Sensory evaluation of organic and conventional fruits and vegetables available to Irish consumers

Abstract

Many Irish consumers believe organic foods taste better; however, scientific data to support this are scarce, inconclusive and limited in scope. Quantitative descriptive analysis was used to compare the sensory properties of nine organic and conventional fruits and vegetables. A panel of nine was trained according to ISO international standards (8586-1, 1993). Independent t-tests of individual descriptor data and a two-way repeated measures Anova of the entire data set were carried out via SPSS 19. Statistical analysis of panel scores failed to show any significant differences between the organic and conventional samples for any of the descriptors assessed (P < 0.05 and less) or between the overall organic and conventional data sets (P < 0.05). In conclusion, the study found no statistically significant differences between the sensory attributes of a range of organic and conventional fruits and vegetables available to the Irish consumer.

Fruits, Ireland, organic food, sensory evaluation, vegetables

Introduction

Foods are generally perceived by appearance, odour/fragrance, consistency/texture and finally flavour; however, in reality, these perceptions often occur in parallel as an overall sensory experience. Sensory evaluation attempts to isolate the sensory properties of foods and other products and can be defined as a scientific method that is used to evoke, measure and interpret responses to the sensory properties through the senses (Lawless & Heymann, 2010). Sensory evaluation assesses these properties using trained panellists or consumer groups for the purpose of rating the quality of a product, or comparing one product to another (Meilgaard et al., 2007). In the last 20–30 years, there has been a significant development in the use of sensory evaluations in the food and beverage industry. These developments have grown from a better recognition of human perception and developments in statistical tools. Sensory evaluation can be an invaluable tool in product development and product improvement, such as the assessment of the effects of formulation changes to the sensory properties. Evaluations can also be used for comparisons to competitor products and also shelf-life establishment (Kemp et al., 2009).

Descriptive analysis is the detection and description of sensory properties of a product; these properties combine to define the sensory profile of the product. Products may have the same sensory profile but may differ in the intensity of the properties; these intensities can be numerically scored, which allows for a number of products to be statistically compared (Meilgaard et al., 2007).

Organic foods can be generally defined as foods that are produced under strict regulations using an agricultural management system that promotes and enhances biodiversity, biological cycles and soil health with a minimum use of off-farm inputs, Organic Trade Association (2011). A 2008 survey by An Bord Bia showed that 27% of Irish consumers believe organic foods taste better, but could these consumers distinguish organic and conventional foods in a blind study? Is it possible that the perception of organic food as a more natural and safer product influences the sensory experience? (Tobin et al., 2011). Scientific studies on the question are limited and cover few food types. A review conducted in 2004 by Trewavas stated that research findings are inconsistent and added that freshness is an important factor often confused with organic because of local sourcing. Table 1 lists some of the most recent sensory comparisons of organic and conventional foods and their findings.

Table 1

Recent studies on the sensory comparison of organic and conventional foods and their findings

References	Foods tested	Findings
Haglund et al. (1999)	Carrots	Conventional carrots higher in carrot taste, sweetness and crunchiness. Organic carrots higher in hardness and pronounced after-taste (P < 0.01 or less).
Wszelaki et al. (2005)	Red skin potatoes	In a triangle test, panellists could distinguish between organic and conventional samples only when the skin was left on.
Gilsenan et al. (2008)	Carrots and Mushrooms	Descriptive analysis of carrots for appearance, aroma, texture and taste found no significant difference (P < 0.05). Analysis of mushrooms for the same descriptors also showed no significant difference; however, it was indicated that organic mushrooms had darker gills and a stronger aroma (P < 0.05).
Gilsenan et al. (2010)	Potatoes	No significant difference in appearance, aroma and taste was observed. However, baked conventional samples were perceived to be significantly softer, less adhesive and wetter than organic baked samples (P < 0.05).
Hajslovà et al. (2005)	Potatoes	In a 4-year study, differences were seen within single crop years; however, pooled results showed that year-to-year, variety and geographical variations were equal or more important factors.

Another survey of Irish consumers by An Bord Bia showed that 21% of organic rejecters state that they do not buy organic foods because they do not know what the benefits are. This would clearly suggest that more scientific research must be carried out to inform consumers of the differences, if any, associated with organically produced foods. In a similar survey carried out by An Bord Bia, organic consumers were asked where they bought their organic products. The results showed that a large majority visit large supermarket chains and buy their organic products along with their general purchases (An Bord Bia, 2008). Many previous sensory studies on the comparison of organic and conventional foods included samples collected directly from farms only. It was also noted that, to a consumer, variations such as soil type, fertilisation and geographical factors are unknown and perhaps unimportant. Therefore, the aim of this research was to provide a consumer-relevant study of organic fruits and vegetables and their conventional counter parts to show whether Irish consumers purchasing organic fruits and vegetables receive a higher sensory quality product or whether no difference exists. The foods included in the study were also chosen based on An Bord Bia survey data on the most popular and frequently purchased organic products (An Bord Bia, 2008).

Materials and methods

Nine fruits and vegetables were tested during the study, namely carrots (raw and cooked), onions, broccoli, vine tomatoes, cherry tomatoes, apples, potatoes, bananas and oranges. Of the fifty-four sample batches purchased for the study, 96% were sourced from large supermarkets (e.g. Tesco, Lidl) and 4% were sampled from a local organic producer to account for a minority of consumers visiting local organic markets.

Sample preparation

Six sample batches of each food (three organic and three conventional) were purchased simultaneously the day before analysis to avoid differences because of degrees of freshness. Preparation was carried out on the morning of analysis in an adjoining preparation kitchen and depended on the nature of the samples (Table 2 shows the preparation carried out for each food type). All samples were washed before preparation. Cooked samples were steamed simultaneously in a segregated household electric steamer to avoid variations because of cooking times and temperatures.

Table 2

Sample preparation

Food	Preparations	Cooking
Carrots (raw)	Ends removed, peeled, chopped in to batons	N/A
Carrots (cooked)	Ends removed, peeled, chopped in to batons	Steamed for 15 min
Onions	Skin removed, sliced	Steamed for 10 min
Broccoli	Uniform florets prepared	Steamed for 15 min
Vine tomato	Vine removed, chopped in to segments	N/A
Cherry tomato	Vine removed, chopped in half	N/A
Apple	Stalk removed, chopped into segments	N/A
Potato	Peeled, chopped into cubes	Steamed for 15 min
Banana	Peeled and broken in half	N/A
Orange	Chopped into segments	N/A

Panel training and selection

The sensory panel was recruited in January 2011. The criteria for panellists were the following: they were available to volunteer for the project, nonsmokers, in good health and regular consumers of the samples included in the study. Participants who were willing to volunteer then completed a consent form and diet survey. Training in sensory analysis principals and quantitative descriptive analysis commenced with ten participants according to ISO international standards (8586-1, 1993). Panellists were trained over 3 months (2 h per week) on their ability to differentiate between basic taste solutions at various levels, aromas and colours. The next stage of training was to describe and discuss descriptors in fruits and vegetables in terms of appearance, aroma, texture and taste. The panel was evaluated for correct detection of basic taste solutions and consistent scoring of descriptors of fruit and vegetable reference samples. Of the ten participants, nine (two men and seven women) panellists performed sufficiently in these evaluations and proceeded to analysis.

Sensory analysis

For each evaluation, the panel would decide on a range of appropriate descriptors that should be used to assess and compare the test samples. This was done using a reference sample for each fruit/vegetable. This session acted as a panel calibration as panellists established benchmark scores for each descriptor. The panellists' sensitivity was also tested before each evaluation using basic taste solutions. Sweet, sour, bitter, salty and savoury solutions at various levels had to be correctly identified before the panellist could sit for evaluation. This ensured all panellists were detecting changes in tastes sufficiently before scoring test samples.

For the analysis, six randomly coded prepared samples (three organic and three conventional) were placed on coded white paper plates, in no particular order, and presented to each panellist in separate white booths. The panellists used mineral water as a palate cleanser between tastings. The test room was also equipped with special white lighting and kept free from any strong smells. The panellists were instructed to assess each test sample in order of aroma intensity, visual appearance, intensity of colour, textures upon bite/chew and individual tastes observed. Each sample was scored between 1 and 9 for each of the descriptors, 1 denoting a low intensity for the descriptor and 9 denoting high intensity.

Data analysis

All panel scores were transferred to an Excel worksheet, and the means and standard deviations were calculated. The data were transferred to SPSS Statistics 19 software (SPSS Inc., Chicago, IL, USA). Independent t-tests were carried out for each fruit/vegetable for each descriptor to determine whether the samples scored significantly different for any individual attribute. The use of multiple t-tests can inflate the probability of type I experiment-wise error (false positives) however, and hence, a Bonferroni correction was applied to ensure significant differences were true differences and not because of error. The correction lowered the significance level of the test by the formula 0.5/N where N is the total number of tests employed.

A two-way repeated measures analysis of variance (Anova) was carried out on the entire data set that consisted of ten fruit/vegetable groups (treating cooked and raw carrots as separate groups) with six samples of each (three organic and three conventional), sixty-seven individual descriptors with 201 vs. 201 mean panel scores for organic and conventional samples in total.

The null hypothesis for the statistical analysis was there is no significant difference between the sensory properties of the organic and conventional fruits and vegetables assessed (P < 0.05).

Results

The results of the quantitative descriptive analysis of the organic and conventional fruits and vegetables tested are shown in Table 3.

Table 3

Results of the quantitative descriptive analysis (QDA) of nine fruits and vegetables; mean panel scores and P-values

Descriptor	Mean Organic	Mean Conventional	P value
Carrots raw
Orange colour	5.8 ± 0.33	6.0 ± 0.64	0.715
Carrot aroma	3.1 ± 0.38	3.7 ± 0.22	0.129
Hardness (on bite)	5.6 ± 0.08	5.9 ± 0.22	0.091
Moistness (on chew)	5.5 ± 0.36	5.4 ± 0.14	0.709
Crunch (on chew)	6.1 ± 0.52	6.2 ± 0.30	0.851
Sweetness	4.4 ± 0.62	4.2 ± 0.46	0.727
Bitterness (after)	2.0 ± 0.50	1.7 ± 0.54	0.488
Carrots cooked
Orange colour	6.1 ± 0.43	6.0 ± 0.68	0.693
Carrot aroma	5.0 ± 0.36	5.0 ± 0.50	0.792
Firmness (on bite)	4.1 ± 1.00	5.2 ± 0.54	0.151
Moisture (on chew)	5.3 ± 0.14	4.8 ± 0.22	0.028
Sweetness	4.6 ± 0.30	4.8 ± 0.44	0.628
Bitterness	1.8 ± 0.82	1.6 ± 0.30	0.802
Onions
White/yellow colour	3.19 ± 0.52	3.63 ± 0.82	0.588
Onion aroma	3.38 ± 0.54	3.38 ± 0.13	1.000
Crunch (on chew)	4.45 ± 0.83	4.96 ± 0.76	0.554
Sweetness	3.50 ± 0.33	3.50 ± 0.25	0.990
Onion after-taste	4.50 ± 0.25	4.63 ± 0.63	0.761
Broccoli
Green colour	6.1 ± 0.26	6.3 ± 0.79	0.740
Compactness (head)	6.7 ± 0.25	6.8 ± 1.04	0.864
Aroma, Cabbage	5.2 ± 0.29	5.1 ± 0.42	0.873
Aroma, Savoury	3.8 ± 0.29	3.7 ± 0.44	0.608
Firmness (on bite)	5.9 ± 0.69	5.8 ± 0.29	0.807
Spongy, of head	4.9 ± 0.25	5.4 ± 0.42	0.194
Tenderness, of stalk	4.1 ± 0.35	3.8 ± 0.88	0.641
Savoury taste	3.9 ± 0.25	3.6 ± 0.39	0.215
Cabbage taste	4.9 ± 0.38	4.7 ± 0.44	0.545
Vine tomato
Red colour	6.6 ± 0.22	7.6 ± 0.41	0.021
Tomato aroma	5.9 ± 0.30	6.3 ± 0.38	0.238
Hardness (on bite)	4.7 ± 0.36	5.3 ± 0.36	0.084
Juiciness (on chew)	6.5 ± 0.44	6.3 ± 0.08	0.615
Savoury taste	4.8 ± 0.46	4.7 ± 0.25	0.763
Sweetness	5.5 ± 0.30	5.9 ± 0.46	0.246
Cherry tomato
Red colour	5.8 ± 0.36	5.0 ± 2.00	0.552
Tomato aroma	4.5 ± 0.58	4.2 ± 0.22	0.542
Hardness (on bite)	5.5 ± 1.15	5.9 ± 1.12	0.737
Juiciness (on chew)	6.1 ± 0.46	5.8 ± 0.22	0.317
Savoury taste	5.6 ± 0.36	5.5 ± 0.68	0.843
Sweetness	5.4 ± 0.76	4.9 ± 0.74	0.405
Apple
Green skin colour	4.8 ± 0.36	5.1 ± 1.25	0.682
Apple aroma	3.3 ± 0.43	2.9 ± 0.65	0.393
Hardness (on bite)	4.2 ± 0.79	5.5 ± 0.50	0.075
Juiciness (on chew)	5.5 ± 0.22	5.4 ± 1.08	0.945
Leathery skin (after)	4.6 ± 1.23	5.8 ± 0.81	0.235
Sourness	4.4 ± 0.44	3.9 ± 0.25	0.144
Sweetness	4.3 ± 0.16	5.4 ± 0.36	0.010
Apple flavour	5.9 ± 0.30	6.1 ± 0.16	0.507
Potato
White/yellow colour	5.0 ± 0.66	4.1 ± 0.25	0.087
Potato aroma	4.1 ± 0.36	4.3 ± 0.19	0.374
Soapy (waxy on chew)	5.9 ± 0.51	5.3 ± 0.32	0.135
Floury (dry on chew)	3.5 ± 0.32	4.8 ± 0.62	0.032
Potato taste	4.5 ± 0.29	5.0 ± 0.47	0.127
Sweetness	3.2 ± 0.26	3.5 ± 0.57	0.447
Astringency (dry after)	4.8 ± 0.36	4.5 ± 0.65	0.539
Banana
Pale yellow colour	4.3 ± 0.00	4.2 ± 0.08	0.116
Banana aroma	5.9 ± 0.54	5.6 ± 0.84	0.644
Firmness (on bite)	4.6 ± 0.52	4.4 ± 0.73	0.732
Astringency/cloyness	6.1 ± 0.22	5.7 ± 0.33	0.171
Sweetness	5.1 ± 0.52	5.4 ± 0.50	0.594
Banana flavour	5.5 ± 0.68	6.0 ± 0.29	0.283
Orange
Orange colour	5.7 ± 0.38	5.4 ± 0.79	0.492
Orange aroma	5.7 ± 0.52	5.8 ± 0.22	0.705
Juiciness (on chew)	6.6 ± 0.73	7.0 ± 0.22	0.485
Fibrosity (after)	6.3 ± 1.36	5.8 ± 1.07	0.690
Sour/acidity	3.6 ± 1.84	3.0 ± 0.86	0.599
Sweetness	4.9 ± 1.89	5.2 ± 1.08	0.861
Orange flavour	5.0 ± 0.38	5.3 ± 0.29	0.330

Discussion

The raw carrots were scored based on the following descriptors: intensity of orange colour, carrot aroma, hardness on bite, moistness on chew, crunchiness on chew, sweetness and bitterness (after-taste). Organic samples scored higher for moistness, sweetness and bitterness. Conventional samples scored higher in all other descriptors; however, differences recorded were not statistically significant (P < 0.05). Gilsenan et al. (2008) also found no significant difference (P < 0.05) between organic and conventional carrots. However, the study by Haglund et al. (1999) found conventional carrots scored significantly higher in sweetness and crunchiness and organic carrots scored higher in hardness and after-taste (P < 0.01), none of which were observed in this study.

Cooked carrots were scored based on the following descriptors: intensity of orange colour, carrot aroma, firmness on bite, moistness on chew, sweetness and bitterness (after-taste). Both sample sets scored equally for aroma and conventional samples scored higher in sweetness, albeit not statistically higher. The organic samples scored higher for colour, moistness, bitterness and firmness. Organic samples were calculated to be significantly higher in moistness at the P < 0.05 significance level; however, when the Bonferroni correction was employed, the significance level was reduced to P = 0.0083, which deemed the P value as no longer significant, and the null hypothesis was then accepted.

Onions were compared based on the following descriptors: intensity of white/yellow colour, onion aroma, crunchiness on chew, sweetness and onion after-taste. Organic samples scored equal for sweetness and lower than conventional samples for all other attributes. None of the differences recorded had statistical significance at the P < 0.05 significance level.

Broccoli samples were scored based on the following descriptors: intensity of green colour of flourette, compactness of flourette, aroma (cabbage-like), aroma (savoury), firmness (overall), sponginess (head), tenderness (stalk), savoury taste and cabbage taste. Organic samples scored slightly higher for cabbage-like aroma, overall firmness, tenderness, savoury taste and cabbage taste. None of the scores were calculated to be statistically significant at the P < 0.05 significance level.

Vine tomatoes were scored based on the following descriptors: intensity of red colour, tomato aroma, hardness on bite, juiciness on chew, savoury taste and sweetness. Organic samples scored just slightly higher in juiciness and savoury taste. Conventional samples scored higher for all other attributes of which colour was calculated to be significantly higher at the P < 0.05 significance level. The Bonferroni correction lowered the significance level to 0.0083, which then excluded the P value from the significance level, and hence, the null hypothesis was no longer rejected.

Cherry tomatoes were analysed for the following descriptors: intensity of red colour, tomato aroma, hardness on bite, juiciness on chew, savoury taste and sweetness. Organic samples scored higher for colour, aroma, juiciness, savoury taste and sweetness. However, none of the differences observed were calculated to have statistical significance (P < 0.05).

Apples were scored based on their intensity of skin colour, apple aroma, hardness on bite, juiciness on chew, leathery skin, sourness, sweetness and apple flavour. Organic samples scored higher in aroma, juiciness and sourness. Conventional samples were higher in all other attributes of which sweetness was calculated to have statistical significance at the P < 0.05 significance level. After the Bonferroni correction, the adjusted significance level of 0.0063 excluded the P value for apple sweetness and the null hypothesis was no longer rejected.

Potatoes were analysed for the following descriptors: intensity of white/yellow colour, potato aroma, soapiness (waxiness), flouriness (dryness/fluffiness), potato taste, sweetness and astringency (drying after-taste). Organic samples scored higher for colour, soapiness and astringency. Conventional samples scored higher for all other descriptors of which flouriness was calculated to be statistically higher (P < 0.05). The Bonferroni correction lowered the level of significance to 0.0071, and hence, the null hypothesis was no longer rejected. Similarly to studies by Wszelaki et al. (2005) and Gilsenan et al. (2010), the panel did not distinguish between the organic and conventional potato samples. Hajslovà et al. (2005) did observe differences but noted other factors as more or equally important to growing method.

The panel assessed the banana samples for intensity of pale yellow colour, banana aroma, firmness on bite, astringency/cloyness, sweetness and banana flavour. Organic samples scored slightly higher for colour, aroma, firmness and astringency/cloyness; however, none of the differences recorded had statistical significance (P < 0.05).

Oranges were scored based on intensity of orange colour, orange aroma, juiciness on chew, fibrosity on chew, sourness/acidity, sweetness and orange flavour. Organic samples scored higher in colour, fibrosity and sour/acidity. Conventional samples scored higher in all other attributes; however, none of the differences observed had statistical significance (P < 0.05).

Independent t-tests (with Bonferroni corrections) of panel scores for individual descriptors found no significant difference (P < 0.05 or less) between the organic and conventional samples. The two-way repeated measures Anova analysed for variance between all organic and all conventional scores (201 vs. 201 mean panel scores) for sixty-seven descriptors assessed. The mean organic score was 4.918, and the mean conventional score was 4.965. The significance value of the within-subjects effects between organic and conventional data sets was 0.475, which indicates that there is no significant variance between all organic and conventional scores (P < 0.05).

It is clear that in this the study, panellists could not significantly distinguish between the organic and conventional fruits/vegetables, which may suggest that opinions on sensory differences in organic foods could be due to a placebo-like effect when it is known that the food is organic rather than a sensory difference that is measurable. It could also be said that differences previously observed by consumers may have been due to differences in freshness as organic foods are often locally sourced and sometimes purchased directly from producers at local food markets, as previously suggested by Trewavas (2004). The results obtained do not strengthen the opinion of some consumers that organic foods taste better and are in agreement with previous studies of carrots and potatoes by Wszelaki et al. (2005), Hajslovà et al. (2005), Gilsenan et al. (2008, 2010).

There is, however, still the question of any health benefits, contaminant levels and environmental benefits of organic foods. Research in these areas has yet to provide clear-cut answers on the benefits of organic foods, if any, but answers are needed for organic market growth and consumer knowledge and awareness.

Further publications are currently being prepared by the researchers, which will include an in-depth biochemical (secondary metabolites) and contaminant analysis, with an aim to determine whether a difference exists between organic and conventional fruits and vegetables in these areas.

Conclusion

The results of the research showed that the sensory panel of nine using the quantitative descriptive analysis method detected.

• no statistically significant differences (P < 0.05 or less) between individual descriptors of a range of organic and conventional fruits and vegetables via independent t-tests of panel scores.

• no significant difference (P < 005), or variance, between all organic and all conventional panel scores via a two-way repeated measures Anova.

In conclusion, the research found no statistically significant differences between the sensory properties of a range of organically and conventionally grown fresh fruits and vegetables available to the Irish consumer.

Acknowledgments

This work is supported by the Irish Research Council for Science, Engineering and Technology (IRCSET) and facilitated by Shannon Applied Biotechnology Center/Sensory Evaluation Centre.