Euphytica 32 (1983) 845-855 IMPROVEMENT OF TOMATO FLAVOR BY GENETICALLY INCREASING SUGAR AND ACID CONTENTS R. A. JONES Department of Vegetable Crops, and S. J. SCOTT University Received of California, 4 January Davis, CA. 95616, USA 1983 INDEXWORDS Lycopersicon esculentum, tomato, flavor, sensory evaluation, soluble solids. SUMMARY Advanced high sugar and acid breeding lines of tomatoes (Lycopersion esculentum MILL.) were rated higher in sweetness, sourness and overall flavor intensity than the standard cultivars Cal Ace or T,. Titratable acidity and soluble solids content were major contributors to differences in overall flavor intensity. The results demonstrate that significant improvement in tomato flavor can be attained by increasing sugar and acid contents in tomato fruits by genetic manipulation. Current evidence indicates that breeding for high soluble solids in horticulturally acceptable tomato cultivars is justified. INTRODUCTION Fresh tomato quality is a complex characteristic, involving not only physical appearance and firmness, but also flavor and nutritive value. Over the past decade considerable research effort has been directed towards understanding the contribution that various fruit constituents make to the edible quality of fresh tomatoes. Continued explorations into the role of organic compounds in determining flavor attributes are pertinent amidst continued reports of consumer dissatisfaction with the quality of fresh fruits and vegetables (MILLER, 1974; HANDY & PFAFT, 1975). Changes in tomato fruit composition during maturation and ripening have been reviewed (HOBSON &.DAVIES, 1971; SALUNKHE et al., 1974). Total solids and total sugars increase progressively during development and maturation (WINSOR et al., 1962a; 1962b). Generally, acidity in tomatoes increases during development and reaches a maximum at the breaker stage, then decreases with further ripening (DAVIES, 1966; STEVENS, 1972). Among the organic substances in tomato fruits, the sugars and acids are major constituents important to tomato flavor differences and affecting tomato flavor quality. SIMANDLE et al. (1966) compared the composition of 6 tomato cultivars with their eating quality and found panel flavor scores to be signiticantly correlated with soluble solids and pH. DE BRUYN et al. (1971) concluded that high sugar and acid contents generally have a favorable effect on taste, and that flavor intensity was more closely correlated with acid content than with sugar content of greenhouse tomato cultivars. Sugars and acids not only contribute to sweetness and sourness of tomatoes (DE BRUYN et al., 1971; STEVENS et al., 1977), but also appear to be important 845 R. A. JONES AND S. 1. SCOTT factors in determining overall flavor intensity (STEVENSet al., 1979). In addition, numerous volatile compounds have a significant effect on flavor variation (DE BRUYN et al., 1971; KAZENIAC &HALL, 1970; KADER et al., 1977). Few investigators have evaluated the impact of fruit maturity at picking on flavor quality of tomato fruits. The inability of table-ripe fresh market tomatoes to withstand the handling necessary for long distance transport has led to the standard practice of harvesting fruits at earlier stages of maturity which will permit handling, and subsequently ripen them in transit or at destination. Evaluating the impact of picking stage on flavor quality of tomato fruits, BISOGNI et al. (1976) reported that vine-ripened tomatoes had better flavor and overall quality than room-ripened tomatoes. Similarly STEVENSet al. (1977) found that fruit picked mature green or at the breaker stage (incipient red color) and ripened off the vine to table-ripe had poorer flavor quality than fruit picked at the table-ripe stage. In the former study no significant differences were found in pH, titratable acidity or soluble solids content, while in the latter study tomatoes picked at earlier stages of ripeness had lower soluble solids content. It is probable that problems with tomato quality, particularly flavor, result from harvest and post-harvest handling procedures, and/or to genotypic differences in responses to these procedures. Tomato fruits picked green and ripened off the plant would appear then to have lower flavor quality than vine ripened fruits. The purpose of these investigations were to determine what impact high concentration of sugars and acids have on flavor quality in advanced high solids breeding lines and in specially developed Ft hybrids and to determine what effect maturity at picking has on tomato composition and flavor at the table-ripe stage. MATERIALSANDMETHODS Genotypes. The genotypes studied included the control lines cv. Cal Ace or Ts, several advanced high sugar breeding lines developed in our fresh market tomato breeding program and selected Fi hybrids between the high soluble solids parents. The high solids lines were: 1) HS89, derived from STEP375; 2) HS70, derived from an interspecific cross (L. peruvianurn); and 3) HS93. Plants were grown in the field at Davis, California using standard cultural practices. Several plantings were made to provide fruits throughout the normal harvest period (August through September). Fruit sampling. Fruits were harvested at the table-ripe stage on the day of evaluation. All samples were havested within a 5 day period to minimize possible environmental effects on fruit composition. Excess fruits were picked, sorted and only fruits in good condition with uniform color, size and firmness were selected for use. Each sample contained a minimum of 20 fruits which were cut in half from stem-end to blossom end. One group of half-fruit was diced (about 1 cm3) and thoroughly mixed for sensory analyses. The remaining lot of half-fruit was retained for chemical analyses. In a separate study, fruits of ‘Cal Ace’, HS70, HS89, HS93, and HS65 were harvested at the full-size, mature-green stage. Fruits were held at 20°C under high relative humidity to ripen. The green fruits were sorted after 5 days in storage and those showing color development were subjectively selected as mature-green (MG) when harvested. When these fruits reached the table-ripe (TR) stage they were used for sensory and 846 Euphytica 32 (1983) TOMATO compositional of each line. FLAVOR IMPROVEMENT evaluations in comparison with freshly harvested, vine-ripened fruits Compositional analyses. The fruit samples retained for chemical determinations were each blended, filtered through Whatman # 1 paper, and the serum used for chemical analyses. Titratable acidity was determined .by titrating 5 ml of the filtered tomato serum to pH 8.1 with 0. IN NaOH and pH was measured with a Corning digital 109 pH-meter. Soluble solids content (SK) was determined as “Brix with a table-top model ABBE-3L Bausch & Lomb refractometer. Reducing sugars were determined by the method of SOMOGYI(1952). Sensory evaluation and statistical analyses. Panelists were instructed in the use of the scoring system, described previously (KADER et al., 1977) for the flavor characteristics of sweetness, sourness, ‘tomato-like’, and overall flavor intensity. Genotype and treatment comparisons were presented in random order to panelists as a complete block design, with blocking by days. For each flavor characteristic, standard deviations of individual judges were calculated by genotype and 1.5judges were selected for lower deviations from the aggregate mean response. Analysis of variance was performed for the sensory scores and the compositional analyses. Significance levels were determined with Duncan’s New Multiple Range test (STEEL& TORRIE, 1980). Results of the sensory evaluations were analyzed by multivariate analysis of multiple covariance (COOLEY& LOHNES, 1971) in two ways. First, all chemical variables, soluble solids (SSC), reducing sugars (RS, total titratable acidity (TTA), total acidity (TA), and pH were used as covariates. Secondly, only SSC and TTA were used as covariates. All four sensory variables were simultaneously evaluated as dependent variables. Results of these analyses were essentially identical as to significance of covariate terms in the model and in adjusted R2 values for each dependent (sensory) variable. Therefore, results using only SSC and TTA as covariates are reported here, except for the correlation matrix from the first analysis. RESULTS Cultivar studies. Comparisons between cultivars revealed significant variations in pH, titratable acidity, reducing sugars and soluble solids content (Fig. 1). Genotypic differences in total soluble solids and reducing sugars showed a similar pattern. Fruits of the high sugar breeding lines had significantly higher pH, reducing sugars and total soluble solids content than ‘T3’. The control cultivar T3 had significantly higher total titratable acidity and a lower pH than the three high sugar lines. HS93 had the highest sugars and was intermediate in both pH and titratable acidity. HS89 and HS70 were intermediate in sugars content. HS89 had the highest pH and the lowest titratable acidity, while HS70 had the highest titratable acidity among the high sugar lines. The F, hybrids had significantly greater levels of reducing sugars and soluble solids content than ‘T3’ (Fig. 1). One hybrid, HS89 x HS70, had soluble solids content comparable to both high sugar parents in the cross, while fruit of the hybrid between HS93 and HS70 had levels similar to the lowest parent in the cross. HS93 and fruit of the hybrid HS93 x HS89 had significantly greater soluble solids and reducing sugar content than any of the genotypes examined. The hybrids had significantly greater Euphytica 32 (1983) 847 R. A. JONES REDUCING SUGARS SOLUBLE CONTENT AND SOLIDS S. J. SCOTT PH TITRATABLE ACIDITY d i d z I a4 P 4.25 5 4 3 4.00 0 1234567 GENOTYPE Fig. 1. Mean composition of ‘Tj’ (l), three high sugar parents (HS93 = 2; HS89 = 3; HS70 = 4) and 3 high sugar hybrids (2 x 4 = 5; 3 x 4 = 6; 2 x 3 = 7). Means among lines not labeled by the same letter differ significantly at the 5% level. pH than ‘T3’, but lower than HS89. Titratable acidity in the hybrids was greater than HS93 and HS89, comparable to HS70, but generally lower than ‘T3’. The exception, HS93 x HS70 had levels similar to ‘T3’. Panelist scores for sweetness, sourness and tomato-like flavor varied significantly among the lines (Fig. 2). The differences for overall flavor intensity were not signiticant. HS90 had the highest sweetness score; HS89 and HS70 were intermediate. ‘T3’ had the lowest sweetness score and the highest sourness score. HS89 was equivalent in sourness to ‘T3’ while that of HS90 and HS70 were lower. HS90 and HS70 had the highest tomato-like scores, while HS89 was intermediate and ‘T3’ was lowest. Generally, the Fi hybrids between the high sugar parents had greater sweetness, sourness, tomato-like and overall flavor intensity than ‘T3’ (Fig. 2). In comparison with the parental lines, the Fi hybrids tended to have greater sourness and overall flavor intensity, but equivalent tomato-like flavor. The exception was hybrid HS89 x HS93 which was rated higher in tomato-like flavor. Hybrid HS89 x HS70 was rated higher in sweetness than either of its parents. Sweetness in hybrid HS89 x HS93 was identical to that of the higher solids parent in the cross, while fruits of HS93 x HS70 were identical to the lower solids parent. Differences in sweetness were related to variations in reducing sugars and soluble solids content (Table 1). Sweetness also showed a strong correlation with tomato-like and to a lesser degree, with overall flavor intensity. Tomato-like flavor was also highly 848 Euphytica32 (1983) TOMATO 7 I 3 FLAVOR SOURNESS SWEETNESS IMPROVEMENT TOMATO-LIKE OVERALLINTENSITY 6! ’ 2 ’ OL 1234567 1234567 GENOTYPE Fig. 2. Mean sensory scores for flavor character of ‘T,‘, three high sugar parents and three high sugar hybrids. Refer to Fig. 1 for genotype entries. Means not labelled with the same letter differ significantly at the 5% level. correlated with overall flavor intensity. In this study sourness showed a low correlation with overall flavor intensity. Sourness was surprisingly not strongly correlated with titratable acidity or with pH, but this probably reflects the lack of a definite trend of acid concentration and the lack of a consistent relationship between titratable acidity and acid concentration among the genotypes. As much as 59”/, of the variance in sweetness, adjusted for degree of freedom, was Table 1. Correlation matrix for flavor 3 F, hybrids harvested at the table-ripe Variable Sensory Sourness ‘Tomato-like’ Overall intensity PH Titratable acidity Reducing sugars Soluble solids * and ** indicates Euphytica 32 (1983) characteristics stage. and chemical components cuhivars and parameter sweetness sourness ‘tomato-like’ overall -0.07 0.53** 0.37* 0.08 -0.05 0.47** 0.49** 0.17 0.37* -0.25 0.32 PO.12 -0.14 0.58** 0.10 -0.02 0.29 0.35* 4.10 0.13 0.11 0.1 I significant in 4 tomato difference intensity at the 5% and 1% level, repectively. 849 R. A. JONES IEDUCING w SUGARS AND S. I. SCOTT SOLUBLE SOLIDS CONTENT (%I PH TlTRATABLE ACIDITY (% CITRIC) q IMG u TR f ? e R5 c 0.4 0.3 12345 12345 12345 il12345 GENOTYPE Fig. 3. Mean composition of five tomato cultivars (1 = Cal Ace; 2 = HS89; 3 = HS93; 4 = HS65; 5 = HS70) harvested at two stages of ripeness. Overall means for each picking stage not labelled with the same letter differ significantly at the 5% level. described by the covariate model (Table 2). Analysis using soluble solids, total titratable acidity, reducin sugars, pH and total acidity as covariates did not increase the adjusted or simple R !?values. The contribution of soluble solids to the taste characteristic of sweetness can be reasonably understood, but additional modifying factors other than titratable acidity are apparently also involved but were not measured. Analysis of sourness indicated that the covariate model accounted for only 41% of the variation observed in panelist responses. ‘Tomato-like’ is even more complex, since Table 2. Multivariate and concentrations analysis of covariance of sensory characteristics of the covariates in the model, acids and sugars. Flavor characteristic (dependent variables) R2 simple adjusted’ Sweetness Sourness Tomato-like Overall flavor intensity 0.75 0.64 0.63 0.59 0.41 0.39 0.63 0.38 ’ Adjusted of freedom. 850 of tomatoes for degrees Euphytica 32 (1983) TOMATO r SWEETNESS FLAVOR SOURNESS IMPROVEMENT TOMATO-LIKE OVERALLINTENSITY IJMG rnTR a ~ 1 2 4 5 12 34 5 12345 GENOTYPE Fig. 4. Mean sensory scores of five tomato cultivars harvested at two stages of ripeness. 3 for entries. Means not labelled with the same letter differ significantly at the 5% level. Refer to Fig. the model described only 39% of the variance observed for this taste attribute. The covariates accounted for only 38% of the variation in overall flavor intensity of tomato fruits. Picking stage. Fruits picked at the mature green (MG) stage and ripened to table ripe (TR) at 20 “C generally had significantly different levels of chemical components than fruit ripened on the plant and picked at the TR stage (Fig. 3). The pH was significantly lower in fruits picked at the MG stage than in the TR fruits in all genotypes, except HS89 where the 2 picking stages were identical. The differences in titratable acidity between the two stages were significantly higher in MG fruits of HS93 and HS65 and although not significantly different in the remaining genotypes, also tended to be higher in MG fruits. Reducing sugars and soluble solids content are also lower in fruits picked at the MG stage than those left to ripen on the vine. Large differences were observed among the lines for both soluble solids and reducing sugars for the two stages of fruit ripeness. The high sugar lines showed a larger decrease in sugars when fruits were picked MG compared to TR than did ‘Cal Ace’ which overall has significantly lower levels of reducing sugars. Even though fruits of the high sugar lines picked MG had approximately 1% less sugars or soluble solids than their respective TR fruits, the MG levels were significantly greater than that found in fruits of ‘Cal Ace’ left to ripen on the vine. The exception was fruits of HS93 picked MG which were equivalent to TR fruits of ‘Cal Ace’. Fruits picked at the mature green (MG) stage and ripened to TR were significantly different in flavor characteristics than those ripened on the plant and picked at the TR stage. Significant differences were observed for all sensory characteristics evaluatEuphytica 32 (1983) 851 R. A. JONES AND S. J. SCOTT ed except overall flavor intensity (Fig. 4). In general, fruits picked at the TR stage were sweeter, less sour and had more tomato-like flavor. The high soluble solids lines had considerably higher sweetness scores in fruits picked on the vine at the TR stage than in fruits picked at the MG stage and ripened off the vine to TR. Large differences in sourness were observed in all lines with fruits picked at the MG stage having signiticantly higher sourness scores than their respective TR fruits, but HS89 showed no significant differences. All the lines showed a difference between fruits at the mature green and the table ripe stage for ‘tomato-like’ flavor, with fruits picked at the TR stage having significantly higher scores. Generally, the high soluble solids lines did not differ among themselves in overall flavor intensity, nor did individual genotypes differ between fruits picked at the two stages of maturity. However, in comparison with TR fruit of ‘Cal Ace’, the high solids lines exhibited significantly greater overall flavor intensity whether they were picked MG or TR. The correlation matrix shows a strong positive relationship between sourness and titratable acidity and a negative relationship with pH (Table 3). Differences in sweetness were related to variations in soluble solids and reducing sugars. Overall flavor intensity was correlated with sourness, titratable acidity and sugars. The tomato-like flavor characteristic was correlated with sweetness and sugars and was also related to overall flavor intensity. DISUSSION Great difficulty is encountered when attempting to improve flavor quality in a breeding program. Perception of flavor is a complex sensation involving the tastes and aroma of numerous compounds (BUTTERY et al., 197 1; STEVENS et al., 1977). The large number of compounds involved and their subtle interactions underscores the difficulty in dealing effectively with flavor attributes. In addition, sensory analysis requires large sample sizes for meaningful statistical comparisons, because of great variability in panelist evaluations and variability in fruit composition. Thus, the ability to discriminate among genetically different progenies becomes a burdensome task not easily accomodated within even a modest breeding program. Attempts to overcome such limitations Table 3. Correlation matrix for mean flavor characterisctics and mean chemical components of five tomato cultivars harvested at two ripeness stages. Variable Sourness Tomato-like Overall Intensity PH Titratable Acids Reducing Sugars Soluble Solids Sensory Parameters sweetness sourness tomato-like overall intensity 0.13 0.60** 0.46** a.10 0.14** 0.38** 0.44** 0.19 0.62** -0.41** 0.52** 0.10 0.12 0.70** -0.10 0.33* 0.58** 0.45** Xl.34** 0.61** 0.44** 0.50** * and **indicate significant difference at the 5% and 1% level, respectively. 852 Euphytica 32 (1983) TOMATO FLAVOR IMPROVEMENT have focused on identifying fruit constituents which have a major impact on flavor and evaluating the effects of genotypic variation in these components. Objective evaluation of key components combined with subjective evaluations by sensory panels may identify the components most responsible for flavor variation. Fruit composition may then serve as useful objective selection criteria within a breeding program to improve flavor attributes. The results presented here clearly underscore the importance of variation in sugar and acid content to genotypic flavor differences. While in agreement with previous studies reporting that sugars and acids are important to tomato flavor (DAVIES & WINSOR, 1969; STEVENS et al., 1979; DE BRUYN et al., 1971; SIMANDLE et al., 1966; LOWER & THOMPSON, 1967), this is the first study to assess quantitatively the impact of significant increases in soluble solids content of commercially acceptable tomato types on flavor qualilty. Changing the genetic determinants which condition sugars assimilation resulted in dramatic improvements in tomato flavor quality. Not only was sweetness improved but also, overall-flavor intensity and tomato-like flavor. Previous reports evaluating tomato flavor quality have utilized a multiple regression model in which tomato-like was the dependent variable and sweetness, sourness, overall intensity and all chemical components were independent variables. Such a model assumes that sweetness, sourness, overall intensity and the chemical parameters are all components of tomato-like flavor. This interpretation would appear reasonable, but depends upon independence of the contributing variables. Examination of the correlation matrix clearly indicates that this assumption is invalid (Table 1). A pattern of correlation is clearly evident among the sensory variables. A more complex model is needed to evaluate relationships between a group of correlated sensory variables and a group of variables that partially represent fruit composition. This study therefore employed multivariate analysis of multiple covariance. Fruit composition variables were considered independent or predictor variables, while all four sensory variables were considered to be dependent on fruit composition. This approach allows simultaneous evaluation of the sensory variables and may more closely describe the complexities of taste. Soluble solids content and titratable acidity were selected as covariates in a model to test whether they can adequately describe or predict flavor attributes. The covariance model accounted for 59% of the variance for sweetness, indicating that inclusion of soluble solids content as a covariate was a major contributor to this flavor characteristic. Similarly the covariate model accounted for a significant portion (41%) of the variance for sourness. It is important to note that inclusion of reducing sugars, pH, and total acidity as additional covariates, failed to further increase R2 values, suggesting that they are either not important or are redundant. The low adjusted R2 values suggest that variables other than those measured may also contribute significantly to flavor. However, that soluble solids content and titratable acidity may serve as relatively simple yet effective screening criteria is evident. Applied to segregating lines, determination of these chemical parameters should help to identify potential progeny with improved flavor attributes. Selection for horticultural traits may further reduce the sample size to a manageable level for sensory analysis. Picking tomatoes at the MG stage significantly alters their sugar and acid content when ripened to table-ripe. These changes contribute significantly to variation in flavor quality compared to fruits harvested table-ripe. Fruits left to ripen on the vine have Euphytica 32 (1983) 853 R. A. JONES AND S. J. SCOTT higher sugar content and are sweeter than fruits picked at the MG stage and allowed to ripen off the vine. This observation is consistent with previous studies indicating that fruits picked at earlier stages of ripeness are lower in soluble solids and reducing sugars than fruit ripened on the vine (DAVIES, 1966; HALL, 1968; KADER et al., 1977). However, it appears that cultivars differ in TR fruit composition compared between fruits picked at the 2 stages of ripeness. Previous work has indicated that some cultivars show significant differences in fruit composition between room ripened and field-ripened tomatoes (KADER et al., 1977), whereas others demonstrate little or no change (BISOGNI et al., 1976). Many of the changes in chemical compostion which condition improved tomato flavor are initiated with the onset of ripening (first color) and tend to be cumulative throughout continued color development of fruits left on the vine (SALUNKHE et al., 1974; HOBSON & DAVIES, 1971). General&, fruits removed from the plant at earlier stages of ripeness do not allow these changes to reach their full potential. Our data clearly indicate that picking fruits at the mature green stage of ripeness adversely affects their TR flavor in a strong negative fashion relative to fruits harvested when table-ripe. In comparison with TR fruits of ‘Cal Ace’, the high solids lines exhibited significantly greater overall flavor intensity whether they were picked MG or TR. Even though fruits of the high solids lines picked MG had approximately 1% less sugars or soluble solids than their respective TR fruits, the MG levels were considerably greater than those found in fruits of ‘Cal Ace’ allowed to ripen on the vine. While ‘tomato-like’ flavor is a complex characteristic, there is little doubt that genetically altering the sugars content in the TR fruits, as in the high solids lines presently studied, also dramatically improves ‘tomato-like’ flavor in comparison with TR fruits of ‘Cal Ace’. Important in this regard are MG fruits of lines HS70 and HS89, which were judged superior to and HS93 which were judged equivalent in ‘tomato-like’ flavor to vine-ripened fruits of ‘Cal Ace’. Sugar content is important to tomato flavor; harvesting and handling techniques that reduce levels of this component have an adverse effect on flavor (KADER et al., 1977). The present study indicates that the problem of reduced tomato quality can be largely overcome by genetically increasing sugars and acids. The significant improvement of tomato flavor quality indicates that tomato breeders interested in high edible quality should undertake the breeding of cultivars with higher levels of soluble solids and acids. REFERENCES BISOGNI, C. A., G. ARMBUSTER & P. E. BRECHT, 1976. Quality comparisons of room ripened and field ripened tomato fruits. J. Food Sci. 41: 333-338. BUTTERY, R. G., R. M. SIEFERT, D. G. GUANDAGNI & D. C. LANG, 1971. Characterization of additional volatile components of tomato. J. Agric. Food Chem. 19: 524-529. COOLEY, W. W. &P. R. LOHNES, 1971. Multivariate data analysis. John Wiley&Sons. New York. DAVIES, J. N., 1966. Changes in the non-volatile organic acids of tomato fruit during ripening. J. Sci. Food & Agric. 17: 39&400. DAVIES, J. N. & G. W. WINSOR, 1969. Some effects of variety on the composition and quality of tomato fruit. J. Hort. Sci. 44: 331-342. DE BRUYN, J. W., F. GARRETSEN & L. K~OISTRA. 1971. Variation in taste and chemical composition of the tomato (Lycopersicon esculentum MILL.). Euphytica 20: 214-227. HALL, C. B., 1068. The effect of maturity at harvest on subsequent ripening characteristics of some tomato 854 Euphytica 32 (1983) TOMATO FLAVOR IMPROVEMENT cultivars. Proc. Fla. Stat. Hort. Sot. 81: 221-225. HANDY, C. R. & M. PFAFT, 1975. Consumer satisfaction with food products and marketing services. U.S. Dept. Agric., Econ. Res. Serv. Agric. Econ. Rpt. 281. HOBSON,G. E. & J. N. DAVIES, 1971. The tomato. p. 437482. In: A. C. HULME (Ed.), The biochemistry of fruits and their products, Vol. 2. Academic Press, New York. KADER, A. A., M. A. STEVENS,M. ALBRIGHT-HOLTON, L. L. MORRIS & M. ALGAZI, 1977. Effect of fruit ripeness when picked on flavor and composition in fresh market tomatoes. J. Amer. Sot. Hort. Sci. 102(6): 724-73 1. KAZENIAC, S. J. & R. M. HALL, 1970. Flavor chemistry of tomato volatiles. J. Food Sci. 35: 519-530. LOWER,R. L. &A. E. THOMPSON,1967. Inheritance of acidity and solids content of small-fruited tomatoes. Proc. Amer. Sot. Hort. Sci. 91: 486-494. MILLER, J., 1974. Agriculture: FDA seeks to regulate genetic manipulation of food crops. Science 192: 240. SALUNKHE, D. K., S. J. JADAV & M. H. Yu, 1974. Quality and nutritional composition of tomato fruit as influenced by certain biochemical and physiological changes. Qual. Plant., Plant Foods, Human Nutr. 24: 85-113. SIMANDLE,P. A., J. L. BROGDON,J. P. SWEENEY,E. 0. MOBLEY & D. W. DAVIS, 1966. Quality of six tomato varieties as affected by some compositional factors. P. Amer. Sot. Hort. Sci. 89: 532-538. SOMOGYI,M., 1952. Notes on sugar determination. J. Biol. Chem. 195: 19-23. STEEL,R. G. D. &J. H. TORRIE.1980. Principles and procedures of statistics. McGraw-Hill, New York. STEVENS,M. A., 1972. Citrate and malate concentrations in tomato fruits: genetic control and maturational effects. J. Amer. Sot. Hort. Sci. 97: 655-685. STEVENS,M. A., A. A. KADER, M. ALBRIGHT-HOLTON& M. ALGAZI, 1977. Genotypic variation for flavor and composition in fresh market tomatoes. J. Amer. Sot. Hort. Sci. 102(5): 680-689. STEVENS,M. A., A. A. KADER & M. ALBRIGHT, 1979. Potential for increasing tomato flavor via increased sugar and acid content. J. Amer. Sot. Hort. Sci. 104(l): 40-42. WINSOR,G. W.: J. N. DAVIES & D. M. MASSEY,1962a. Composition of tomato fruit. III. Juices from whole fruit and locules at different stages of ripeness. J. Sci. Food & Agric. 13: 108-l 15. WINSOR,G. W., J. N. DAVIES & D. M. MASSEY, 1962b. Composition of tomato fruit. IV. Changes in some constitutents ofthe fruit walls during ripening. J. Sci. Food & Agric. 13: 141-145. Euphytica 32 (1983) 855
© Copyright 2026 Paperzz