit to a 39:13:923 ratio of green P2: yellow P9: green P : yellow Pl: assuming indepen ence Components of Chi—square test for goodness of fit to a 39zl3 9:3 ratio between petiole color and pinnae number Chi-square test for goodness of fit for pooled F2 populations from reciprocal crosses of MSU 130 x Golden Plume based on a one factor pair hypothesis for leaf shape. Chi—square test for goodness of fit to a two gene additive model for fruit shape assuming independence F2 genotypes, phenotypes, and zygotic frequencies for fruit shape Observed and expected frequencies of fruit shape phenotypes when linkage is concerned . . . . vi Page 15 l6 23 25 26 27 37 55 57 Table ll. l2. 13. 14. l5. l6. 17. 18. 19. 20. 21. 22 Chi-square test for goodness of fit to a two gene additive model for fruit shape assum— ing a linkage intensity value of 0.2 Gametic types and their frequencies for fruit shape and beaked fruit Chi-square test for goodness of fit to a three gene model for fruit shape and beaked fruit with a linkage intensity value of 1.0 Chi—square test for goodness of fit to a 3:6:3:1:2:1 ratio of sepal length and beaked fruit with a linkage intensity value of 0.212. Chi—square test for goodness of fit to a 1:2:1:A:3:A:6:12:6:5:10:5 ratio for sepal length and fruit shape assuming a linkage intensity value of 0.52 Chi—square test of goodness of fit to a 9:3:3:l ratio for sepal shape and beaked fruit assuming a linkage intensity value of 0.32 Chi-square test for goodness of fit to a 3:6:3:1:2:l ratio for sepal length and sepal shape assuming a linkage intensity value of 0.60 Chi—square test for goodness of fit to a 3:1:12:A:18:6:15:5 ratio for fruit shape and sepal shape assuming a linkage intensity of 0.68. Possible types of gametes bearing the genes ii: en, and Bk—2 Phenotypes in the F2, observed individuals in the phenotypic classes, and calculated cross—over individuals for sepal length and shape and beaked fruit. One postulated order and distance between the genes Eff, Lgrf, Bk-2, en, and §§ Correlation coefficients between pedicel length, sepal length, sepal shape, fruit shape, and beaked fruit. . . . vii Page 59 61 66 7O 71 73 74 77 80 Figure 1. LIST OF FIGURES Stages in seedstalk development. Left, not bolting; center and right, bolting; growing points outlined for ease of identification . . Percentages of plants bolting after various days of treatment. Squares, 980 (MSU 130 x Golden Plume); triangles, 981 (Golden Plume x MSU 130) . . Number of pinnae per petiole: top eight, bottom ten . . . . . . . Leaf shape of the two parental types. Left, Golden Plume (P2) shallowly toothed; right, MSU 130 (P1) deeply toothed . . . Scatter diagram of leaf shape data, regres- sion line (y = —2.67A2 + 2.2602x), and 1% confidence belts about the regression line for the Golden Plume parent Scatter diagram of leaf shape data, regression line (y = —2.0132 + 1.87l8x), and 1% confidence belts about the regres— sion line for the MSU 130 parent. Scatter diagram of leaf shape data for 980 (MSU 130 x Golden Plume) and parental regression lines . . Scatter diagram of leaf shape data for 981 (Golden Plume x MSU 130) and parental regression lines . . . . 1% confidence belts for both parents showing leaf five shape classes in the F popula- tion. (All observations in overlapping confidence belts—-shaded area—-were dropped). viii Page 12 18 22 29 3O 31 33 35 Figure 10. ll. l2. 13. 14. 15. Relative frequency of MSU 130, Golden Plume, and pooled F2 from the reciprocal crosses of Golden Plume x MSU 130 in various dis— criminant value classes. . . . . Parental and F1 fruits, calyxes, and pedicels. Upper left, P ; upper right P2; and lower center, F (seé text for fruit length: breadth ratios) Relative frequency of fruit shape ratio for parental, F1’ and F2 populations. Relative frequency of sepal lengths (in mm) for parental, F1’ and F2 populations Sepal shapes. Top row, recessive en showing ensiform sepals of P1; second row, heterozygous showing acuminate sepals of F ; bottom row, homozygous En with acumin- ate sepals of P2 Relative frequency distributions of pedicel length for parental, F1’ and F2 populations. . . . . . . ix Page 39 51 52 6A 68 ln|l A THE INHERITANCE OF SEVERAL CHARACTERISTICS IN CELERY AND TOMATO I THE INHERITANCE OF VERNALIZATION RESPONSE, PINNAE NUMBER, AND LEAF SHAPE IN CELERY (Apium graveolens L. var. dulce) INTRODUCTION Exposure to a period of cold temperature is known to be a prerequisite for flower initiation in many bi- ennial species. This phenomenon,referred to as vernali— zation, is commonly defined as the acquisition by a plant of the ability to hasten or promote flowering by exposure to low temperature. This definition will be used in this study. Celery (Apium graveolens L. var. dulce) set in the field without protection early in the spring or ex— posed to cold temperatures in the greenhouse prior to transplanting, often bolts before reaching marketable size. The occurrence of bolters varies depending on the duration of the cold period during the growing season. This results in a considerable loss to the growers since the bolted plants are not marketable and the cost of pro- duction is high for early celery. One solution to this problem is the use of paper or plastic coverings. The cost of these protective coverings is high, and the re- sults are not entirely satisfactory. Although they pro— vide some protection, they are not adequate for long dark cool periods. ..I..llrl| ‘ A more satisfactory solution to this problem is through the use of varieties resistant to bolting. Al- though genetic control of bolting has been known for some time, no study to determine the mode of inheritance of the response to vernalization has been reported in celery. The objectives of the following study were: 1. To identify the factors controlling the inheri— tance of pinnae number, leaf shape, and response to vernalization. 2. To establish linkage relationships of these three characters as well as the factor conditioning yellow petiole color. REVIEW OF LITERATURE Vernalization There have been several recent review articles and books on the subject of vernalization and flower induction (7, 20, 26, A6, 72). No attempt will be made to summarize and compile all of the literature on vernalization, how— ever, those articles pertinent to this study will be in- cluded. Bolting has been known in biennial vegetable crops for many years. Krasan, as reported by Thompson (6A) in 1890 called attention to the importance of temperature in the initiation of flowering in many species. Starring (51) and Thompson (63) reported that bolting in celery could be induced by a two to three week treatment at A0 — 60° F. Several factors which check growth (such as frost, drought, poor soil conditions and crowding the plants in the seed tped) were also studied. Both researchers concluded that iri celery prolonged exposure of the plants to cold temper- zature was responsible for premature seed stalk development. Several investigators have shown that all stages of plant growth are not equally responsive to the vernalizing stimulus. Pawar and Thompson (35), working with celery, A conclJided that the age of the plant was more important than plant size. Sarkar (A7) reported that, in the bien- nial Hyoscyamus niger, seedlings less than ten days old could not be vernalized. The sensitivity to vernaliza— tion increased until the plants were thirty days old. Older plants showed no increase in sensitivity over the thirty day old plants. Wellensiek (72) proposed that some biennials have a "juvenile phase" and, during this phase, plants were insen- sitive to the vernalization stimulus. Cheiranthus allioni (3) and Arabidopsis thaliana (32) also show differences in response to vernalization at the various stages of develop- ment. Chouard (7) concluded that sensitivity generally increased with age. Several workers (8, 38, 63) have reported that the receptive site for vernalization was the stem apex. Curtis and Chang (13), working with celery, cooled the growing point by circulating cold water in tubing wrapped around the plant at the base of the petioles. The cooled plants grown in a warm greenhouse bolted while the uncooled con- trol plants remained vegetative. Plants with their apices warmed by circulating warm water failed to bolt while the control plants bolted when grown in a cool greenhouse. Wellensiek (72), working with Lunaria biennis, con- cluded that mitosis was necessary for vernalization. He based his conclusions on the fact that leaf cuttings from veiwuilized plants regenerated a flowering plant only if the leaves were quite young. Older leaves from a vernal- ized plant regenerated vegetative plants. His findings, however, do not preclude the possibility that vernaliza- tion is dependent on young, undifferentiated cells and not mitosis as such. Although several authors (28, 3A, 50, 73) noted varietal differences in response to vernalization, there are few studies concerning the genetics of the response. Hall (17) reported that bolting was not simply inherited but that non—bolting appeared to be dominant in sugar beets, mangolds and leeks. Van Heel (67), working with sugar beets, concluded that bolting was recessive and was controlled by a single gene. Sutton (60) found that bolt— ing in cabbage was conditioned by recessive factors. Barber (2) concluded that in peas a single gene, gn, was conditioned for the response to vernalization. Response was dominant to non-response. It might be noted, however, that peas do not require vernalization in order to flower but that vernalization only makes peas flower sooner. Nilsson—Ehle (33) and Cooper (10), studying crosses of winter and spring wheat, found that the spring type was dominant to the winter type giving a 3:1 ratio. COOper also reported a 13:3 ratio using another winter wheat par- ent. Takahashi (61) reported that the spring type was ciominant to the winter type in barley giving a 3:1 ratio. gaixmis (16), however, reported a 13:3 ratio of spring to winter types in barley. Tschermak (66) found a 3:1 ratio of spring to winter types in rye but noted a difference in reciprocal F2 populations. Purvis (37) reported similar results with rye. Elmsweller (15) demonstrated a difference in bolting response among inbred lines of celery and suggested that bolting resistance was quantative and recessive. Thompson (63) also concluded that varieties and strains differed in their response with reference to bolting. Starring (52) made similar observations and suggested that it was possible to select and breed for resistance to bolting. Leaf Shape The existence of genes controlling shape was demon- strated with cucurbit fruits by Sinnott (A8, A9) who noted that ”shape" genes did not control individual dimensions but controlled the correlations between growth rates in different dimensions or planes. Several genetic studies of leaf shape have been re- ported. Tedin (62) found leaf shape in Camelina to be controlled by two genes. The pinnatified parent was designated as AABB while the entire-leafed parent was aabb. There were also two true breeding intermediate types which were designated AAbb and aaBB. lmai (22) reported more than twenty genes affected the leaf shape in Pharbitis nil, the Japanese morning glory. Joyner and Pate (2A) found in Hibiscus cannabinus that deeply lobed leaves were due to a single gene which was dominant to the gene conditioning shallowly lobed leaves. Ryder (A5) reported three genes affecting leaf shape in lettuce. Extensive genetic studies on leaf shape have been made in the genus Gossvpium (l8, 19, 5A, 55, 56, 57, 58). Stephens (SA) has shown that leaf shape is controlled by a system of four alleles in amphidiploid cotton. In this system, the allele for the more deeply lobed form is al- ways dominant to the allele for the less deeply lobed form. Rhyne (A0) demonstrated that phenocopies of the amphidiploid, New World cottons could be produced by com- binations of alleles for leaf shape in the diploid and the Asiatic series. Thus the four allelic leaf shapes in the amphidiploid cotton might have arisen from combinations of two leaf types in each of the progenitors of the amphidi— ploid. Several genes have been reported which affect leaf shape in tomato. They are potato leaf, 9 (5), entire leaf, 3 (5), curl, Cu (76), trifoliate, 2: (AA), mouse ear, Me (A2) and lanceolate, La (29, 30, 59). MATERIALS AND METHODS In 1960 a program was started at the Michigan Agri- cultural Experiment Station to obtain materials to study the mode of inheritance of bolting in celery. The general plan was to establish diverse lines so that the characters could be readily identified in the segregating population. The easy bolting line was obtained by exposing the yellow petioled variety Golden Plume to two weeks of vernal- izing temperatures (A0o F) and selecting plants that bolted. These plants were selfed in the greenhouse and were again subjected to the same treatment. After five generations, a line which bolted after two weeks of vernalization was established which henceforth will be referred to as Golden Plume. The hard—to—bolt line was obtained from a cross between Early Fortune and Emerson Pascal. The F2 genera— tion was exposed to six weeks of vernalizing temperatures and green petioled plants that failed to bolt after this treatment were selected, given additional cold treatment, and selfed in the greenhouse. Plants resulting from self- ing were again subjected to the same treatment; this was repeated (four generations) until a line that would not 10 boli. after six weeks of cold treatment was established. This line is henceforth referred to as MSU 130. Reciprocal crosses were made between one plant of Golden Plume and one of MSU 130 according to the method outlined by Honma (21). The plants were also selfed in the greenhouse to obtain parental seeds for use in the study. The F1 plants were grown in the field, lifted, vernalized, and selfed in the greenhouse to obtain the F2 Seeds. The F2 seed from the cross MSU 130 x Golden Plume was labelled 980 and its reciprocal, 981. The importance of including the F and back cross populations was 1 realized; however, celery crosses are difficult to make and if successful only a few seeds are obtained per cross (21). Seeds from the parents and the two F2 populations were sown in vermiculite. At the first true leaf stage, the seedlings were planted in flats on 1 1/2 inch centers into a sterilized soil mixture of 1/2 muck, 1/2 sand. Cold treatments were started when the plants were at the five to seven true leaf stage. The experiment consisted of seven treatments with three replications of each treatment. Each replication consisted of 2A plants of each parent and 72 plants of each of the F2 populations. The treatments consisted of different durations of exposure to a constant AO° F temperature and a 12 hour day length. The shortest ll eXDCDSure was 20 days and the exposure time increased at five day intervals to 50 days. In addition, extra seed- lings of MSU 980 were given a ten and fifteen day exposure. After the cold treatment, the plants were kept in a greenhouse at 70° — 75° F for a week in order to stabilize the vernalization. The plants were then transplanted to a plot at the MSU Muck Experimental Farm near Bath, Michigan. The transplanting was delayed until July 15 in order to eliminate the possibility of the plants receiving an ex— posure to natural cold temperatures. During the growing season, the plants were fertilized and irrigated as necessary. Records were taken on the number of bolted and yellow petioled plants in the field. Plants with a long duration of treatment had a shorter normal growing season than plants with a shorter treatment time. In order to elimi— nate the variable of rate of seed stalk development, all plants without visible seed stalks were cut longitudinally to observe the apices. All plants with pointed apices were considered as bolters (Figure 1). One petiole from each plant from the 25 and 30 day treatments was sampled and the pinnae number was recorded. The terminal leaflets from these petioles were saved and pressed in an herbarium press for measurements. The data on the length, weight, and leaf area of the terminal leaflets were taken when the samples were dried. The differences in leaf area were measured using Figure l.——Stages in seedstalk development. Left, not bolt— ing; center and right, bolting; growing points outlined for ease of identification. 13 an apparatus similar to that described by Potter (36). The images of one hundred leaves were reproduced with a Xerox copying machine and their areas were measured with a planimeter. The data from the light meter apparatus and from the planimeter measurements were used to obtain a linear regression equation. A correlation coefficient of 0.968 was obtained between the measured leaf area and the units from the apparatus. Since the r-value was highly significant, the light meter reading of the leaflet samples was transformed using the linear regression equation (leaf area in cm2 = 0.78 + 0.38 light units) prior to analysis. The estimate of genetic differences of certain char- acters in this cross was based on the ratio obtained by apportionment of the F2 data where separation was suggested by the bimodal distribution. Observed frequencies were compared with theoretical ratios using the Chi-square method. Data from individual plants were assembled and entered on IBM cards. Means, variances, standard devia- tions, and correlations were obtained from individual plant data and calculated by the use of the Control Data Corporation 3600 computer. The results were statistically interpreted by using statistical tables in Steel and Torrie (53). RESULTS AND DISCUSSION Bolting The results from the bolting experiment for the two F2 populations studied are shown in Table 1. These results suggest that the character "bolting response" is not a simply inherited trait. The variance analysis computed for bolting is shown in Table 2. The significant F value for F2 populations suggests a cytoplasmic difference in the inheritance of this character. Cytoplasmic differences might be explained by the current theories on vernalization and gene action if one considers vernalization in celery to be an inducible ge- netic system. Such a system would require a change in concentration of some chemical or biochemical constituent in the cell which in turn would regulate the action of a gene. Celery requires a prolonged (2-A week) cold treat- ment for bolting. A short, very cold treatment (i.e., a frost) will not induce bolting, but a longer cool treat- ment causes bolting. This suggests that some accumulative chemical changes may be taking place. 1A l5 m.w> m.mw N.mm s.mm m.Hm H.ww 3.50 Hmm m.mm :.ss s.ws w.ws m.sm m.om m.mw m.am H.3m 0mm oa ma om mm om mm o: m: om coupsasaoe mm pcoEpwohp paoo mo mama .oma mm: x madam sooaow mo mommomo Hmoommfiomu Eonm mGOdeHonm mm Eonm mpcwam zhoamo pom onEpmmup taco mo mCOHpmpzp mSOHLw> pom momMucoosod mcfluaon cmmzll.H mqmfimsHocH memEpmomo amp om cmSOLSp om no mfimzamc< .Hm>ma moo.o pm undefiuuewum*** a: saamma.o Hmpoe mmnaoo.o mm szomo.o LOLLM .m ***mam.m mmm:ao.o H mmmsao.o mcoapmfiseom mm oomHHo.o m Hmmmmo.o hmflpo ***wwm.mm bmwfiwo.o H waamo.o kw®CHq ***Hma.ma mmwamo.o m wmmama.o mucoEpmeB Nmm.o JMHHO0.0 m pmmmoo.o mmmm m m: no mm oopsom .OMH 3m: x macaw outflow mo mommoso Hmoomdfioop Eopm mQOHpmHBQOQ mm CH mowproopod mcfipaon mo mflmzamcm mosmflpm>ll.m mqmuompo m madam smcaom omH 2m: mwccflm .Oflpmh mnma m 0p paw mo mmocpoow pom pmop ohmsomlano a one boners owccfid mom omH 3m: x wesam smpaow mo mommopo Hwoopgfioos Eonm mQOHpmHSQOQ mm poaoom tam ampsmama mo Apcmo mod Gav moccddohmll.m mqm p > 0.5 26 TABLE 5.--Chi-square test of goodness of fit to a 39:13: 9:3 ratio of green szyellow szgreen Pl: yellow P1 assuming independence. P8138? Pfigze Observed Expected 5:356 Green P2 693 675 0.A80 Yellow P2 211 225 0.871 Green Pl 186 156 5.769 Yellow P1 18 52 22.231 29-351 ____________________________._____________________ P < 0.0001 27 TABLE 6.——Components of Chi-square test for goodness of fit to a 39:13:9z3 ratio between petiole color and pinnae number. Chéazqggre df 823:;e Probability Petiole color 1 10.633 0.001 — 0.002 Pinnae number 1 0.028 0.95 — 0.90 Interaction (linkage) 1 18.690 less than .0001 28 is due to linkage. The high Chi—Square value for linkage suggests that linkage is present between petiole color and pinnae number. Leaf Shape The leaf shape of the two parents is shown in Figure A. The MSU 130 parent on the right is deeply toothed, while the Golden Plume parent is shallowly toothed. Visual inspection of the two leaf shapes suggested that there was a difference in the area per units of length of the leaf- let. For the leaf shape character regression analysis was used. Figures 5 and 6 show the scatter diagrams of the parental types and their regression equations. The best equations in each case were the logarithmic functions ex- pressing the 10g of the length of the leaflet in milli- meters as a function of the log of the leaf area in square centimeters. The r—values of the lines are 0.87 and 0.8A for Golden Plume and MSU 130 respectively. The regression lines were tested for differences with a t—test using the formula of Steel and Torrie (53): t: 2 2 2 / Sp (i/zx + i/zxzj) lj where: b1 and b2 are the regression coefficients of lines 1 and 2 respectively. 29' Figure A.——Leaf shape of the two parental types. Left, Golden Plume (P2) shallowly toothed; right, MSU 130 (P1) deeply toothed. LOG LENGTH 30 |_J_L__ .05 1.70 1.7:: 1.80 1.35 Inn 1.95 2.00 2.05 L_OC3 AFWEA GP LI: SHAPE Figure 5.——Scatter diagram of leaf shape data, regression line (y = —2.67A2 + 2.2602x), and 1% confidence belts about the regres- sion line for the Golden Plume parent. 2. 10 LOG LENGTH .00 .QO .80 .70 .b0 1 .50 .uO .30 31 .J 1.. l l l L l 4—1—4 .J 1 J l LAP‘I- 1.b5 1.70 1.75 1.80 1.85 1.50 1.=5 2.00 2.05 2.10 LOO AREA MSU 1E0 LI: SHAPE Figure 6.——Scatter diagram of leaf shape data, regression line (y = —2.0l32 + 1.87l8x), and 1% confidence belts about the regres— sion line for the MSU 130 parent. SS is the weighted average of the error sum of squares for each line divided by the sum of the error degrees of freedom. 2x2. and 2x2 are the sum of the squared l.) 23' deviations from the mean of the_x variable for lines 1 and 2 respectively. The t—value was 7.27 which is significant at the 0.01 level indicating that the regression lines are different. Con- fidence belts about the regression lines were determined using the formula of Steel and Torrie (53): — /i (x — If y + bx i tOLSyQC H T where: y + bx is the regression line. t is Student's t for n-2 degrees of freedom. Syox is the square root of the error mean square of the regression line. X is some value of the horizontal variable. E is the mean of the horizontal variable. 2x2 is the sum of the squared deviations from the mean. Division of the F2 into classes was accomplished by use of the parental regression lines and their confidence belts. Figures 7 and 8 show the scatter diagrams of the two F2 populations. The pooled F population was divided 2 into the various classes as shown in Figure 9. The number LOG LENGTH 33 2.00 .00 1 F 2 LEAF SHAPE Figure 7.——Scatter diagram of leaf shape data for 980 (MSU 130 x Golden Plume) and parental regression lines. LOG LENGTH LED 1.55 1.70 1.75 1.33 1.35 1.‘=D 1.55 2&1) 2.5 2. I: .2 LEAP SHAPE Figure 8.-—Scatter diagram of leaf shape data for 981 (Golden Plume x MSU 130) and parental regression lines. LOG LENGTH LEAP SHAPE Figure 9.——l% confidence belts for both parents showing five leaf shape classes in the F2 population. (A11 observations in overlapping confidence belts--shaded area——were dropped). of‘F2 individuals falling in each class is shown in Table 7. The F populations were tested for homogeneity prior 2 to pooling the data. A bimodal distribution was obtained when the class values were plotted against the frequencies in each class. The dividing point or separation class appears to be at class 3 since it contained the smallest number of individuals. Classes 1, 2, and half of class 3 are of the P phenotype (MSU 130) and classes A, 5, and 1 half of class 3 are of the other parental phenotype (Golden Plume). The division intd two parental phenotypes sug- gested a segregation characteristic of a single gene pair with a probability of 0.20 - 0.15, suggesting an accept— able fit. This character will be designated d2 for deeply toothed; the other allele 2:, is the dominant form which results in shallowly toothed leaves. _Further genetic analysis of leaf shape data was made by means of discriminant analysis. Webber (69, 70, 71) suggested the use of discriminant analysis when the trait in question was complex and the character could best be described by more than one measurement. Since leaf shape cannot be described by a single measurement, it was felt that discriminant analysis might be useful. The method used was described by Miller (31). The discrimi- nant function used was X = .07107 a + .A9l21 b — .86813 c, where: TABLE 7.--Chi-square test for goodness of fit for pooled F2 populations from reciprocal crosses of MSU 130 x Golden Plume based on a one factor pair hypothesis for leaf shape. Class Observed Leaf type Expected SSEAPe 1 52 2 10A 169 Pflfl) 185.25 3 26 A 230 572 P2(2£) 555.75 0.A75 1.900 5 329 0.20 > p > 0.15 a is the dry weight of the leaflet in milligrams. b is the length of the leaflet in millimeters. c is the area of the leaflet in square centimeters. Application of this function to the parental and F2 data gave the relative frequencies as shown in Figure 10. The frequency distribution of the F2 populations showed a bimodal character when plotted against the discriminant values (X). This dip occurred at a value where the paren— tal distributions showed a dip. The dip in the parental distributions occurred between 18 and 19. The separation point between 18 and 19 gave a 181:562 ratio in the F A 2' Chi—square of 0.165 with a P value of 0.70 - 0.60 gave a good fit to a 3:1 ratio. Independence tests for leaf shape as determined by regression analysis showed that petiole color and vernali— zation response were independent of leaf shape. 39 F 2 ’ hIIIIII—T—VT—TWTT'T—WWWH— .tItI .0 DUDDDUUUUUHUUUHHHH >_ lib. ‘ -io ‘ -5‘.o 000 500 100 15‘ U + I I ' I ’ I I I LE ._ a E % E MSU 1E0 L E 1L1 I > C 111 0+ I (:1 Li _I :0 L 11111 J J Li- L, -1 __J .a o o oo 5.00 10 0 15 M l I I 0 -_LS -10. -5.0 0.CD 5.CD 10.0 15. DISCHIMIN Figure 10.——Relative frequency of MSU and pooled F2 from the re Golden Plume x MSU 130 in value classes. SUMMARY 1. The progenies of reciprocal crosses between Golden Plume and MSU 130 together with the parents were evaluated to determine the mode of inheritance of vernali— zation response, petiole color, pinnae number, and leaf shape. 2. The "easy" bolting condition of the Golden Plume . was effected by one major dominant gene, designated Er for vernalization response, and minor genes affecting the ac- tion of the major gene. Cytoplasmic factors were also present. 3. Yellow petiole color was conditioned by a single gene which was recessive to green petiole, as previously reported. The factor for petiole color was independent of that for vernalization response. 4. Two genes controlling pinnae number were estab- lished which segregated in a 13:3 ratio. Six to eight pinnae is dominant to 10 to 12 pinnae. Pinnae number was independent of vernalization response but is thought to be linked with petiole color. 5. Leaf shape was conditioned by a single gene pair when the data were analyzed by either regression analysis “0 41 01‘ discriminant analysis. This gene pair was designated. §§_for deeply toothed. Shallowly toothed leaves (23) are dominant to deeply toothed. Leaf shape as determined by regression analysis was independent of vernalization re- sponse and petiole color. THE INHERITANCE OF SEVERAL CHARACTERISTICS IN CELERY AND TOMATO II THE INHERITANCE OF FRUIT SHAPE, BEAKED FRUIT, SEPAL LENGTH, SEPAL SHAPE, AND PEDICEL LENGTH IN TOMATO (Lycopersicon esculentum Mill) INTRODUCTION Mechanical harvesting of tomatoes is increasing each year. One of the problems associated with mechanical har— vesting is fruit damage. One type of fruit damage is punctured fruit which may occur when a pedicel from one fruit punctures another fruit. These punctures are sites for rot infestation and, if soiled, are difficult to wash, thus increasing the bacterial and fungal content of the processed product. Puncture damage can be eliminated by breeding vari- eties that are free of pedicels when the fruits are har- vested. The separation of the pedicel from the fruit at harvest can be effected when one of the jointless genes is present in the variety. The jointless characters in existent stocks are often associated with undesirable characters such as reduced yield and poor quality. An— other more serious problem is that the fruit is often very difficult to separate from the pedicel. Short pedicel length may help solve the fruit punc- ture problem. If the pedicels were short (i.e., below the shoulders of the fruit), they would avoid puncturing other fruits. The fruit would also separate easily from the 43 A4 IiLants. For the fresh market, the calyx and short pedicel Would add to the attractiveness of the harvested fruit. In the Cleveland, Ohio and Grand Rapids, Michigan area, the greenhouse growers are clipping the pedicel and using the calyx as a "trademark" for greenhouse grown tomatoes. The objectives of this study were: 1. To identify the factors controlling the inheri- tance of pedicel length, sepal length, sepal shape, and fruit shape. 2. To establish linkage relationships between the above characters and with factors conditioning beaked fruit. REVIEW OF LITERATURE Genetically, the tomato is more widely studied than any other horticultural species. In spite of this wealth of information, there are few genetic studies of sepal shape or length. Young and MacArthur (77) described a gene affecting sepals which they designated macrocalyx, mg. The recessive allele expresses itself as very large sepals, almost the size of leaflets. Young (75) des— cribed a character called fleshy calyx, fl, which ex- presses itself in the recessive condition as fleshy calyx bases. Rick and Robinson (44) reported that Cl-l (cleis— togamous flowers) or a closely linked allele affected sepal length; homozygous Cl-l produces longer sepals than homozygous cl-l or the heterozygote. Reynard (39) noted differences between tomato lines for pedicel length. He reported tomato lines with pedicel lengths ranging from 9 to 22 millimeters. Although no genetic studies were reported, he suggested the symbol lg to designate long pedicel. Culp (ll) found pedicel length was controlled by a single gene in castor beans, Ricinus communis L. us Several investigators have attempted to determine the mode of inheritance of fruit shape in tomato. Yeager (7“) reported the gene lg for locule number, which af— fected fruit shape. Few (two or three) locules was domi— nant to many and in the presence of 9 produced round fruit. lg and g together produced oval fruit and g and lg produced oblate or flat fruit. Yeager postulated that the g and lg combination should produce round fruit; how- ever, he did not observe plants of this genotype in his investigations. Young and MacArthur (77) reported the combination of g and lg produced a giant plum-shaped phenotype. Dennett and Larson (14) concluded from their data that fruit shape was determined by a single gene pair designated as Ql. Non-ovate fruit was dominant over ovate fruit. The allele, gl, or a very closely linked gene also caused a marked increase in locule number. In their studies a non—ovate, few loculed or an ovate, many loculed plant was not observed. Warren (68) investigated fruit shape in tomatoes and found that two complementary genes were responsible for the expression of round vs. oblate fruit shape. He concluded that A_B_ gave a deep fruited phenotype while A_bb, aaB_, and aabb all produced shallow fruited pheno- types. Currence (12) showed 0 and g (compound inflores— cence) were linked with a cross-over value of 20.0%. “7 Yeager (74) reported lg to be located 20.5 cross—over units from the g locus on the opposite end of the chromo— some from the o locus. Young and MacArthur (77) reported a recessive gene, gg, which expresses itself as a sharp beak on the blossom end of the fruit. Boswell (U) discovered a gene with similar effect, g, for nipple-tip fruit. This gene is inherited as a simple recessive to normal fruit. There are many instances of recessive and dominant mimics in the literature of tomato genetics. Rick (41) reported a dominant gene, fD, which causes fasciated fruit to be allelic to f which causes fasciation in the reces- sive condition. The gene, V-3, which causes virescence, is inherited as a dominant character (27), while v—2 causes virescence in the recessive condition (25). Crn-l and crn—2, crimson l and 2, have similar effects on fruit color but are not inherited in the same way (6). MATERIALS AND METHODS Plants with short and long pedicels and sepals and contrasting sepal shapes and fruit characteristics were derived from a cross between VFlB—L,,VFlu5—B, and a MSU jointless (j-l) line. The parents were selfed and the re- sulting plants were grown to establish uniformity. Plants from uniform lines were selfed again prior to using them for hybridization in the greenhouse. Seeds of both parents, the Fl’ and F2 were planted in vermiculite. When the cotyledons had fully expanded, the seedlings were transplanted into flats. The treatment contained 70 plants of both parents and the F1 and 280 plants of the F When the plants reached an appropriate 2. size, they were transplanted to the field. During the growing season, the plants were fertilized and irrigated as necessary. Data were taken in the following manner: Two fruit clusters were removed from each plant; at least one fruit and, whenever possible, two mature green fruits were sampled from each cluster; the length, equatorial dia- meter (to be called breadth for the remainder of this thesis), pedicel length, and fruit characteristic (beaked U8 or non—beaked) were recorded for each fruit. The means of two to four observations from each plant were used for all calculations. A calyx, considered to be typical of the plant, was pressed in an herbarium press. When the calyx had dried, the samples were measured and scored for acuminate or ensiform sepal shape. The mean of two sepal measurements was used to describe sepal length. All data were punched on IBM cards for statistical analysis. Observed fre- quencies were compared to theoretical frequencies by means of the Chi—square test. RESULTS AND DISCUSSION Fruit Shape Parental and F fruit shapes are illustrated in 1 Figure ll. The P1 parent mean length:breadth ratio was 2.08 i 0.253 while the P2 parent mean ratio was 0.99 t 0.039. The F1 mean ratio was 1.52 i 0.007 and the F2 mean ratio was 1.52 i 0.358. The frequency distribu- tions for these populations may be seen in Figure 12. Hypotheses on the inheritance of fruit shape were based on the following observations: 1. The F1 mean was approximately midway between the parental means (1.5“ calculated, 1.52 ob- served). 2. The distribution of the F1 slightly overlaps the distribution of the long-fruited parent (Pl). 3. Fruit shapes occurred in the F2 which were not present in the P P2, or the F1 populations. 1’ The F2 population was divided into classes accord— ing to the distributions of the P P and F . All 1’ 2’ 1 plants with length:breadth ratios of 0.90 to 1.05 were considered class 1 or the P2 phenotype. Plants whose 50 Figure ll.——Parental and F1 fruits, calyxes, and pedicels upper left P , upper right P2, and lower center Fl (see text for fruit length: breadth ratios). “I RELATIVE FREQUENCY Fi 52 ULUUUUUUU UUUUUUUUU UUUUUUU UUUUUUUUEL ME .513 . 270 PET.) F l E‘ J J._.,_J RHUHHHHHHUJ J .J ._J J J J l J ”J _J._+ :0 l lo 1.30 1.20 l 70 ‘43 3C) .0 .70 2.% P 2 P l m UEE U UUUUUUUUUUUUUEUUE 0.50 1.10 1.33 LENGTH Z COBREADTH aoRATlO PRU I T SHAPE gure l2.——Relative frequency of fruit shape ratios for parental, F1, and F2 populations. 53 ratios were 1.10 to 1.35 are classified class 2, a class not represented by the P1’ P2, or F1 populations. Plants with ratios of 1.40 to 1.75 are classed in class 3 or the F phenotype. Plants with ratios greater than 1.80 were 1 placed in class A, the Pl phenotype. In the F the mean length:breadth ratio of 1.01 t 2’ 0.042 for class 1 did not differ significantly from the value of 0.99 i 0.039 for the P2 population (t = 2.09). The length:breadth ratios 1.21 t 0.07“ for class 2 and 1.53 i 0.108 for class 3 did not differ significantly from the F ratio of 1.52 i 0.007 (t = 0.538). The ratio 1 of 1.97 i 0.227 for class A did not differ significantly from the P ratio of 2.08 i 0.253 (t = 2.50). 1 It appears that two genes are responsible in deter- mining fruit shape and that the Pl parent carries the re— cessive alleles at both loci, while the P2 parent is domi- nant at both loci. The fruit shape genes are designated as orf and Lgrf. Postulated genotypes and phenotypes are listed below: gUfgfflgrfggrf round fruit, P2 phenotype, class 1 g 9 lg lg long fruit, Pl phenotype, class A _rf2 lgrflg intermediate fruit shape, Fl pheno— type, class 3 This leaves two genotypic classes; those with one dominant allele and those with three dominant alleles. Based on the intermediate length:breadth ratios of the F it is postulated that the genes operate additively l, and show no dominance when zero, one, or two dominant alleles are present. Since the distribution of the F1 slightly overlaps the Pl parent distribution, it is pos— tulated that three and four dominant alleles behave pheno- typically alike and make up class A. Table 8 shows a poor fit to a two gene additive model for fruit shape and suggests the presence of link- age. Since the two postulated genes are indistinguishable phenotypically, the standard method for determining link— age was not used. Linkage was determined in the following manner . The genotypes of the two parents were 0 lc /o 10 (long fruit) and orflgFf/grchrf (round fruit). The F — —— 1 genotype g lg /grfgg?f, at meiosis produces four types of gametes. The frequency of each of the gametes when linkage is present is influenced by the strength of the linkage. These frequencies are 1/U (1 + L), where L is a term denoting the strength of linkage, for the parental type gametes. The recombinant type gametes have a fre- quency of l/U (1 — L). The F2 phenotypic classes, pos- tulated genotypes, and the zygotic frequency of each pheno— type in terms of L may be seen in Table 9. Combining the zygotic frequencies for each phenotypic class, the ex- pected frequency for each class (illustrated in Table 10) is obtained. Equating the expected frequencies to the 55 TABLE 8.--Chi-square test for goodness of fit to a two gene additive model for fruit shape assuming independence. Class Frequency Observed Expected s23:;e 1 1/16 25 15.6 5.6614 2 4/16 59 62.5 .196 3 6/16 95 90.8 .194 4 5/16 71 78.1 .644 676—919 0.10 < p < 0.08 TABLE 9"‘F1 gametic frequency and F2 genotypes, pheno- types, and zygotic frequency for fruit shape.- Phenotypic Class Genotype 10 1c Lc Lcrf/o rf lc LCrf Lcrf/orflc Lcrf/o Lc Gametic Frequency 1/4(1 + 1/4(1 l/4(i - 1/4(1 + Zygotic Frequency [1/u(1 + 1.)]2 [1/u(1 + L)]2 l/4(1 + L)l/4(1 l/4(l + L)l/4(1 [1/u<1 + L)]2 [1/u(1 + 1.)]2 1/4(l + L)l/4(l l/4(l + L)l/4(L 1/4(1 + L)1/4(1 l/4(l + L)1/4(l [1/u(1 — L)]2 [1/u(1 — L)]2 l/4(1 + L)l/4(1 1/4(l + L)1/4(l [1/u<1 — L)]2 [1/u(1 — L)]2 57 TABLE 10.——Observed and expected frequencies of fruit shape phenotypes when linkage is considered. Class Expected Frequency Observed Frequency 1 1/16(1 + 2L + L2) = .100 2 1/16(4 - 4L2) = .236 3 1/16(6 - 4L + 6L2) = .380 u l/l6(5 + 2L - 3L2) .28u EE observed frequencies yields the equation 1/16(1 + 2L + L2) = 0.100. The equations can be solved by use of the formula E—b : (b2 — 4ac)l/2l/ for solving quadratic equations: x 2a. Since there are four equations, four estimates of L are obtained. They are: for class 1, L = 0.265; class 2, L = 0.237; class 3, L = 0.020; and class 4, L = 0.180. The mean of these four estimates (0.176) was used as a starting point in calculating the expected frequency for each class and the Chi-square value was calculated. The value of L was increased or decreased and a new Chi- square value was calculated. The value of L was adjusted until a value of L was obtained at which the Chi—square would be lower than the Chi—square for any other value of L. This value was considered the best estimate of L. The best estimate of L for the fruit shape genes was 0.2, which gives the expected values as shown in Table 11. The Chi—square value has two degrees of freedom since one is attributed to linkage. The relationship between L and r, the recombination percentage, can be determined by comparing gametic fre- quencies. Bailey (1) gives the gametic frequencies of 1/2 (1 — r) for the parental type gametes and 1/2 r for the recombinant gametes. Therefore, 1/4 (1 + L) equals 1/2 (1 — r) and 1/4 (1 — L) equals 1/2 r. Thus, the rela— tion L = 1 - 2r between L and r is obtained. Substituting TABLE 11.——Chi—square test for goodness of fit to a two gene additive model for fruit shape assuming a linkage intensity value of 0.2. Class Observed 1 25 2 59 3 95 4 71 0.3 < p < 0.2 59 22. 60 82. Expected 5 5 Chi— square 0.278 0.017 1.176 1.603 3007“ 60 0.2 for L, a value of r = 0.40 is obtained which is approx— imately the reported distance between 0 and lg. Beaked Fruit Young and MacArthur (77) reported beaked fruit to be conditioned by a recessive gene designated gg. This character appears as a sharp point or beak on the blossom end of the fruit. The data in this study suggest that beaked is controlled by a single dominant gene which is designated Bk—2. Attempts were made to locate Bk—2 on the chromosome map with orf and Lgrf. Since orf and lgrf are phenotypi— cally indistinguishable, the standard methods could not be applied. The analysis was accomplished using the model f on Table 12. The parental genotypes are g? lgffbk-2/gff f Lcr orchrfbk—2/g lc Bk—2 and produces eight types of bk—2 and 0 1c Bk-2/g lc Bk—2. The F1 genotype is gametes with the frequencies illustrated in Table 12. L1 is designated as the linkage between 0 and lg and L2 the linkage between ggzg and the position midway between the loci of g and lg. The linkage between beaked and g and beaked and lg is difficult to determine since the indi— vidual effects of g and lg cannot be measured. Substituting 0.2 for L1’ the gametic frequencies as sfnown in the right column of Table 12 are obtained. The eiEght gametes recombine to produce 64 possible zygotes Z211d. eight phenotypic classes (each of the four fruit shape 61 TABLE 12.-—Gametic types and their frequencies for fruit shape and beaked fruit. Gametes Frequencies Frequencies 0 LC Bk-2 l/8(l + Ll + L2) 1/8(l.2 + L2) 0 Lcrka-2 1/8(1 — Ll — L2) 1/8(.8 — L2) rch Bk-2 1/8(1 — Ll + L2) l/8(.8 + L2) orchrka—2 1/8(1 + Ll — L2) 1/8(1.2 — L2) 0 Lc bk-2 1/8(1 + Ll — L2) 1/8(1.2 — L2) o Lcrfbk-2 1/8(1 - Ll + L2) 1/8(.8 + L2) orch bk—2 1/8(1 — Ll - L2) l/8(.8 - L2) orchrfbk-2 1/8(1 + Ll + L2) 1/8(1.2 + L2) 62 classes is divided into beaked and non-beaked fruit). Summing the frequencies of each class and solving for L2 as in "Fruit Shape” gives an estimate for linkage between Bk—2 and the position midway between the.loci of gff rf Lc . Table 13 shows the eight phenotypic classes, the and observed and expected frequencies, and Chi-square test. A linkage value of 1.0 suggests that Bk-2 is linked to the position midway between the loci of or.f and Lgrf, or 20 recombination units from orf on one side and Lg?f on the other. This corresponds closely to the position of g5 from o and lg, as shown on the tomato chromosome map (9). This information suggests that gg may be allelic to Bk—2, 0 may be allelic to orf, and lg may be allelic to Lcrf. The effect of Lgrf and lg genes on locule number was not studied in the F2, although the Lcrchrf parent (P2) had few locules (2-3) and the lg lg parent (P1) had many (4 or more). Sepal Length The frequency distributions for sepal length of the parents, F1, and F2 generations are shown in Figure 13. The distribution of the sepal length of the F1 was inter- mediate to the two parents. The F2 was divided into three classes of 16 — 20, 21 — 27, and 28 — 40 mm corresponding tC) P F1’ and P phenotypes, respectively. A 61:124z65 2’ 1 )réirtio of shortzintermediatezlong sepals was obtained which TABLE l3.—-Chi—square test of goodness of fit to a three gene model for fruit shape and beaked fruit with a linkage intensity value of 1.0. Fruit shape Class ESL/UUU 0.6 < p < 0.5 Beaked Observed 21' 27 32 81 14 68 Expected 3.6 18.9 32.5 27.5 71.6 13.4 79.85 2.65 64 F' 2 .111 213111.1le.. > Lb.0 1H?) 21.0 23.") 31.0 33.5 32.0 35.5 h1.0 D Z 0 I 0 F 1 L E: u. 01 > H q -. U UUUU - EEE _] Lb.0 155.5 21.0 23.5 22.0 285 31.0 33.5 $.0 38.5 L¢1.0 0] GI P 2 p 1 EE L41.D U U...UU;U.Ufl UU .r. UUUH LL20 SEPAL LENGTH Figure l3.-—Relative frequency of sepal lengths (in mm) of parental, F1, and F2 populations. 65 gave a Chi—square value of 0.144 and a P value of 0.95 — 0.90. The ratio obtained suggests.that a single gene pair with incomplete dominance was responsible for sepal length. This gene is designated §x for sepal extender. The geno- type §x §x produces long sepals; §_ §_ produces short' sepals; and EE sx produces intermediate sepals. The Chi-square test of an independent gene model for beaked fruit and sepal length gave a Chi—square value of 10.882 with a P value of 0.05 — 0.03. The low probability suggested linkage. Linkage intensity was determined by the method as outlined under "Fruit Shape." The best estimate of L was 0.212, which was used to calculate the expected frequencies illustrated in Table 14. This link- age intensity value corresponds to a recombination value of 39.4%. The linkage intensity between sepal length and round fruit was determined similarly to ”Beaked Fruit" except that the linkage was estimated from §x to the nearest fruit shape gene. Table 15 shows the Chi—square test for a one gene additive model for fruit shape with a linkage factor of 0.52. This linkage intensity value corresponds to a recombination value of 24%. §gpal Shape The sepal shapes of the parents and Fl may be seen in Figure 14. The top row depicts the recessive condition, called ensiform or sword—like, designated as 22. The 66 TABLE 14.-—Chi-square test for goodness of fit to a 3:6:3 1:2:1 ratio of sepal length and beaked fruit with a linkage intensity value of 0.212. Short Intermediate Long Short Intermediate Long 0.8

50 195 1.35 2 1 . 1.1..... E E 7 . 1 1 g > 7. E 1:. - 1 11 . 1 1,111,111. . . . 111111111 ”111111111; PEDICEL LENGTH Figure 15. ——Relative frequency distributions of pedicel length for parental, F1’ and F2 populations. 79 l+ to be polygenic. The mean length of the F1 is 12.99 0.810 which is approximately midway between the parental means of 7.20 t 0.525 and 19.7 i 2.205. The mean pedicel length of the F is 12.28 i 2.809 and does not differ 2 significantly from the mean of the Fl (t = 1.952). The correlation coefficients between pedicel length and the characters described above were calculated to de— termine if pedicel length was associated with any of the characters. These coefficients are illustrated in Table 22. All correlation coefficients are significant at the 0.005 level. It may be noted that the magnitude of the correlation coefficients is in the following order: Sepal length, §£3 sepal shape, en; fruit shape and beaked fruit, Bk—2. This is the order postulated from the recombination data. 80 TABLE 22.——Correlation coefficients between pedicel length, sepal length, sepal shape, fruit shape, and beaked fruit. Beaked fruit 0.5u8 —0.l92 0.181 0.266 Fruit shape ~0.360 0.370 0.408 Sepal shape -0.562 -0.H38 Sepal length 0.5u5 __—______—___—__.—_———————————— All coefficients significant at 0.005 level. SUMMARY 1. The progenies of two genetic lines MSU A7 and MSU B12 together with the parents were evaluated to de— termine the mode of inheritance of fruit shape, beaked fruit, sepal length, sepal shape, and pedicel length. 2. The character fruit shape was conditioned by two genes, designated orf and Lcrf , which may be allelic to o and lg, respectively. These genes behaved in the following way: orf rf r f 9 £3 fggr produces round fruit (length:breadth ratio = 1.01 i 0.042). lo and o grfggrfggrf produces slightly oval fruit (length:breadth ratio = 1.21 i 0.07M). -rf_ Egrfig , grfgrflg l8: and g 9 ESrfEE-rf produces an oval fruit (length:breadth ratio = 1.53 i 0.108). orfo lc lc , 9 9 lc Lcrf, and o 0 lg 10 produce a long fruit (length:breadth ratio = 1.97 i 0.227). The two fruit shape genes were linked with a recombi- nation value of H0%. 81 82 3. The character beaked fruit.was conditioned by a single dominant gene, designated Bk-2, and may be allelic to bk. Bk—2 is postulated to be located approximately midway between the loci of orf and ggff. 4. Sepal length was controlled by a single gene pair with no dominance. §§, for sepal extender, when homo- zygous, effects a long sepal (30.60 t 3,075 mm); s5, when homozygous, effects short sepals (19.36 t 1,808 mm); and 1+ §£ sg effects an intermediate sepal length (25.24 2,112 mm). This character was linked to Bk—2 with a recombina- tion value of 39.M% and to the nearest fruit shape gene with a recombination value of 2Mo 5. Sepal shape was determined by a single gene pair with acuminate sepal shape being dominant to ensiform or sword-shaped sepal. This character was designated ego En was linked to the nearest fruit shape gene with a recombi- nation value of 20%, with beaked fruit Bk-2 (recombina- tion value 3U%) and with ii (recombination value 20%). 6. The character pedicel length was inherited poly— genically. Correlation values suggest some of the genes for this character might be linked to the genes above. CONCLUSIONS Several important characters were shown to be heri— table. Knowing the inheritance of bolting would be help- ful in a breeding program since bolting resistance in celery is an important horticultural character. Leaf shape, pinnae number, and petiole color in celery are not associated with bolting. Fruit shape and pedicel length in tomato are im- portant horticultural characters, and knowledge of their inheritance and linkage relationships is useful in a breeding program. Presently sepal size and shape are not economically important characters. If short pediceled varieties are introduced, the shape and size of the calyx would become more important and knowledge of their inheri- tance and linkage relationships would be useful. In addition to the practical importance of the infor- mation in this thesis there is also a contribution to the knowledge about the species. lO. BIBLIOGRAPHY Bailey, N. T. J. 1961. Introduction to the mathe— cal theory of genetic linkage. Clarendon Press, Oxford. 298 p. Barber, H. N. 1959. Physiological genetics of Pisum. II. The genetics of photoperiodism and vernalization. Heredity 13:33-60. Barendse, G. w. M. 1965. Vernalization in Cheiran- thus allionii. Hort. Mededel Landbouwhogeschool Wageningen 65(1u):1-6A. Boswell, V. R. 1937. Improvement and genetics of tomatoes, peppers, and eggplant. U. S. Dept. Agr. Yearbook: 176-206. Butler, L. 1951. New linkage groups in tomato. J. Hered. A2leO-104. . 1962. Crimson: a new fruit color. Tomato Genet. Coop. Rep. 12:17—18. Chouard, P. 1960. Vernalization and its relations to dormancy. Ann. Rev. Plant Physiol. 11:191- 238. Chroboczek, E. 193“. A study of some ecological factors influencing seedstalk development in beets (Beta vulgaris L.). New York (Cornell) Agr. Exp. Sta. Mem. 15A. Clayberg, C. D., L. Butler, E. A. Kerr, C. M. Rick, and R. w. Robinson. 1966. Third list of known genes in the tomato with revised linkage map and additional rules. J. Hered. 57: 189—196. Cooper, H. P. 1923. The inheritance of spring and winter growing habit in crosses between typical spring and winter wheats and the response of wheat plants to artificial light. J. Amer. Soc. AgrOn. 15:15—24. 8A 11. Culp, T. W. 1966. Inheritance of capsule drop resis- tance and pedicel length in Castor beans Ricinus communis L. Crop Sci. 6:280-283. 12. Currence, T. M. 1938. The relations of the first chromosome pair to the data of fruit ripening in the tomato. Genetics 23:1—11. 13. Curtis, D. F., and H. T. Chang. 1930. The relative effectiveness of the temperature of the crown as contrasted with that of the rest of the plant upon the flowering of celery plants. Amer. J. Bot. 17:10A7-1OA8. 14. Dennett, R. K., and R. E. Larson. 1953. Inheritance of fruit shape, locule number, and stem color in derivatives of a hybrid of Lycopersicon. esculentum and L co ersicon peruvianum. Penna. Agr. Exp. Sta. Bull. 56341-30. 15. Elmsweller, S. L. 193A. Premature seeding in inbred celery lines. Amer. Soc. Hort. Sci. Proc. 31: 155—159. 16. Gaines, E. F. 1917. Inheritance in wheat, barley, and oat hybrids. Wash. Agr. Exp. Sta. Bull. 135. 17. Hall, A. Do 1928. Bateson's experiments on bolting in sugar beets and mangolds. J. Genet. 20:219- 231. 18. Hammond, D. 1941. The expression of genes for leaf shape in Gossypium hirsutum L. and Gossypium arboretum L. . The expression of genes for leaf shape in Goss ium hirsutum L. Amer. J. Bot. 28:12A-138. l9. . 1941. The expression of genes for leaf shape in Gossypium hirsutum L. and Gossypium arboretum L. II. The expression of genes for leaf shape in Gossypium arboretum L. Amer. J. Bot. 28:138—150. 20. Hillman, W. S. 1963. The physiology of flowering. Holt, Rinehart, and Wilson, New York. 164p. 21. Honma, S. 1959. A method for celery hybridization. Amer. Soc. Hort. Sci. Proc. 73:345-3A8. 22. Imai, Y. 1930. A genetic monograph on the leaf form of Pharbitis nil. Zeit. Induc. Abstam. Verer. 55:1—107. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 35. 86 Immer, F. R. 1930. Formulae and tables for calculat— ing linkage intensities. Genetics 15:81-98. Joyner, J. F. and J. B. Pate. 1956. Inheritance of leaf shape in kenaf. J. Hered. 47:199-204. Kerr, E. A. 1962. Another virescent gene. What shall we call it? Tomato Genet. Coop. Rep. 12: 30. Lang, A. 1952. Physiology of flowering. Ann. Rev. Plant Physiol. 3:265-306. Lesley, J. W. and M. M. Lesley, 1965. A dominant virescent, V-3.. Tomato Genet. Coop. Rep. 15: uo-Ll'l a Lorenz, O. A. 1946. Response of Chinese cabbage to temperature and photoperiod. Amer. Soc. Hort. Sci. Proc. 47:309—319. Mathan, D. S. and R. D. Cole. 1964. Comparative bio— chemical studies of two allelic forms of a gene affecting leaf shape in tomato. Amer. J. Bot. 51:560-566. . and J. A. Jenkins. 1962. A morphogenic study of lanceolate, a leaf-shape mutant in tomato. Amer. J. Bot. 49:504-514. Miller, R. G. 1962. Statistical prediction by dis- criminant analysis. Meteorl. Monogr. 4(25): 1—54. Napp—Zinn, K. 1961. Vernalisation and verwandth Erscheinungen. Handbuch der pflanzenphysiologie. Berlin, Springer XVI:24—75. Nilsson-Ehle, H. 1917. Selection of spring wheat in Sweden. Sveriges Utsdeforenings Tidskrift 28:51 (Abstr. in Int. Rev. Sci. and Practice in Agri. 1917). Owen, F. V. and J. S. McFarlane. 1958. Successive annual backcrosses to a nonbolting inbred line of sugar beets. J. Amer. Soc. Sugar Beet Tech. 10:124-132. Pawar, S. S. and H. C. Thompson. 1950. The effect of age and size of plant at the time of exposure to low-temperature on reproductive growth in celery. Amer. Soc. Hort. Sci. Proc. 55:367—371. 36. Potter, G. F. 1937. Photoelectric measurements of apple leaf areas. Amer. Soc. Hort. Sci. Proc. 35:260-262. 36a. Powers, L. 1942. The nature of a series of envir- onmental variances and the estimation of the genetic variances and the geometric means in crosses involving species of Lycopersicon. Genetics 27:361—375. 37. Purvis, O. N. 1939. Studies in vernalization of cereals. V. The inheritance of the spring and winter habit in hybrids of Petkus rye. Ann. Bot., N. S. 3:719—729. 38. . 1940. Vernalization of fragments of embryo tissue. Nature 145:462-467. 39. Reynard, G. B. 1960. Long pedicels. Tomato Genet. Coop. Rep. 10:29—30. 40. Rhyne, C. L. 1960. Leaf lobation in Gossypium amphidiploids as determined by leaf shape genes of Gossypium diploid species. Genetics 45: 63-7 0 41. Rick, C. M. 1965. A dominant allele at the f locus. Tomato Genet. Coop. Rep. 15:50. 42. . and A. L. Harrison, 1959. Inheritance of five new tomato seedling characters. J. Hered. 50:90-98. 43. . and J. Robinson. 1951. Inherited defects of floral structure affecting fruitfulness in L co ersicon esculentum. Amer. J. Bot. 38: 539—252. 44. Robinson, R. W. and C. M. Rick. 1954. New tomato seedling characters and their linkage relation- ships. J. Hered. 45:241-247. 45. Ryder, E. J. 1965. Inheritance of five leaf char- acters in lettuce (Lactuca sativa L.). Amer. Soc. Hort. Sci. Proc. 86:457-461. 46. Salisbury, F. B. 1964. The flowering process. Per— gammon Press, Oxford. 234p. 47. Sarkar, S. 1958. Versuche zur physiology der ver- nalisation. Biol. Zbl. 77:1-49. 88 48. Sinnott, E. W. 1935. Evidence for the existence of genes controlling shape. Genetics 20:12-21. 49. . 1936. A developmental analysis of inherited shape differences in cucurbit fruits. Amer. Nat. 70:245-254. 50. Smith, 0. 1939. Premature seeding studies with beets. Amer. Soc. Hort. Sci. Proc. 37:793— 798. 51. Starring, C. C. 1924. Premature seeding of celery. Montana Agr. Exp. Sta. Bull. 168. 52. . 1934. The premature seeding of celery a second report. Montana Agr. Exp. Sta. Bull. 292. 53. Steel, R. G. D. and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw—Hill, New York. 481p. 54. Stephens, S. G. 1944. The genetic organization of leaf shape development in the genus Gossypium. J. Genet. 46:28-51. 55. . 1945. A genetic survey of leaf shape in New World cottons. A problem in critical identification of alleles. J. Genet. 46:313— 330. 56. . 1945. The modifier concept. A develop— mental analysis of leaf shape "modification" in New World cottons. J. Genet. 46:331—344. 57. . 1945. Canalization of gene action in the Gossypium leaf shape system and its bearing on certain evolutionary mechanisms. J. Genet. 46:345—357. 58. . 1945. Some observations on leaf shape ex- prission in the Malvaceae. Amer. Nat. 79:380- 38 . 59. Settler, R. F. 1964. Dosage effects of the lanceo— late gene in tomato. Amer. J. Bot. 51:253-264. 60. Suttom, E. P. F. 1924. Inheritance of "bolting" in cabbage. J. Hered. 15:257-260. 61. 62. 66. 70. 71. 72. Takahashi, N. 1924. On the inheritance of the spring versus winter form in barley (In Japanese, English summary) Jap. J. Genet. 3: 22— 28. Tedin, O. 1923. The inheritance of pinnatifid leaves in Camelina. Hereditas 4:59—64. Thompson, H. C. 1929. Premature seeding in celery. New York (Cornell) Agr. Exp. Sta. Bull. 480. 1939. Temperature in relation to vegeta- tive and reproductive development in plants. Amer. Soc. Hort. Sci. Proc. 37:672—679. Townsend, G. R., R. A. Emerson, and A. G. Newhall. 1946. Resistance to Cercospora apiia Fres. in celery (A ium raveloens var. dul ce Phy- topathology 36: 980- 982. Tschermak, E. 1906. Uber zuchtung neuer getrerder- assen mettles kunstlicher kreuzung, II. Kreuzungstudien am roggen. Zeits. Fd. Land. Versuchswesen in Osterreich 9:699-705. van Heel, J. P. D. 1927. Inheritance of bolting in sugar beet. Genetica 9:217—236. Warren, P. A. 1924. Genetic studies in Lycopersicon I. The heredity of fruit shape in the garden tomato. Mich. Acad. Sci. Arts and Letters. 4: 357—394. Weber, E. 1959. The genetical analysis of charac- ters with continuous variability on a Mendelian basis I. Monohybrid segregation. Genetics 44: 1131—1139. . 1960. The genetical analysis of characters with continuous variability on a Mendelian basis II. Monohybrid segregation and linkage analysis. Genetics 45:459-466. 1960. The genetical analysis of characters with continuous variability on a Mendelian basis III. Dihybrid segregation. Genetics 45:567-572. Wellensiek, S. J. 1965. Recent developments in ver— nalization. Acta Bot. Neer. 14:308-314. M 9O 73. Wester, R. E. and R. Margruder. 1937. Varietal and strain differences in bolting of turnips. Amer. Soc. Hort. Sci. Proc. 35:594—598. 74. Yeager, A. F. 1937. Studies on the inheritance and development of fruit size and shape in the tomato. J. Agr. Res. 55:141-152. 75. Young, P. A. 1952. New characters: green Jelly, white flesh, cold resistance, rooty stems, pink and gold fruits, and curled sepals. Tomato Genet. Coop. Rep. 2:12-13. 76. . 1955. Curl, a mutant tetratism of the tomato. J. Hered. 46:243—244. 77. . and J. W. MacArthur. 1947. Horticultural characters of tomatoes. Texas Agr. Exp. Sta. Bull. 698. ' 11111111):murmmmum111an , 31293101650186