2005 Volume 9 Issue 1 Pages 32-39
Tomatoes (Lycopersion esculentum Mill.) are one of the most familiar vegetables in Japan and most other countries. Due to the recent progress in biotechnology in selective breeding, a variety of tomatoes with distinctive features such as high-sugar-content tomatoes, cherry tomatoes, and cooking tomatoes have been developed in Japan. In particular, high-sugar-content tomatoes are an excellent target for new product development for many tomato growers. However, it has not been clarified how consumers eat high-sugar-content tomatoes in actuality. We studied the tomato preferences of university students using extensive questionnaire surveys and a sensory evaluation (Honda et al., 2003; Isazawa et al., 2004; Kinoshita et al., 2002; Tamaki et al., 2000ab, 2003ab and 2004ab). Our previous questionnaire surveys showed that university students like red, ripe, firm-textured tomatoes that are rich in taste but not too strong-flavored or too sweet (Tamaki et al., 2004a). It was also found using sensory evaluation that consumers who usually purchase tomatoes at the supermarket tend to like high-sugar-content tomatoes more than those who usually consume home-grown tomatoes (Honda et al., 2003).
This study considers factors affecting the preference for high-sugar-content tomatoes compared with the regular tomatoes most widely consumed in Japan using sensory evaluation. Analytical sensory evaluation is usually conducted by comparing different samples to discriminate among small differences between samples. The data obtained in this manner do not always reflect preferences in actual food intake because people do not usually eat different kinds of tomato at the same meal. Furthermore, it is difficult to compare the features of tomatoes because they vary enormously from part to part. Palatability changes according to the time course of eating and how much is eaten. A tomato that is palatable at first bite may rapidly satiate the eater, but another tomato that is initially unpalatable may become more palatable with each bite. Matching with other ingredients is also an important factor at the time of purchase since tomatoes are not always eaten alone. Thus, it is important to extensively study eating conditions to evaluate tomatoes, especially new products.
This paper evaluates tomatoes individually in a natural manner as they are usually eaten. Tomatoes are evaluated both alone and in a salad. Evaluation was conducted over the time course of eating. The objective of the study is to identify the relative palatability and acceptance of high-sugar-content tomatoes and regular tomatoes under actual food intake conditions to maximize the number of consumers who are satisfied with their tomatoes.
This research was done as part of a contract research project in fiscal year 2004 for the quality assessment of tomatoes sponsored by the Japan Speciality Agriculture Products Association. The results of the project research will be published elsewhere.
Four kinds of tomato, two high-sugar-content (HS1&HS2) and two regular (R1&R2), were evaluated. They were grown in different places of production: R1 and HS2 in Yamagata prefecture, and R2 and HS1 in Fukushima Prefecture. They were delivered the day before the experiment directly from the production farmhouse and not through a retailer.
2. Chemical and physical analysesChemical analyses, which included proximate analysis (water, protein, lipid, ash, carbohydrate and dietary fiber), Brix (total soluble solid content), acidity, Vitamin A, ascorbic acid, lycopene and free amino acid analysis, were conducted by Japan Food Research Laboratories. Rupture properties were measured using Rheoner RE-33005 (YAMADEN Co. Ltd., Tokyo, Japan) under the following conditions. The firmness of the fruit was measured according to the method of Kanno et al. (1985). Sample pieces of skin and pericarp (about 1.5 × 1.5cm) were cut from each tomato, and the height was measured automatically. Plunger P-4 (φ3 × H22mm) was used to measure the skin. Plunger P-1 (φ30 × t8mm) was used to measure the pericarp. Compression was applied for the skin and the pericarp at a speed of 1mm/sec, at a strain rate of 90%; the breaking force (firmness) and the absolute value of the differential (crispness) at failure were read from the stress-strain curve obtained from the measurement according to the method of Omura et al. (1978). The equatorial surface region of the fruit was measured for color using a Minolta Chroma Meter CR-200 (MINOLTA Co. Ltd., Osaka, Japan). Readings are reported using the L*, a*, b* system.
3. Sensory evaluationSensory evaluation was conduced in July 2004. The panel consisted of 120 members recruited from the students of the Department of Nutritional Science of Tokyo University of Agriculture. All of them had previous experience as sensory evaluation panelists.
Sensory evaluation was conducted at lunchtime in a large cooking practice room of the Department of Nutritional Science of Tokyo University of Agriculture. Each group of four panelists sat down at one table. They were asked not to speak but to make themselves comfortable during the test session. The panel was given time before the evaluation to read and understand a list of questions. The tomatoes were evaluated under two eating conditions.
The tomatoes were first evaluated alone and then in a salad. First, 2 pieces of comb-shaped cut tomato) about 35 g per piece were given to each panelist who was then asked to score the list of questions while eating all of one piece. Evaluation was repeated twice. After evaluating each piece, the panelists were asked how many more pieces of tomato they wanted to eat. Then, the same tomato was evaluated in a salad with 3 pieces of the tomato (20 g per piece), 15 g cucumber, 20 g lettuce, and 10 g French dressing (McCormick-Lion Ltd., Shizuoka, Japan). Since a standardized amount of each sample was given, total tomato consumption was 130 g, approximately equivalent to one whole medium-sized tomato. The evaluation was conducted to measure the intensity or liking of attributes using a 7-point category scale: very weak (left), weak, slightly weak, neutral, slightly strong, strong, very strong (right), or dislike-like, bad-good, soft-hard, etc. in place of weak-strong (Table 1). At the end evaluation, the panelists were asked how often they would want to eat a sweeter tomato if it were available.
4. Data AnalysisMicrosoftⓇ Excel 2002 was used for subtotaling and drawing figures. Principal component analysis was conducted using JUSE-StatMaster/V2.0 Plus (The Institute of JUSE, Tokyo Japan). Correspondence analysis was performed using SPSSⓇ 11.5J Base System and Categories (SPSS Japan Inc., Tokyo, Japan).
Before the sensory evaluation, the panelists were asked about how they liked tomatoes at a conceptual level. As for the liking of the samples, a total of 17.5% of panelists liked high-sugar-content tomatoes very much, and a further liked them, so 87.5% liked high-sugar-content tomatoes. In their preference for the two types of tomato, 38.3% of panelists selected high-sugar-content tomatoes, 38.3% selected regular tomatoes, and 23.3% were neutral. The main reason that the panelists liked high-sugar-content tomatoes was because they are sweet and tasty. The main reason that they preferred regular tomatoes was that they were familiar with them.
Sensory characteristics of tomatoes evaluated using sensory evaluation
Means of the instrumental traits of the tomato samples
The results of the chemical analyses are shown in Table 2. The components of the two samples of the regular tomatoes, R1 and R2, were similar, and Brix was 5.7 and 5.8, respectively. As for high-sugar-content tomatoes, HS1 was 6.1 and HS2 was 11.0. The Brix of HS1 was almost the same as that of the regular tomatoes, but both free amino acid content and lycopene content were higher than that of the regular tomatoes.
The results of the sensory evaluation showed that the intensity of all flavors, i.e., sweetness, acidity, umami and thickness of flavor, of HS2 was outstandingly the highest among the four kinds of tomato when they were eaten alone (Fig. 1). The mean liking scores for taste attributes at each eating stage are plotted in Figure 2 as spider diagrams, and overall palatability scores are re-plotted in Figure 3 to show the differences between the samples more clearly. In Figure 2, HS2 obtained high evaluation scores in all items, followed by HS1 when eaten alone. As for regular tomatoes, the mean liking scores for sweetness, umami and thickness of flavor were slightly negative.
However, the evaluation scores of the regular tomatoes for after taste fresh-tasting were high compared with the high-sugar-content tomatoes. Moreover, when the tomatoes were evaluated in the salad, the evaluation scores of the high-sugar-content tomatoes increased moderately although the evaluation scores of the other tomatoes increased markedly compared to when the tomatoes were eaten alone (Fig. 3).
Means and SEs of the attributes (sweetness, acidity, umami and thickness of flavor) of the tomatoes obtaine from sensory evaluation. ■：1st piece; □：2nd piece.
Spider diagram of the mean liking rating of the sensory attributes of tomatoes. ●：1st piece; ○：2nd piece; ■：Salad.
As for color, samples HS1 and HS2 had higher a* values and a*/b* ratios and high concentrates of lycopene (Table 2). This indicates that high-sugar-content tomatoes are redder as well as darker than the regular tomatoes. However, the color did not affect how much the tomatoes were liked in the sensory evaluation.
The results of the mechanical measurements are shown in Table 3. The fruit of R1 was significantly firmer than the others excpet HS2 (P < 0.05). Sample HS2 had significantly firm skin, and it was significantly less crisp than the others. The large absolute value of the differential represented by crispness in Table 3 means a high rate of rupture speed at failure point. Thus, the skin of HS2 was regarded to be firm rather than crispy. It was shown in the sensory evaluation that the evaluation score of the firmness of the skin of HS2 was high for intensity and low for liking (Fig. 4). As for the pericarp, HS2 showed a significantly high value compared with the others. However, no significant difference was observed in the results of the sensory evaluation of the firmness and how much the tomatoes were liked.
Means and SEs of overall palatability scores of the tomatoes obtained from sensory evaluation.
■：1st piece; □：2nd piece; □：Salad.
Principal component analysis (PCA) was run on the correlation matrices using 24 sensory and hedonic attributes for the tomatoes being eaten alone (2nd piece) and 15 hedonic attributes for the tomatoes being eaten in a salad. When the tomatoes were eaten alone, 5 factors explain 65.5% of the variability. The factor-loading matrix is shown in Table 4. PC1 was characterized mainly by the panelists liking of flavor, taste and texture attributes. PC2 was mainly characterized by intensity of sweetness, umami and thickness of flavor against fresh-tasting and liking of texture. PC3 was mainly characterized by the sharp flavor and the intensity of acidity. PC4 was mainly characterized by appearance and maturity. PC5 was mainly characterized by firmness. Graphing the first two PC scores visually separated HS2 from the others by PC2, showing that HS2 tomatoes had a heavy flavor and were not fresh-tasting (Fig. 5).
In case of the salad, 3 factors explaining 66.3% of the date variability factor-loading matrix are shown in Table 5. PC1 was related to almost all hedonic attributes. PC2 was mainly related to the liking of sweetness, umami and thickness of flavor against firmness of the skin and feeling in the mouth. PC3 was mainly related to the liking of green flavor. In the case of the salad, at least the first two components had almost the same meanings to the case of the tomatoes eaten alone. Graphing the first two PC scores visually separated HS2 from the others by PC2, showing that HS2 tomatoes had a liking of sweetness, umami and thickness of flavor and were firmness of the skin (Fig. 6). More than a half of HS2 samples were positioned in the left side by PC1. This means that the high-sugar-content tomatoes were not especially highly evaluated compared with others when eaten in the salad.
means and SEs of mechanical properties of various parts of the tomatoes
Means and SEs of the attributes (firmness, liking of firmness) of the tomatoes obtained from sensory evalua
■：1st piece; □：2nd piece; □：Salad.
Loadings of attributes resulting from PCA (tomato alone)
Plot of the individual tomato samples of 4kind of tomato eaten alone in the plane spanned by the first two principal components. ○：R1, ○：R2, ●：HS1, ●：HS2.
Loadings of attributes resulting from PCA (in salad)
Correspondence analysis (CA) (Uchida, 2003) computed with the frequencies of the 4 kinds of tomato evaluated and how much people want to eat, which has been categorized as 'none,' '1 piece,' '2 pieces,' '3 pieces,' '4 or more' for each sample, shows the correspondence between the sample and how much of it people want to eat.
The first plot of CA explains 95.4% of the variation (Fig. 7). HS2, which had the highest concentration of taste substances including sugars, was close to 'none,' which means that the testers did not to want to eat any more after eating the 2nd piece, whereas it was close to '1 piece or 2 piece' after eating the 1st piece. The other 3 samples were close to '3 pieces' or '4 or more' after eating the 1st piece, and they were close to '1 piece' or '2 pieces' after eating the 2nd piece. This suggests that the high-sugar-content tomatoes that were very sweet and thick satiate the eater rapidly and cannot be eaten in large amounts.
Plot of the individual tomato samples of 4kind of tomato eaten in the salad in the plane spanned by the first two principal components. ○：R1, ○：R2, ●：HS1, ●：HS2.
Correspondence analysis of the association of the 4 kinds of tomato evaluated and how much people want to eat reported after evaluating the 1st and 2nd pieces of the tomato. ●：none, 1 piece, 2 pieces, 3 pieces, 4 or more; ■ 1st：after eating first piece；□ 2nd：after eating second piece.
Correspondence analysis of the association of the 4 kinds of tomato evaluated and the degree of desire for a sweeter tomato reported after evaluating the tomatoes. ●：always, occasionally, never; ■：the four kinds of tomato.
CA computed with the frequencies of the 4 kinds of tomato and the degree of desire for a sweeter tomato reported after evaluating the tomatoes, which has been categorized as 'always,' 'occasionally,' and 'never' for each sample, shows the correspondence between the samples and the desire for a sweeter tomato.
The first plot of CA explains 100% of the variations (Fig. 8). The regular tomatoes, R1 and R2, are close to 'always,' high-sugar-content tomato HS1 is close to 'neutral,' and the more typical high-sugar-content tomato, HS2, is close to 'never.'
As mentioned above, at a conceptual level, more than 87.5% of the panelists wanted to eat a sweeter tomato and also wanted to eat it as a regular tomato in everyday meals. However, it is shown in the figure that the desire weakened after eating actually a high-sugar-content tomato. We also demonstrated that regular tomatoes, which had a rather weak taste, made the salad more palatable than the high-sugar-content tomatoes (Fig. 3).
We conclude that high-sugar-content tomatoes are strong in flavor and have a very attractive image at a conceptual level, and are sufficiently palatable when eaten in small amounts. However, it they are eaten in large amounts for everyday meals, they rapidly satisfy the eater. When they are eaten in a salad, the palatability score is also high, but the palatability is not always superior to regular tomatoes. It is important to choose tomatoes to suit the occasion such as for enjoying the flavor in a special meal or for regular use as a salad vegetable.
The authors are grateful to Director Toshihiko Nishio and Executive director Susumu Yamamoto of the Japan Speciality Agriculture Products Association for their support to acomplish this research. We also thank Section Director Ken-ichiro Kanaya of Japan Food Research Laboratories who helped us with this research. The students of the Tokyo University of Agriculture are thanked for the hard work of the sensory evaluation.
