-! A, xgfi-h.‘ .lr." "\ - J g’f‘ A V ' ‘ a) . 4 .. P. l -.>_. _ _.'_—.- \. _ «7 “' ~‘ hm" -"— m1. 0" This is to certify that the thesis entitled THE INHERITANCE OF CARPEL SEPARATION IN MATURE FRUITS OF PICKLING CUCUMCBER presented by Jill E . Wilson has been accepted towards fulfillment of the requirements for Ph . D. degree in Horticulture Date May 29, 1974 0-7639 1 ABSTRACT THE INHERITANCE OF CARPEL SEPARATION IN MATURE FRUITS OF PICKLING CUCUMBER By Jill E. Wilson Progenies of crosses between the pickling cucumber (Cucumis sativus L.) lines SC601H and MSU381 were evaluated to determine the inheritance of mature fruit carpel separa- tion. Reciprocal cross differences for carpel separation were not detected. Weak carpel suture strength, as expressed by carpel separation, exhibited dominance over non-separation. Additive genetic variance exceeded domi- nance variance, and a significant non-genetic or environ- mental component of variation was observed. Estimates of narrow-sense heritability ranged from 39 to “5% suggesting that carpel suture strength, and consequently fruit quality, could be improved through selection. Although a genetic model consistent with the observed frequency distributions was not found, the data can be interpreted as a 2-gene, or at the most a 3-gene, system with low heritability. High frequencies of carpel separation were not necessarily associated with gynoecious or high yielding phenotypes. THE INHERITANCE OF CARPEL SEPARATION IN MATURE FRUITS OF PICKLING CUCUMBER By Jill E. Wilson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Horticulture 197A ACKNOWLEDGMENTS The author wishes to express her appreciation to Dr. L. R. Baker for serving as her major professor and providing advice and direction throughout the course of her research and study. She is indebted to Dr. C. E. Peterson, presently with the U.S.D.A., for his guidance during the initial stages of her graduate training and research. Special thanks goes to Mr. Amos Lockwood of the Horti- cultural Research Center for his invaluable help planting and maintaining research plots. And my thanks also goes to all those student workers who aided in data collecting, and especially to Mr. David Jandik for his assistance with the computer analysis of that data. ii Guidance Committee: The body of this thesis has been condensed into a manuscript intended for publication in the Journal of the American Society_for Horticultural Science. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . vi ABSTRACT . . . . . . . . . . . . . 1 INTRODUCTION . . . . . . . . . . . 2 MATERIALS AND METHODS . . . . . . . . . 10 RESULTS AND DISCUSSION . . . . . . . . . l6 LITERATURE CITED . . . . . . . . . . . 29 iv Table 1. LIST OF TABLES Page Means, standard errors, and variances for mature fruit carpel separation ratings in the family SC60lH X MSU381G, preliminary study . . 1? Estimates of components of variation, degree of dominance, heritability, and probable num- ber of effective factors for mature fruit separation in the family SC601H X MSU381, preliminary and detailed studies . . . . . 18 Relative frequency distributions (%) for mature fruit carpel separation in cucumbers from the pooled family of SC601H X MSU381 compared to established varieties; detailed study . . . . . . . . . . . . . . 22 Means, standard errors, and variances for mature fruit carpel separation in cucumbers from the pooled family of SC601H X MSU381; detailed investigation . . . . . . . . 2“ Analysis of variance for mature fruit carpel separation in cucumber for parent and F1 generations of the pooled family SC60lH X MSU381; detailed investigation . . . . . . 26 LIST OF FIGURES Figure Page -1. Cross-sections of mature cucumber fruits exhibiting separation of the carpels . . . . 3 2a. Longitudinal sections of immature cucumber fruits showing carpel separation . . . . . 5 2b. Longitudinal sections of immature cucumber fruits showing no carpel separation . . . . 7 3. Typical family derived from cross of two single plants for use in the inheritance study . . . . . . . . . . . . . . 13 U. Frequency distributions (%) comparing carpel separation in mature fruits of cucumber plants with early immature sepa— ration ratings . . . . . . . . . . . 2O 5. Frequency distributions (%) for carpel separation in mature fruits of cucumber from the family of SC601H X MSU381 . . . . 23 vi ABSTRACT THE INHERITANCE OF CARPEL SEPARATION IN MATURE FRUITS OF PICKLING CUCUMBER By Jill E. Wilson Progenies of crosses between the pickling cucumber (Cucumis sativus L.) lines SC601H and MSU381 were evaluated to determine the inheritance of mature fruit carpel separa- tion. Reciprocal cross differences for carpel separation were not detected. Weak carpel suture strength, as expressed by carpel separation, exhibited dominance over non-separation. Additive genetic variance exceeded domi- nance variance, and a significant non-genetic or environ- mental component of variation was observed. Estimates of narrow—sense heritability ranged from 39 to “5% suggesting that carpel suture strength, and consequently fruit quality, could be improved through selection. Although a genetic model consistent with the observed frequency distributions was not found, the data can be interpreted as a 2-gene, or at the most a 3-gene, system with low heritability. High frequencies of carpel separation were not necessarily associated with gynoecious or high yielding phenotypes. INTRODUCTION Carpel separation in mature fruits of the pickling cucumber (Cucumis sativus L.) is expressed when the sutures of the 3 fused carpels separate forming a hollow or void which runs through part or the entire length of the fruit (Fig. l). Economically, this character is important because it is associated with carpel separation in the immature fruits used for pickling (Fig. 2). At present approximately “5% of the pickling cucumbers harvested in the USA is utilized immediately in various pasteurized, fresh pack products (11). Carpel separation lessens the consumer appeal of these products lowering their economic value. The remain- ing 55% is cured by fermentation in brine for later processing. Early workers (8) suggested a causal relation- ship between green stock characteristics and bloater forma— tion during curing of salt stock. Later, researchers showed that carpel suture strength was positively correlated with the brining quality of pickling cucumbers (7,15). As carpel suture strength increased, the frequency of green stock carpel separation decreased, and likewise the frequency of balloon bloater formation during fermentation decreased. It has been reported that bloater formation resulted in a 5 million dollar loss to the pickle industry in the 1971-72 packing season (11). Figure l. Cross-sections of mature cucumber fruits exhibiting separation of the carpels. Figure 2a. Longitudinal sections of immature cucumber fruits showing carpel separation. J ‘ hl§§5,h% I : 'luoiuit‘v‘f“ .9 I " (I, 2'. .5: ’ . -~’.J . 33;: g GREEN FRUIT: CARPEL SEPARATION Figure 2b. Longitudinal sections of immature cucumber fruits showing no carpel separation. _r - ,Pr‘fl D t ‘ 0 U (:33, GREEAI FRUIT: NO CARPEL SEPARATION The frequency of carpel separation is determined by a number of factors. It has been demonstrated that percent carpel separation was associated with length of processing time and delay of processing after harvesting of the raw product (13) and with the date of harvest (1“). Carpel separation was markedly increased with each step in mechanical harvesting and grading (9,10) and sometimes by increasing levels of impact (5). In each of the above studies, varietal differences in susceptibility to carpel separation were suspected or demonstrated suggesting genetic control. The present study was designed to determine the inheritance of carpel suture strength as measured by carpel separation in mature fruits. 10 MATERIALS AND METHODS It is difficult to determine objectively the degree of carpel separation in immature fruits suitable for proc— essing. The informal "thumb test", by which one estimates the force necessary to separate the carpels of a cross- sectional slice with the thumb, is too subjective. In contrast, the technique utilizing the Instron Universal Testing Machine [Hooper, g§_al. (6)], although sensitive and objective, cannot be conveniently used when large numbers of fruits must be evaluated. Therefore, in this study susceptibility to carpel separation was Judged by the degree of this defect in mature fruits. The use of mature fruits permitted full expression of the character being studied. The rating of up to 3 fruits per plant allowed sampling of fruits which developed under different environ- ments. This partly offset the observation that carpel separation ratings of fruits on the same plant may differ with environmental conditions during fruit development. Preliminary Investigation (1968). Two parent lines were used. SC601H (P1), a monoecious line developed at Clemson University, produced mature fruits which were pre- dominantly non-hollow with apparent resistance to carpel separation. The second parent, MSU38lG (P2), was a gynoecious line characterized by a high degree of separation 11 in both the immature and mature fruits. Seeds of the parents, F1’ F2, and reciprocal backcross generations were produced by hand pollinations in the greenhouse. Each generation was made up of seed bulked from several plants. All populations were field grown near Dansville, Michigan, in a randomized complete block design with A replications. Each replicate consisted of one or more 3.05 m (10 ft) rows of each generation spaced 61 cm (2 ft) between rows, 15.2 cm (6 in) between plants giving 107,593 plants per hectare (43,560 per acre). Fertilizer and irrigation schedules approximated those recommended for commercial pickle produc- tion. When the majority of the fruits had reached maturity, data were collected from the first and, whenever possible, from the second fruit of each plant. Only mature fruits were evaluated. Each fruit was cut transversely through the center and visually classified for degree of carpel separation using the following rating scale: Class Loglo Class Description 1 0.00 No separation between adjacent carpels. 2 0,30 Separation only at the vertex juncture of the 3 carpels. 3 0.A8 Any separation greater than the above but involving less than one-half of each carpel suture. 4 0.60 Any separation greater than above. 12 Although flood damage caused poor stands and variable fruit maturation, the data collected were sufficient to permit tentative conclusions concerning the inheritance of carpel separation and to indicate the magnitude of the environmental effects on this character. The need for a more detailed study utilizing a larger population was apparent. Detailed Investigation (1970). Three parents were used to produce populations for study. The P1 and P2 generations were selected from the S generations of SC601H l and MSU3BIG used in the preliminary investigation. The P3, MSU381M, was a monoecious line isogenic to gynoecious P2, differing from it only by sex expression. Genetic families (Fig. 3) were originated from single plant crosses by hand pollinations. The F1 and reciprocal F generations were made both between P and P2 and between 1 l P and P3. Three such single plant families were created 1 and utilized. Parents and progeny were field grown near Dansville, Michigan, in a randomized complete block design with 3 replications. Commercial varieties were included in each replication for comparison. Cultural practices used in the preliminary study were followed. In July a randomly chosen sample of #17 plants from the parental, F1’ and F2 genera- tions was labeled for use in comparing immature and mature 13 Figure 3. Typical family derived from cross of two single plants for use in the inheritance study. SC601H-1 MSU381M-l MSU3?lG-1 ® 0 + + + P1 P2 P3 (P1 X P2)Fl (Pl X P3)F1 (P2 X P1)Fl (P3 X P1)Fl XP (1 XP XP 1 XP 1 2 1 U! +. \l ¢r + Si BCl F2 BC2 BCl F2 BC3 1“ fruit carpel separation. Grade size 3 (3.8 to 5.1 cm dia.) immature fruits were cut longitudinally and scored visually as being either carpel separated or non—separated. In August, all plants in all generations were observed for mature fruit carpel separation. Data were collected from at least 1, and whenever possible from 2 or a maximum of 3, mature fruits on each plant. Each fruit was cut cross-sectionally into 4 segments so that the center, blossom end, and stem end could be classified according to the rating scale given previously. Thus a plant could be measured by as few as 3 determinations to as many as 9 readings. The mean of these 3, 6, or 9 determinations was used to measure the carpel separation of each plant. Data Analysis. The non-parametric Mann-Whitney U— and Kruskal-Wallis tests as well as parametric analyses of variance were employed for comparing distributions or means. The distributions of the segregating generations were continuous, exhibiting no clear-cut modes which would permit division into phenotypic classes. Consequently, data were analyzed biometrically using the methods outlined by Mather and Jinks (12). The original data were transformed to natural logarithms, common logarithms, and square roots. Mather's ABC scaling test (12) was applied to each trans- formation to determine on which the additive—dominance model was adequate. 15 The ratio of additive genetic variance to total pheno- typic variance was used to estimate narrow-sense heritability. The minimum number of genes differentiating the parents was computed using the methods reported by Castle (3), Wright (16), Burton (2), and Mather and Jinks (l2). 16 RESULTS AND DISCUSSION Preliminary Investigation. Original data were trans- formed to common logarithms because this transformation best conformed to the assumptions of Mather's additive-dominance model for partitioning of variance. Means and variances for carpel separation were calculated (Table l) for the parents, F F2, and backcrosses. The means of the F and F2 genera- l’ 1 tions fell between the means of the two parents. The means of the BC with P approached the mean of P likewise the 1 13 mean of the BC with P2 approached P2. No reciprocal cross differences were detected for the BC1 generation. Dominance for weak carpel suture was suggested by the fact that the F mean exceeded the midparent value (Table l 2). However, the comparatively higher variance of the BC with P2 than the BC with P1 suggested that P1, carpel sutured parent, carried a predominance of the domi- the strong nant genes. The estimated values for the components of variance (Table 2) suggest that additive and dominant effects as well as environmental effects were important in the inheritance of mature fruit carpel separation. Relatively high esti- mates for degree of dominance and heritability were obtained. Estimates of k suggest that a minimum of 2 gene pairs conditioned this character. Detailed Investigation. In both the non-parametric and parametric tests no differences (5% level) were found 1? Table 1. Means, standard errors, and variances for mature fruit carpel separation ratings in the family SC601H X MSU381G, preliminary study. Population Total Pedigree Gen ngngg Meanz Phenotypic Variance SC601H P1 15 0.00 i .000 0.000 MSU38lG P2 13 0.u2 1 .023 0.007 SC601H X Fl 23 0.25 1 .035 0.028 MSU381G Fl self F2 48 0.20 i .030 0.0A2 Pl X F1 BCl 66 0.12 i .019 0.023 F1 X Pl BClrecip 59 0.10 i .021 0.026 P2 X Fl BC2 32 0.27 i .033 0.035 ZData transformed to loglo. 18 Table 2. Estimates of components of variation, degree of dominance, heritability, and probable number of effective factors for mature fruit separation in the family SC601H X MSU381, preliminary and detailed studies. Preliminary Detailed Environmental variance (E) 0.017 0.010 Additive variance (D) 0.051 0.015 'Dominance variance (H) 0.019 0.009 H Degree of dominance Q/g) 0.61“ 0.750 Midparent value 0.212 0.283 F1 mean 0.253 0.387 Narrow-sense heritability (h2) In F2 0.39 In F2 and BC gen 0.62 0.A5 Number of effective factors (k) Castle (1921) 1.6 1.u Wright (193A) 1.9 Burton (1951) 1.6 1.6 Mather and Jinks (1971) 1.2 19 1'5, and between plants maturing l and 3 or more fruits. Data were pooled between P2 and P3, between reciprocal F accordingly. Original data were transformed to common logarithms as in the preliminary study. Immature-Mature Fruit Carpel Separation Comparison. Frequency distributions comparing carpel separation ratings of mature fruits harvested from plants producing separated immature fruits and those producing non-separated immature fruits were developed (Fig. A). Plants scored as non- separators in the immature fruit stage had an average mature fruit carpel separation rating of 0.1“; whereas plants having separated immature fruits produced mature fruits with a 0.4“ average mature fruit carpel separation rating. The difference between these two means was highly significant (.01 level), showing that there was good agree- ment between mature and immature carpel separation. Con- sequently, degree of carpel separation in the mature fruits seemed a reliable indicator of the potential for this defect in the immature fruits used for pickling. This conclusion is reinforced by the work of Hooper (A) in which the carpel suture strength of immature fruits of SC601H (P1) and MSU381 (P2&3) was measured with an Instron puncture technique (6). Using 6 mm slices from No. 3 fruits, he found that a mean force of 318.6 E 55.1 g was required to separate the carpels of SC601H; whereas the carpel suture of MSU381 required only 222.7 E 32.2 g. c. omit. ms oub UmzmHHompmm on Umhmdsoo mebmz x mHomom mo mHHEmm UoHooa on» Eoum pmoezozo CH COHumLonm Hmopmo uH3pm opsumfi pom Axv mCOHpanpume hocmSUmpm m>HpmHmm .m mHomB 23 75 75 50 'fi PI 50 Fl 25 25 I—-1 0 0 a LoowOOO 020 040 0.60 manaoo 0.20 040 0.00 :0 LG 2.5 4.0 to us 4.0 g 75 75 I.” g 50 P2+3 50 F2 E "i E 25 25 I: _ , j 0 , '1 0 I I I f ”L 000 0.20 0.40 0.60 L 000 0.20 040 0.60 “5 o"'“Io I.6 2.5 4.0 0°” I.0 lb 2.5 4.0 75 75 50 BC. 50 BC, 25 25 FIT—IT-Ij L 000 0.20 0.40 0.60 L 000 0.20 0.40 050 0°” LO :5 2.5 4.0 0°” LG LG 2.5 4.0 CARPEL SEPARATION RATING CARPEL SEPARATION RATING Fig. 5. Frequency distributions (%) for carpel separation In mature fruits of cucumber from the family of SC SOIH at MSU 38L 2U Table A. Means, standard errors, and variances for mature fruit carpel separation in cucumber from the pooled family of SC601H X MSU381; detailed investigation. Population Total Pedigree Gen No of Meanz Phenotypic Plants Variance SC601H Pl 100 0.11 i .012 0.01“ MSU381 P2&3 168 0.A5 i .006 0.006 SC601H X Fl A84 0.39 i .OOA 0.009 MSU381 F1 self F2 723 0.31 i .005 0.020 F1 X P BCl 528 0.2“ i .006 0.019 Fl X P BC2 39A 0.Al t .006 0.013 Pioneer Fl 60 0.28 i .013 0.010 Spartan F A7 0 A3 i 011 0 006 Progress 1 ° ° ' Pixie Variety 300 0.27 i .007 0.01M zData transformed to loglo. 25 midparent value (Table 2) and the BC to P2 exhibited a lower variance than the BC to P1. Estimated values of components of variation, degree of dominance, heritability, and minimum number of effec— tive factors were computed (Table 2). The additive portion of the variance exceeded the dominance variance but was almost equaled by the environmental, or non-heritable variance. This value for environmental variance was large, but realistic. The results of the initial analysis of variance in which environmental variance was measured by the differences between replications was highly signifi- cant and accounted for a large portion of the total pheno- typic variation in carpel separation (Table 5). The calculated degree of dominance suggested partial dominance. Narrow-sense heritability, which reflects the effectiveness of selection for carpel suture strength, ranged from 39 to 45%. Four estimates of k, the minimum number of effective factors controlling carpel separation, were computed. Each gave a value between 1 and 2. This is an underestimate of the actual number of gene pairs involved because a comparison of the s(dh) and /DH (12) indicated that one or more of the following was true: the increments due to each gene were unequal, the distribution of genes was not completely isodirectional, linkage was present, or the genes involved could not be phenotypically distinguished. 26 Table 5. Analysis of variance for mature fruit carpel separation in cucumber for parent and F genera- tions of the pooled family SC601H x MSU381; detailed investigation. Sources of Degrees of Mean F Variation Freedom Squares Total 751 Generations 2 76.28 345.7** Replications 2 5.17 23.4** Error 747 0.22 **P < 0.01. 27 The values calculated suggested that a 2 gene pair system was in operation. A large number of gene models were tried and rejected because of low probabilities in chi-square tests for goodness of fit. However, the data suggest that segregation at two loci, or at the most 3 loci, accounted for most of the variation in carpel sepa- ration in the crosses studied. A clear-cut dihybrid ratio may have been obscured by incomplete penetrance and possibly by incomplete dominance, linkage, and genotype x environment interactions. The Relationship of Carpel Separation to Sex Expression and Yield. Recently it has been implied by growers and processors that predominantly female cucumber hybrids produced on gynoecious seed parents, when compared to monoecious varieties, have more carpel separation and higher frequencies of bloated fruits in brine stock. The isogenic lines used as P2 and P3 were compared with each other in terms of carpel separation ratings. No differ- ences (5% level) were found between P2 and P3, nor between the F F2, or backcross populations resulting when each 13 was crossed with P1' Segregating populations resulting from the cross P2 X Pl (MSU38lG X SC601H) contained gynoecious segregates whereas those resulting from the cross P3 X P1 (MSU381M X SC601H) contained only monoecious plants. Nonetheless, the 28 populations containing gynoecious segregates did not have significantly higher levels of mature fruit carpel separa- tion than those populations containing no gynoecious segregates. These data suggest that weak carpel suture strength is not necessarily associated with gynoecious sex expression, and are in agreement with Barnes (1), who analyzed data collected over several seasons, and concluded that F hybrids made with gynoecious seed parents brined as 1 well as their respective monoecious male parents. Tests for significance also indicated no differences (P > .05) in degree of carpel separation between those plants maturing one fruit and those maturing 3 or more fruits. Since the number of mature fruits per plant is an indication of yield capacity, these results suggest that weak carpel strength is not always associated with high yielding capacity. 10. 29 LITERATURE CITED Barnes, W. C. 1970. Testing brine qualities of new varieties. The Pickle Pak 30:3. Burton, 0. W. 1951. Quantitative inheritance in pearl millet (Pennisetum glaucum). Agron. J. 43:409- 14170 Castle, W. E. 1921. An improved method of estimating the number of genetic factors concerned in cases of blending inheritance. Science 54:223. Hooper, A. W. 1971. A study of the carpel strength of various varieties of pickling cucumbers. Unpub- lished paper submitted to Ag. Engineering Dept., M.S.U. 8 pp. Hooper, A. W. 1973. The effect of impact on brine stock quality and green stock carpel strength for cucumbers, Cucumis sativus L. Thesis for M.S. Degree, Dept. of Ag. Engineering, M.S.U. 46 pp. Hooper, A. W., D. E. Marshall, L. R. Baker, D. R. Heldman. 1972a. A method for measurement of carpel strength in pickling cucumbers. ASAE Paper No. 72-379. 12 pp. Hooper, A. W., L. R. Baker, D. E. Marshall, and D. R. Heldman. 1972b. Determination of force required to separate carpels of cucumber fruit slices. Hort- Science 7:336,337 (Abstract). Jones, I. D., J. L. Etchells, and R. J. Monroe. 1954. Varietal differences in cucumbers for pickling. Food Tech. 8:415-418. Marshall, D. E., B. F. Cargill, and J. H. Levin. 1971. Mechanical harvesting and handling of pickling cucumbers - an evaluation of green stock and brine stock quality. ASAE Paper No. 71—348. 8 pp. Marshall, D. E., L. R. Baker, J. H. Levin, and B. F. Cargill. 1972. The effect of mechanical harvesting and handling on pickling cucumber quality. ASAE Paper No. 72-885. 7 pp. ll. 12. 13. 14. 15. 16. 30 Marshall, D. E., J. H. Levin, and D. R. Heldman. 1973. Density sorting of green stock cucumbers for brine stock quality. ASAE Paper No. 73-304. 15 pp. Mather, K. and J. L. Jinks. 1971. Biometrical Genetics. Cornell University Press, Ithaca, N. Y. 382 pp- Nicholas, R. C. and I. J. Pflug. 1962. Variety response to some variables in fresh cucumber pickle production. Michigan Quarterly Bulletin 44:739-750. Sneed, F. D. 1969. A study of several green fruit characters of Cucumis sativus L. as related to certain quality determinations in brine stock. Thesis for M.S. Degree, U. of Arkansas. 56 pp. Sneed, F. D. and J. L. Bowers. 1970. Green fruit characteristics of cucumber as related to quality factors in brine stock. J. Amer. Soc. Hort. Sci. 95: 489-491. Wright, S. 1934. The results of crosses between inbred strains of guinea pigs, differing in number of digits. Genetics 19:537-551. ”'TITII‘IIITIIIIIIIIflifllfli’lfllfllflljilllll“ 1293 0