- Email: [email protected]
Consumers' segmentation based on the acceptability of meat from entire male pigs with different boar taint levels in four European countries: France, Italy, Spain and United Kingdom
Meat Science 114 (2016) 137–145
Contents lists available at ScienceDirect
Meat Science journal homepage: www.elsevier.com/locate/meatsci
Consumers' segmentation based on the acceptability of meat from entire male pigs with different boar taint levels in four European countries: France, Italy, Spain and United Kingdom N. Panella-Riera a,⁎,1, M. Blanch a,1, Z. Kallas b, P. Chevillon c, A. Garavaldi d, M. Gil a, J.M. Gil d, M. Font-i-Furnols a, M.A. Oliver a a
IRTA-Monells, Finca Camps i Armet, 17121 Monells, Spain CREDA, Parc Mediterrani de la Tecnologia, 08860 Castelldefels, Spain c IFIP, La Motte au Vicomte, 35651 Le Rheu, France d CRPA, Viale Timavo 43/2, 42121 Reggio Emilia, Italy b
a r t i c l e
i n f o
Article history: Received 21 April 2015 Received in revised form 17 December 2015 Accepted 23 December 2015 Available online 28 December 2015 Keywords: Acceptability Androstenone Boar taint Consumers Clusters Skatole
a b s t r a c t Two consumer studies were conducted to know the acceptability of pork with different boar taint levels: test 1 performed in Spain (n = 126) and United Kingdom (n = 146), and test 2 performed in France (n = 139) and Italy (n = 140). Each test had 3 types of pork: ‘Female meat’, ‘Low boar tainted meat’, and a third type was ‘Medium boar tainted meat’ or ‘High boar tainted meat’. Three main clusters were identified on the basis of ‘How delicious do you find this meat?’: 1—Pork lovers, 2—Boar meat lovers, 3—Reject boar tainted meat. Additionally, in test 2, a fourth cluster was identified: ‘Reject low tainted meat’. A group of 16.2–38.2% of consumers rejected meat from boars, and another group of 12.4–21.7% rated the meat with medium or high levels of boar taint better than the meat from females, identifying a niche for meat from medium and high levels of boar taint, and suggesting the need to select carcasses on the basis of boar taint. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Nowadays there is a growing concern about the negative effect of surgical castration of pigs without anaesthesia on animal welfare. In 2008, across Europe the majority of male piglets intended for pork production were castrated to avoid potential consumer dissatisfaction because of boar taint. At that time, castration was performed in most of the EU countries on 80–100% of the male pigs in conventional production, and surgical castration without anaesthesia was the most common technique (Fredriksen et al., 2009). The exceptions were United Kingdom and Ireland, where castration was hardly performed, and some southern countries such as Cyprus, Portugal and Spain, where a limited percentage of the male pigs were castrated in comparison to the other European countries. Since then, some countries have taken an action to avoid piglet castration without anaesthesia not only for welfare reasons, but also because the production of entire male pigs decreases production costs and results in higher meat content of the carcass (EFSA, 2004; PIGCAS, 2009). Norway and Switzerland have ⁎ Corresponding author. E-mail address: [email protected] (N. Panella-Riera). 1 The first and the second authors had an equal input to the paper.
http://dx.doi.org/10.1016/j.meatsci.2015.12.017 0309-1740/© 2015 Elsevier Ltd. All rights reserved.
already banned piglet castration without anaesthesia by law, and other countries such as The Netherlands and Germany have signed letters of intentions (Declaration of Noordwijk 2007 and Düsseldorf Declaration 2008, respectively) which aimed to avoid the need for piglet castration in the long term. In 2010, at the EU level, representatives of European farmers, meat industry, retailers, scientists, veterinarians and animal welfare NGOs, committed themselves to voluntarily end the practise of surgically castrating pigs in Europe by January 2018 by means of the European Declaration on alternatives to surgical castration of pigs (DG-SANCO, 2010). Consequently, entire male production is one of the alternatives and therefore, an increase in the production of entire males would result in a positive impact on production efficiency and carcass composition. But as a consequence, more tainted carcasses could enter the meat chain. A number of consumer studies have been carried out in order to ascertain the acceptability of pork from entire male pigs in different countries and using different methodologies as was reviewed in EFSA (2004) and more recently, by Font-i-Furnols (2012). The two main compounds responsible for boar taint are androstenone (Patterson, 1968) and skatole (Vold, 1970; Walstra & Maarse, 1970). The respective contributions of these two compounds vary depending on the characteristics of the meat evaluated, procedures for preparing the meat, the
138
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
ability of consumers to perceive androstenone among others, as has been reviewed by Font-i-Furnols (2012), and also it may vary depending on the habit of eating meat from castrated or entire male pigs. Therefore, it would be interesting to study the impact that meat from boars could have in the European market (consumer satisfaction), which is traditionally used to eating meat from castrated pigs, and identify if there is any difference comparing to other countries where the utilization of entire male pigs is more common. The present work had two main objectives: 1) to evaluate and update the results on sensory acceptability of meat from entire male pigs (as an alternative to the production of castrates), involving four European countries that produce a different proportion of castrated pigs in two consumer studies: Spain and United Kingdom, France, and Italy, and 2) to identify potential niche markets for meat with different boar taint levels. 2. Materials and methods 2.1. Consumer studies Two consumer studies were conducted in which each consumer assessed pork with various levels of androstenone and skatole. Participants in each study were: ▪ Consumer test 1. Spain (ES, n = 133, performed in Barcelona) and United Kingdom (UK, n = 146, performed in Reading); April, 2009. ▪ Consumer test 2. France (FR, n = 139, performed in Cain (n = 74) & Paris (n = 70)) and Italy (IT, n = 140, performed in Matelica (n = 70) & Reggio Emilia (n = 70)); April and May 2009. Considering the two consumer tests, a total of 558 consumers from the different European countries participated in this project. They were stratified by age (according to each country profile) and sex (approximately 50:50 ratio between men and women). Consumers were required to eat pork on a regular basis.
From each animal, the muscle Longissimus lumborum (with the subcutaneous fat) was taken from the 1st lumbar vertebra to the last rib, divided in two sections (min. 15 cm) and distributed in one of the two countries. 2.3. Androstenone and skatole measurements in fat Back fat samples from entire male pigs and females were collected on the day of slaughter at the level of the neck, vacuum packed and deep frozen until analysed for androstenone (AND) and skatole (SKA). The determination of SKA levels was performed using HPLC-FLD and the determination of AND levels using GC-MS (for ES and UK samples; Ampuero et al., 2011) or by HPLC-FLD (for FR and IT samples; Pauly, Spring, O'Doherty, Ampuero, and Bee, 2008). Results were expressed as μg/g, on pure fat basis. Meat samples were classified as FE (meat from gilts), Low boar taint (LBT), Medium boar taint (MBT) or High boar taint (HBT) depending on the sex and the levels of boar taint compounds (Blanch et al., 2012; Table 1, Fig. 1). In each consumer test 3 samples were evaluated in each session: samples from gilts (FE) as a reference, and two more samples with different levels of boar taint: in consumer test 1 (ES & UK) meat from LBT and MTB was used; and in consumer test 2 (FR & IT), meat from LBT and HBT. 2.4. Preparation of samples In each location, loins were defrosted at 2–4 °C during 24 h, and cut into 0.5 cm thick slices with 5 mm of subcutaneous fat (when it was possible). Each slice was cut in two pieces, and cooked using a cooking plate at 180 °C (which was greased with maize oil). A different cooking plate was used for each type of meat presented within a session. The meat was turned upside down regularly up to a core temperature of 80 °C measured with a flexible temperature probe (TESTO, Germany) and the meat was salted after cooking (1 g salt/200 g meat), reproducing home preparation. Samples were then immediately served to the consumers for evaluation.
2.2. Animals and samples 2.5. Sensory evaluation of samples The meat and fat samples used for the consumer studies were obtained from populations of pigs slaughtered according to commercial practise. They were collected from boars and gilts in commercial abattoirs from Spain and France. Carcass weights and genotypes were representative of current practise in these countries (carcass weight of 79.4 kg in Spain, and 84.0 kg in France; (2009)). The meat collected in Spain was used for the consumer studies carried out in ES and UK; and the meat collected in France was used for the consumer studies carried out in FR and IT. A minimum 20 loins from gilts and 20 loins from boars with Low boar taint (LBT) were collected in Spanish and French abattoirs. Furthermore, a minimum of 20 loins from boars with Medium boar taint (MBT) were collected in Spanish abattoirs and a minimum of 20 loins with High boar taint (HBT) were selected in French abattoirs (Table 1).
Sessions of 10–12 consumers were organized for meat evaluation. Each consumer assessed 3 pieces of meat. A 10 min interval was used between each of the three samples presented within a session. The order of presentation of samples was rotated using a partial Latin square design to avoid any first sample and carry-over effect and the identity of the samples was not given to consumers (Macfie, Bratchell, Greenhoff, & Vallis, 1989). Consumers were asked for hedonic questions (questions 1 to 4) and intensity evaluations (questions 5–11) as described in Table 2, using a modified nine-point intensity scale (from 1 to 9). The intermediate level (5) was not included to stimulate consumers to commit themselves and not to allow a neutral assessment which is not informative (Guerrero, 1999). Table 2 also shows the scale used to assess each attribute.
Table 1 Mean and standard error of the androstenone and skatole levels on the pure fat basis of the meat used for the consumer test. N of loins
Androstenone (ppm pure fat) [range]
Skatole (ppm pure fat) [range]
Consumer test 1: ES & UK Gilt (FE) Low boar taint (LBT) Medium boar taint (MBT)
18 18 18
b0.04 0.20 ± 0.07 [0.04–0.29] 1.07 ± 0.40 [0.58–2.28]
0.04 ± 0.02 [0.02–0.07] 0.06 ± 0.02 [0.02–0.08] 0.18 ± 0.07 [0.11–0.39]
Consumer test 2: FR & IT Gilt (FE) Low boar taint (LBT) High boar taint (HBT)
18 18 18
b0.2 b0.2 2.39 ± 1.07 [0.59–5.18]
b0.03 b0.03 0.11 ± 0.07 [0.02–0.28]
ES: Spain; UK: United Kingdom; FR: France; IT: Italy.
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
139
eat the pork with the fat), FREQ and GENMOD procedures were applied. The MIXED procedure was used to analyse the acceptability of odour and flavour by consumers: the model included the type of animal, country and their interaction as fixed effects, session as blocking variable and consumer as random effect. Interaction was kept in the model only if it was significant (P b 0.05). Differences were considered at P b 0.05 and tendencies at P b 0.10. In order to establish different clusters of consumers a hierarchical cluster analysis was performed with the CLUSTER procedure and the Ward method. The cluster analysis was conducted on the basis of the attribute ‘delicious’. For each cluster, the MIXED procedure was then used to analyse the hedonic attributes (delicious, odour, taste, and in-mouth feeling) and the intensity attributes (strength of odour and strength of taste, abnormal odour and abnormal taste) scored by the consumers. The model included the type of animal and country as fixed effects, session as blocking effect and consumer as random effect. The interaction type of animal and country was removed from the model because it was not significant (P N 0.05). Differences were also declared at P b 0.05. 3. Results 3.1. Description of consumers Table 3 shows the description of consumers according to the gender, age and educational level. 3.2. Habits of European consumers regarding pork consumption Fig. 1. Androstenone and skatole levels of the meat used in the two consumer studies.
2.6. Description of EU pork consumers At the end of the test, consumers also answered socio-demographic questions and frequency of consumption of different pork products, the most common purchasing place for fresh pork meat, if they were responsible for buying fresh pork at home, if they were responsible for cooking at home, and if they usually eat the pork with the fat.
2.7. Statistical analysis Data analyses were conducted using SAS Statistical Package (SAS Inst., Inc., Cary, NC, USA, version 9.2). To analyse the parameters regarding classification of consumers (age, gender, educational level, frequency of consumption of different pork products, the most common purchasing place for fresh pork meat, if they were responsible for buying fresh pork at home, if they were responsible for cooking at home, and if they usually
In general, over 90% of consumers ate fresh pork N 2 times/week (ES 95.5%; UK 97.3%; IT 92.9%) except for FR (34.8%) (Table 4). The most consumed product was dry cured ham in ES and IT, cooked ham in FR, and sliced bacon in UK (Fig. 2). In general, the percentage of respondents responsible for buying fresh pork in their household was 86.4% (Table 4). In all countries, women were more responsible for buying fresh pork than men, and they were mostly between 41 and 60 years old. Eighty-eight percent of respondents were partially responsible for cooking at home. Women were more responsible for cooking at home than men. France was an exception, where percentages were very similar 49.6% women and 50.4% men cooked at home. Considering all respondents, 43.5% were used to eat the pork with the fat in all the countries (36.4% of women and 53.1% of men). Considering the place of purchasing fresh pork meat, the supermarket was the most common one, followed by the butcher and finally the traditional market. Table 3 Description of consumers according to gender, age and educational level.
Table 2 Questionnaire and scoring scales for the evaluation of meat samples during eating. Questions Hedonic evaluations 1. How delicious do you find this product? 2. What do you think of the odour? 3. What do you think of the taste? 4. What do you think of the in-mouth feeling? Intensity evaluations 5. Strength of odour 6. Abnormal odour 7. Strength of taste 8. Salty taste 9. Meaty taste 10. Fatty taste 11. Abnormal taste a
All scoring scales were from 1 to 9, without 5.
Consumer test 1 n = 279
Scoring scalesa 1. Not good at all 9. Really delicious 1. Not pleasant at all 9. Very pleasant
1. Barely perceptible 9. Very strong
N Gender (%) Male Female Age (%) 18–25 26–40 41–60 N60 Educational level (%) Primary studies not completed Primary studies Secondary studies University studies
Consumer test 2 n = 279
ES
UK
FR
IT
133
146
139
140
Total
558
54.9 45.1
50.0 50.0
49.6 50.4
74.3 25.7
42.8 57.2
10.5 30.8 43.6 15.0
19.9 19.2 39.0 21.9
15.1 28.8 41.0 15.1
10.0 25.7 60.0 4.3
14.0 26.0 45.9 14.2
3.0 12.1 53.8 31.1
0.0 4.8 61.6 33.6
11.7 3.6 46.0 38.7
2.9 25.0 29.4 42.6
4.4 11.3 47.9 36.5
ES: Spain; UK: United Kingdom; FR: France; IT: Italy.
140
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
Table 4 Frequency (%) of consumption of fresh pork meat and behaviour of the consumers of each country participant in the study. Consumer test 1
Consumer test 2
ES
FR
UK
Table 5 Significance levels within a consumer test: consumer test 1 carried out in Spain and United Kingdom, and consumer test 2 carried out in France and Italy.
Total
IT
n 133 146 139 140 558 How many times a week do you eat fresh pork meat? 1 time per week 0.0 0.0 29.7 1.4 7.7 2 times per week 4.5 2.7 35.5 5.7 12.1 More than 2 times per week 95.5 97.3 34.8 92.9 80.2 Are you partially responsible of doing the shopping of fresh pork eaten in your household? Yes 78.9 81.5 93.5 91.4 86.4 Are you partially responsible for cooking in your household? Yes 80.5 82.9 97.8 91.4 88.2 When you eat the pork meat, do you eat it… without the fat 58.6 43.8 61.2 62.9 56.5 with the fat 41.4 56.2 38.8 37.1 43.5 Where do you buy fresh pork meat? (multiple choice question) At the butcher 58.6 21.9 15.8 14.3 27.2 At the supermarket 72.2 95.9 82 94.3 86.4 At the traditional market 42.1 4.8 0.7 1.4 11.8 ES: Spain; UK: United Kingdom; FR: France; IT: Italy.
3.3. Sensory evaluation The hedonic evaluation of in-mouth feeling and of the intensity of salty, meaty and fatty tastes did not result in any consistent difference between meat sample groups and therefore it is not presented.
3.3.1. Differences between meat sample groups, within experiments Table 5 shows significance levels within a consumer test: Consumer test 1 carried out in Spain and United Kingdom, and Consumer test 2 carried out in France and Italy. Results showed no significant interaction between the boar taint level and the country in each consumer test. In consumer test 1 (ES & UK), significant differences (P b 0.05) were observed for the hedonic attributes assessed by the consumers (delicious, odour and taste). In all the cases, the score given to the meat from FE and MBT was significantly higher (more favourable)
Hedonic evaluations Delicious Odour Taste Intensity evaluations Strength of odour Abnormal odour Strength of taste Abnormal taste
Consumer test 1 (ES & UK) n = 279
Consumer test 2 (FR & IT) n = 279
BT
C
BTxC
BT
C
BTxC
** *** ***
*** *** **
ns ns ns
† *** *
ns ns ns
ns † ns
*** ns * ns
*** ns † **
ns ns ns ns
ns ns ** *
ns ns ns ns
ns ns ns ns
BT: type of meat; C: country. ns: P N 0.10; +: P b 0.10; *: P b 0.05; **: P b 0.01; ***: P b 0.001.
than the one given to the meat from LBT (Table 6). With regard to the intensity evaluations, significant differences were observed between type of meat when considering strength of odour and strength of taste (P b 0.05) whereas no differences were observed when taking into account the attributes abnormal odour and abnormal taste. Results showed that the strongest odour was observed in MBT and FE meat, and the highest strength of taste was observed in MBT in comparison to LBT. In consumer test 2 (FR & IT), significant differences (P b 0.05) were observed for the hedonic odour and taste (Table 7). For all of them, HBT meat was rated worst (lower scores) than FE meat. Considering the intensity scores, significant differences (P b 0.05) were observed for the attributes strength of taste and abnormal taste. In both cases, HBT received the highest score (strongest taste and abnormal taste). 3.3.2. Consumer segmentation in consumer test 1: Spain and United Kingdom Due to the fact that the interaction between type of meat and country was not significant, results are presented considering consumers from the two countries together. A total of three clusters were identified
Fig. 2. Frequency (%) of consumption of different pork products by the consumers from each country who participated in the study.
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145 Table 6 Least square means (LSM) and standard error (S.E.) in consumer test 1 (ES & UK) of the scores given by consumers to meat from gilts (FE), from entire males with low levels of skatole and androstenone (LBT), and entire male pigs with medium levels of skatole and androstenone (MBT). FE
Hedonic evaluations Delicious Odour Taste Intensity evaluations Strength of odour Abnormal odour Strength of taste Abnormal taste 1
LBT
Sig1
MBT
LSM
S.E.
LSM
S.E.
LSM
S.E.
6.4a 6.5a 6.5a
0.11 0.10 0.11
5.9b 6.0b 6.0b
0.12 0.10 0.11
6.4a 6.5a 6.5a
0.12 0.10 0.11
0.0015 0.0003 0.0005
4.8a 2.6 5.7ab 2.6
0.12 0.11 0.10 0.11
4.3b 2.7 5.5b 2.7
0.12 0.11 0.10 0.11
4.8a 2.6 5.9a 2.7
0.12 0.11 0.11 0.11
0.0003 0.6171 0.0197 0.4422
Different letters within a line indicate significant differences (P b 0.05).
on the basis of ‘how delicious do you find this meat’. Fig. 3 shows least square means of the scores that consumers gave to each attribute (hedonic attributes: delicious, odour and taste; intensity attributes: strength of odour and strength of taste) and each type of meat (FE, LBT, MBT), for each cluster identified: ▪ The first cluster (n = 169, 62.1% of consumers; 57.4% from Barcelona and 42.6% from Reading) comprises respondents that gave high scores to all types of samples. These consumers were labelled as ‘Pork lovers’. ▪ The second cluster (n = 59, 21.7% of consumers; 30.5% from Barcelona and 69.5% from Reading) comprises respondents that like the boar taint and therefore, the higher the level of boar taint, the higher the score to the attribute. These consumers were considered as ‘Boar meat lovers’. ▪ The third cluster (n = 44, 16.2% of consumers; 25.0% from Barcelona and 75.0% from Reading) comprises respondents that did not like the boar taint and therefore, the higher the level of boar taint, the lower the score to the attribute ‘Delicious’. These consumers were labelled as ‘Reject boar tainted meat’. The three clusters did not display any relevant difference on the basis of demographic variables (age, gender and educational level) and consumers' behaviour such as frequency of consumption of fresh meat (Table 4), pork products (Fig. 2), habits related to doing the shopping, cooking at home, eating meat with or without fat and place of purchasing fresh pork (Table 4). 3.3.3. Consumer segmentation in consumer test 2: France and Italy As observed in consumer test 1, due to the fact that the interaction between type of meat and country was not significant, results are presented considering consumers from the two countries together. In this consumer test, a total of four clusters were identified on the basis of ‘how delicious do you find this meat’ (Fig. 4). Table 7 Least square means (LSM) and standard error (S.E.) in consumer test 2 (FR & IT) of the scores given by consumers to meat from gilts (FE), from entire males with low levels of skatole and androstenone (LBT), and entire male pigs with high boar taint (HBT). FE
Hedonic evaluations Delicious Odour Taste Intensity evaluations Strength of odour Abnormal odour Strength of taste Abnormal taste 1
LBT
Sig1
HBT
LSM
S.E.
LSM
S.E.
LSM
S.E.
5.7 5.9a 5.8a
0.11 0.11 0.11
5.6 5.9a 5.7ab
0.11 0.11 0.11
5.4 5.4b 5.4b
0.11 0.11 0.11
0.0518 0.0002 0.0153
4.4 2.3 4.8ab 2.2b
0.12 0.11 0.11 0.12
4.5 2.4 4.7b 2.5ab
0.12 0.11 0.11 0.12
4.5 2.5 5.1a 2.6a
0.12 0.11 0.11 0.12
0.3047 0.1537 0.0102 0.0309
Different letters within a line indicate significant differences (P b 0.05).
141
▪ The first cluster (n = 136, 49.5% of consumers; 57.4% from France and 42.6% from Italy) comprises respondents that gave high scores to all types of samples. These consumers were labelled as ‘Pork lovers’. ▪ The second cluster (n = 34, 12.4% of consumers; 23.5% from France and 76.5% from Italy) comprises respondents that liked the boar taint and therefore, the higher the level of boar taint, the higher the score to the attribute. These consumers were considered as ‘Boar meat lovers’. ▪ The third cluster (n = 55, 20.0% of consumers; 49.1% from France and 50.9% from Italy) comprises respondents that did not like the HBT meat and therefore, the higher the level of boar taint, the lower the score to the attribute ‘Delicious’. These consumers were labelled as ‘Reject boar tainted meat’. ▪ The fourth cluster (n = 50, 18.2% of consumers; 52.0% from France and 48.0% from Italy) comprises respondents that gave low scores to the meat with LBT. These consumers were labelled as ‘Reject low boar tainted meat’. Like in consumer test 1, the four clusters did not display any relevant difference on the basis of demographic variables (age, gender and educational level) and consumers' behaviour (Table 4 and Fig. 3). 4. Discussion In the present paper, consumer studies were carried out with the aim to ascertain consumer acceptability of boar meat with different levels of boar taint. Due to practical reasons, meat was collected in Spain and France, chemically analysed in terms of androstenone and skatole levels, and then distributed to two other countries: meat from Spain was used for the consumer test performed in Spain and United Kingdom, and meat from France was used for the consumer test carried out in France and Italy. It is interesting to highlight that, consumer test 1 was planned in two countries where, as pointed out by Fredriksen et al. (2009), consumers are used to the commercialization of meat from entire male pigs, whereas consumer test 2 was carried out in two countries where consumers are not used to this type of meat: in 2008, the percentage of production of castrated male pigs was estimated as 33.2% in ES, 2.1% in UK, 97.5% in FR and 100% in IT. It is also interesting to underline that in the two studies consumers could taste meat from entire male pigs with such low levels of boar taint comparable to the levels found in meat from gilts (Bonneau & Chevillon, 2012; type of meat: LBT – low boar taint – and FE ‐ meat from gilts). Additionally, each study had a third type of meat with medium levels (MBT in study 1) or high levels (HBT in study 2) of boar taint. 4.1. Methodology used As reviewed by Font-i-Furnols (2012), the methodologies used in the already published papers are very diverse and therefore, it is difficult to compare the results among them. Issues such as type of meat (loin slices, chops, cutlets, minced meat, fat, bacon, dry cured ham...), location of the test (hall or home test), cooking procedure (including cooking device, cooking time and temperature), type of meat samples (meat from gilts, castrated pigs and/or from entire male pigs), level of boar taint in the assessed meat, and the type of attributes and scales used during the consumer test are only some examples. As stated in the methodology section, in this paper we used meat from the Longissimus lumborum muscle from the 1st lumbar vertebra to the last rib. We served loin samples due to the fact that loin cuts are widely sold and consumed. On the day of the consumer test, the meat was cut into 0.5 cm thick slices with a maximum of 5 mm of subcutaneous fat. The acceptance of boar meat is assumed to be dependent on the sample preparation and presentation. In this study, meat samples were cooked on a hot plate and were served uncovered to the participants to simulate a meal at home. This kind of preparation is not very
142
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
Fig. 3. Least square means for hedonic traits and strength of odour/taste given by consumers of each of the three clusters observed in consumer test 1 (Spanish and English consumers).
common in previously published papers although it is very similar to the type of preparation that consumers usually do at home. Since AND and SKA are predominantly released during heating (Lunde et al., 2008), the intensity of these volatile compounds is expected to be higher during the preparation than at consumption. In our study, consumers were not involved in the cooking process. Therefore, the fact that the meat came from a very lean piece of meat, the thickness of the sample and the fact that it was served uncovered, could minimize the perception of boar taint. Regarding the androstenone and skatole levels of the samples used in this study, the highest level of boar taint was different in each consumer test: consumer test 1 had a group with medium levels whereas in consumer test 2 there were high levels of boar taint (Table 1 and Fig. 1). As can be seen, only very few pigs have high skatole concentrations which means that the variation in androstenone is the main background for classifying the pigs as high or medium boar taint level. 4.2. Overall acceptability Taking into account that different methodologies have been used in the already published consumer tests carried out in different countries (reviewed by Font-i-Furnols, 2012), results obtained in a given country cannot be extrapolated to another one. The present paper, however, provides the results from 4 countries using the same methodology. In terms of hedonic assessment, in consumer test 1 (ES & UK) meat with low levels of boar taint (LBT) was rated worse than meat with medium boar taint levels (MBT). Considering consumer test 2 (FR & IT),
meat with high levels of boar taint (HBT) was rated worse than meat from females (FE), and meat with low levels of boar taint (LBT) was scored in between. In the two countries where consumers are used to eating boar meat (ES & UK, consumer test 1) the meat with medium levels of boar taint was scored better than the meat with low levels of boar taint. This can be explained by the resulting lower boar taint levels in consumer test 1 compared with consumer test 2, but also can be explained by the habit of consumers to this type of meat that let the consumers perceive boar odour/flavour as normal as has been hypothesised in a previous study (Matthews et al., 2000) or may be also explained by differences in sensitivity to boar taint compounds. In the two countries where consumers are not used to eating meat from entire male pigs (FR & IT, consumer test 2), the higher the level of boar taint, the worst score given by the consumers. Another interesting result found in this study is that we observed significant differences on the hedonic scores of LBT and FE only in the case of the first consumer test (ES & UK), where the LBT meat was scored with significantly lower values (less delicious, and lower score of odour and taste) than meat from FE. These results contrast with the second consumer test (FR & IT) where no differences between LBT and FE were observed, and also with previous findings such as Bonneau and Chevillon (2012), who found that pork from entire males with very low levels of skatole and androstenone was well accepted as gilt pork by consumers evaluating it during eating. Differences between females and entire male pigs have been described by different authors. For example, Gispert et al. (2010) described differences between meat from females and meat from entire male pigs in terms
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
143
Fig. 4. Least square means for hedonic traits and strength of odour/taste given by consumers of the four clusters observed in consumer test 2 (French and Italian consumers).
of electrical conductivity, meat colour, but no differences in intramuscular fat were observed. In the literature, there are numerous studies that compare meat quality traits of entire male pigs and females but none of them distinguish entire male pigs with low levels and high levels of boar taint. Recently, Mörlein and Tholen (2015) focused their research on comparing entire male pigs with low levels and high levels of boar taint, but the authors of this study did not include females. One explanation of the different acceptability between meat from female and meat from entire male pigs with low levels of boar taint could be based on the intramuscular fat content or maybe tenderness of the meat, but in the present study, none of these attributes were analysed. Therefore, more research is needed to understand the different acceptability of these two types of meat.
4.3. Consumer segmentation based on the hedonic attribute deliciousness To the best of our knowledge, the cluster analyses with the aim to identify consumers' segmentation based on the question ‘How delicious do you find this meat (with different levels of boar taint)’ has not been applied yet. So far, consumer segmentation was identified on the basis of the sensitivity test described by (Weiler et al., 2000) based on smelling crystals of pure androstenone. To identify the existence of market segmentation, a hierarchical cluster analysis was conducted for each consumer test. As a result, a similar cluster solution was identified as best explaining data for each consumer study. Three groups of clusters were identified in consumer test 1 (ES & UK), and four in consumer test 2 (FR & IT). Interestingly, although the boar taint level of the most tainted type of meat (Table 1) was different in the two studies, clusters with similar behaviour were
observed: the first three clusters identified were labelled as Pork lovers, Boar meat lovers and Reject boar tainted meat: ▪ Cluster 1 (Pork lovers) rated the three types of meat with relatively high scores (all of them over 5 in a 9-point scale in which the middle point was not presented to the consumers). This cluster represents 62.1% of consumers in consumer test 1, and 49.5% in consumer test 2. In both cases, compared to the other clusters identified, this cluster comprises the higher percentage of consumers. Despite the fact that significant differences between types of meat have been observed in the different attributes in both studies, it is remarkable that all the scores were relatively high. This segment of consumers was found in the two consumer tests, despite the levels of boar taint in the type of meats assessed. In the case of consumer study 1 (ES & UK), scores given to the LBT meat were significantly lower (worst score, but still over 5 in the 9-point scale) than the meat from FE and MBT, which could explain the results when considering the overall acceptability. Salt addition during cooking, thickness of the loin and cooking methodology could have had an influence in these results, reducing the perception of boar taint to levels that are acceptable by consumers. Lack of differences in consumers' acceptability in meats from boars and gilts has also been reported (Aluwe et al., 2009; Smith, Ellis, Clark, & Innes, 1983). Considering the percentage of consumers represented in this cluster, 62.1% of consumers in the first consumer test and 49.5% of the consumers in the second one were indifferent towards boar taint. These findings could be related to the androstenone sensitivity since it is reported that about 40% of consumers are not able to perceive androstenone (Panella-Riera et al., 2010). Blanch et al. (2012) estimated the percentage of consumers that could reject tainted meat, and also
144
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
reported a range of consumers classified as insensitive or low sensitive to androstenone in ES, UK and FR (no data was provided from IT): 52.5–79.2% in ES, 52.4–79.6% in UK and 43.1–73.6% in FR. In the two consumer tests, the percentage of consumers that were indifferent towards boar taint were close to the values associated to insensitive or low sensitive to androstenone. ▪ Cluster 2 (Boar meat lovers) rated the meat from gilt (FE) with lower scores, and the meat with MBT or HBT with higher scores. Meat with LBT was scored in-between in both cases. This cluster represents 21.7% of consumers in consumer test 1, and 12.4% in consumer test 2. The lower percentage of consumers that are part of this cluster in study 2 may be explained because the highest level of boar taint meat was higher in comparison to the levels in consumer study 1. In fact, it is reported that some of the sensitive consumers like the androstenone smell, and such preference could influence their acceptability of the meat (Blanch et al., 2012; Bonneau & Chevillon, 2012; Font-i-Furnols, Gispert, Diestre, & Oliver, 2003). Blanch et al. (2012) estimated the percentage of consumers that were highly sensitive to androstenone and that revealed that they liked the meat (middle and high sensitive + I like/neutral) in ES, UK and FR: 6.9% in ES, 17.0% in UK and 16.0% in FR. These percentages are close to the percentage of consumers represented in this cluster. ▪ In cluster 3 (Reject boar tainted meat) meat from gilt received the highest (best) score in both consumer tests, while meat with medium levels (MBT in study 1) or high levels (HBT in study 2) received the lowest (worst) score. This cluster represents 16.2% of consumers for the consumer test 1, and 20.0% in consumer test 2. These results are also in line with the results obtained in cluster 2. Since the levels of boar taint were higher in consumer test 2, and due to the fact that the consumers that tasted this meat were not used to boar meat, it was expected to find a higher percentage of consumers rejecting the boar tainted meat. It is interesting to highlight that the type of meat scored below 5 (in a 9-point scale) was HBT for consumer test 2, and MBT for consumer test 1. In both cases, the meat with LBT was scored in-between, but with scores around 5. Therefore, these consumers rejected meat with medium or high boar taint, but not the type of meat with low boar taint. The results of this cluster are in agreement with those reported by Babol, Squires, and Gullett (2002) and Lundström, Malmfors, Fjelkner-Modig, and Szatek (1982). Additionally, in consumer test 2 (FR & IT) there was a fourth cluster identified and labelled as consumers that Reject low boar tainted meat, representing 18.2% of the sample. In this one, the meat with low boar taint received the lowest (worst) score in all the attributes (between 4.3 and 5.1). The identification of this cluster could be expected if we take into account that it has been found in the two countries where consumers are not used to the meat from entire male pigs although it is strange that meat with HBT had high acceptability scores and similar to the acceptability observed for the FE meat. Considering the percentage of consumers being part of each cluster, a percentage of 16.2% (from cluster 3) in consumer study 1 (ES & UK) and 38.2% (from cluster 3 plus 4) in consumer study 2 (FR & IT) could react negatively when eating meat from entire male pigs. Again, to understand these differences two facts should be taken into account: (i) meat with boar taint used in FR and IT had higher boar taint levels than the meat used for ES and UK, and (ii) consumers from FR and IT were not used to eat meat from entire male pigs. 5. Conclusions and implications The present study shows that, when eating fresh pork (loin 0.5 cm thick with 5 mm back fat, cooked in a hot plate and served uncovered), different groups of consumers could be identified on the basis of ‘How delicious do you find this meat’. Among them, a group of consumers that like pork meat (49.5–62.1% of the participants), a group of
consumers that prefer meat with boar taint (12.4–21.7%) and a group of consumers that reject boar tainted meat (16.2–20.0%) have been identified. Results suggest that there is a group of consumers who may reject meat with boar taint, but there is also a niche for meat from medium and high levels of boar taint in the cities where the consumer tests were performed. Further research is needed to understand how to address this niche. The present research also identifies a need to develop tools to select and classify carcasses on the basis of boar taint level so that it can be properly addressed to the meat market. This situation can influence pork consumption so that it might guarantee that consumers do not purchase pork with inadequate sensory quality, according to their preferences. Acknowledgements This study was financially supported by the Health and Consumers Directorate-General, Animal Health and Welfare (DG-SANCO) of the European Commission (Contract No. SANCO/2008/D5/018). Complementary financial supports were provided at the national levels by INAPORC in France and IRTA in Spain. References Aluwe, M., Millet, S., Nijs, G., Tuyttens, F. A. M., Verheyden, K., De Brabander, H. F., ... Van Oeckel, M. J. (2009). Absence of an effect of dietary fibre or clinoptilolite on boar taint in entire male pigs fed practical diets. Meat Science, 82(3), 346–352. http://dx.doi.org/ 10.1016/j.meatsci.2009.02.001. Ampuero, S., Verkuylen, B., Dahlmans, H., Hortós, M., García-Regueiro, J. A., Dahl, E., ... Harlizius, B. (2011). Inter-laboratory comparison of methods to measure androstenone in pork fat. Animal, 5(10), 1–9. http://dx.doi.org/10.1017/S1751731111000553. Babol, J., Squires, E. J., & Gullett, E. A. (2002). Factors affecting the level of boar taint in entire male pigs as assessed by consumer sensory panel. Meat Science, 61(1), 33–40. http://dx.doi.org/10.1016/S0309-1740(01)00159-0. Blanch, M., Panella-Riera, N., Chevillon, P., Font-i-Furnols, M., Gil, M., Gil, J. M., ... Oliver, M. A. (2012). Impact of consumer's sensitivity to androstenone on acceptability of meat from entire male pigs in three European countries: France, Spain and United Kingdom. Meat Science, 90(3), 572–578. Bonneau, M., & Chevillon, P. (2012). Acceptability of entire male pork with various levels of androstenone and skatole by consumers according to their sensitivity to androstenone. Meat Science, 90(2), 330–337. http://dx.doi.org/10.1016/j.meatsci. 2011.07.019. DG-SANCO (2010). European Declaration on alternatives to surgical castration of pigs. Retrieved 5 August 2013, from http://www.alternativepig.eu/partnership/ declaration.html EFSA (2004). Report AHAW/04–087: Welfare aspects of the castration of piglets. Font-i-Furnols, M. (2012). Consumer studies on sensory acceptability of boar taint: A review. Meat Science, 92(4), 319–329. http://dx.doi.org/10.1016/j.meatsci.2012.05.009. Font-i-Furnols, M., Gispert, M., Diestre, A., & Oliver, M. A. (2003). Acceptability of boar meat by consumers depending on their age, gender, culinary habits, and sensitivity and appreciation of androstenone odour. Meat Science, 64(4), 433–440. http://dx. doi.org/10.1016/S0309-1740(02)00212-7. Fredriksen, B., Font-i-Furnols, M., Lundström, K., Migdal, W., Prunier, A., Tuyttens, F. A. M., & Bonneau, M. (2009). Practice on castration of piglets in Europe. Animal, 3(11), 1480–1487. http://dx.doi.org/10.1017/S1751731109004674. Gispert, M., Oliver, M.À., Velarde, A., Suarez, P., Pérez, J., & i Furnols, M. F. (2010). Carcass and meat quality characteristics of immunocastrated male, surgically castrated male, entire male and female pigs. Meat Science, 85(4), 664–670. Guerrero, L. (1999). Estudios de consumidores: análisis de los errores más habituales. In T. C. A. Almeida, G. Hough, M. H. Damasio, & M. A. A. d. Silva (Eds.), Avances en Análisis Sensorial (pp. 121–129). Sao Pablo: Ed. VArela. Lunde, K., Egelandsdal, B., Choinski, J., Mielnik, M., Flatten, A., & Kubberod, E. (2008). Marinating as a technology to shift sensory thresholds in ready-to-eat entire male pork meat. Meat Science, 80(4), 1264–1272. http://dx.doi.org/10.1016/j.meatsci. 2008.05.035. Lundström, K., Malmfors, B., Fjelkner-Modig, S., & Szatek, A. (1982). Consumer testing of boar meat in Sweden. Swedish Journal of Agricultural Research, 13, 39–46. Macfie, H. J., Bratchell, N., Greenhoff, K., & Vallis, L. V. (1989). Designs to balance the effect of order of presentation and first-order carry-over effects in hall tests. Journal of Sensory Studies, 4(2), 129–148. http://dx.doi.org/10.1111/j.1745-459X.1989. tb00463.x. Matthews, K. R., Homer, D. B., Punter, P., Béague, M. P., Gispert, M., Kempster, A. J., ... Bonneau, M. (2000). An international study on the importance of androstenone and skatole for boar taint: III. Consumer survey in seven European countries. Meat Science, 54(3), 271–283. Mörlein, D., & Tholen, E. (2015). Fatty acid composition of subcutaneous adipose tissue from entire male pigs with extremely divergent levels of boar taint compounds — An exploratory study. Meat Science, 99, 1–7. Panella-Riera, N., Blanch, M., Gonzàlez, J., Gil, M., Gispert, M., Font I Furnols, M., & Oliver, M. A. (2010, 18–19 March). Consumers' sensitivity to androstenone in Spain and UK.
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145 Paper presented at the EAAP Working Group Meeting. Production and utilization of meat from entire male pigs, Bristol, UK. Patterson, R. L. S. (1968). 5α-Androst-16-ene-3-one:—Compound responsible for taint in boar fat. Journal of the Science of Food and Agriculture, 19(1), 31–38. http://dx.doi.org/ 10.1002/jsfa.2740190107. Pauly, C., Spring, P., O'Doherty, J. V., Ampuero, S., & Bee, G. (2008). Performances, meat quality and boar taint of castrates and entire male pigs fed a standard and a raw potato starch-enriched diet. Animal, 2(11), 1707–1715. http://dx.doi.org/10.1017/ s1751731108002826. PIGCAS (2009). Report on recommendations for research and policy support. Deliverable 4.1 of the EU project PIGCAS: Attitudes, practices and state of the art regarding piglet castration in Europe. Smith, W. C., Ellis, M., Clark, J. B. K., & Innes, N. (1983). A comparison of boars, gilts and castrates for bacon manufacture. 2. Curing characteristics, bacon yield and quality. Animal Production, 37, 11–15.
145
Vold, E. (1970). Fleischproduktionseigenschaften bei ebern und kastraten IV: organoleptische und gaschromatographische untersuchungen wasserdampfflüchtiger stoffe des rücken-speckes von ebern. Meldinger fra Norges Landbrukshoegskole, 49, 1–15. Walstra, P., & Maarse, G. (1970). Onderzoek gestachlengen van mannelijke mestvarkens. Researchgroep voor Vlees en Vleeswaren TNO, IVO-rapport C-147, Rapport, 2, 30. Weiler, U., Font-i-Furnols, M., Fischer, K., Kemmer, H., Oliver, M. A., Gispert, M., ... Claus, R. (2000). Influence of differences in sensitivity of Spanish and German consumers to perceive androstenone on the acceptance of boar meat differing in skatole and androstenone concentrations. Meat Science, 54(3), 297–304.
Contents lists available at ScienceDirect
Meat Science journal homepage: www.elsevier.com/locate/meatsci
Consumers' segmentation based on the acceptability of meat from entire male pigs with different boar taint levels in four European countries: France, Italy, Spain and United Kingdom N. Panella-Riera a,⁎,1, M. Blanch a,1, Z. Kallas b, P. Chevillon c, A. Garavaldi d, M. Gil a, J.M. Gil d, M. Font-i-Furnols a, M.A. Oliver a a
IRTA-Monells, Finca Camps i Armet, 17121 Monells, Spain CREDA, Parc Mediterrani de la Tecnologia, 08860 Castelldefels, Spain c IFIP, La Motte au Vicomte, 35651 Le Rheu, France d CRPA, Viale Timavo 43/2, 42121 Reggio Emilia, Italy b
a r t i c l e
i n f o
Article history: Received 21 April 2015 Received in revised form 17 December 2015 Accepted 23 December 2015 Available online 28 December 2015 Keywords: Acceptability Androstenone Boar taint Consumers Clusters Skatole
a b s t r a c t Two consumer studies were conducted to know the acceptability of pork with different boar taint levels: test 1 performed in Spain (n = 126) and United Kingdom (n = 146), and test 2 performed in France (n = 139) and Italy (n = 140). Each test had 3 types of pork: ‘Female meat’, ‘Low boar tainted meat’, and a third type was ‘Medium boar tainted meat’ or ‘High boar tainted meat’. Three main clusters were identified on the basis of ‘How delicious do you find this meat?’: 1—Pork lovers, 2—Boar meat lovers, 3—Reject boar tainted meat. Additionally, in test 2, a fourth cluster was identified: ‘Reject low tainted meat’. A group of 16.2–38.2% of consumers rejected meat from boars, and another group of 12.4–21.7% rated the meat with medium or high levels of boar taint better than the meat from females, identifying a niche for meat from medium and high levels of boar taint, and suggesting the need to select carcasses on the basis of boar taint. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Nowadays there is a growing concern about the negative effect of surgical castration of pigs without anaesthesia on animal welfare. In 2008, across Europe the majority of male piglets intended for pork production were castrated to avoid potential consumer dissatisfaction because of boar taint. At that time, castration was performed in most of the EU countries on 80–100% of the male pigs in conventional production, and surgical castration without anaesthesia was the most common technique (Fredriksen et al., 2009). The exceptions were United Kingdom and Ireland, where castration was hardly performed, and some southern countries such as Cyprus, Portugal and Spain, where a limited percentage of the male pigs were castrated in comparison to the other European countries. Since then, some countries have taken an action to avoid piglet castration without anaesthesia not only for welfare reasons, but also because the production of entire male pigs decreases production costs and results in higher meat content of the carcass (EFSA, 2004; PIGCAS, 2009). Norway and Switzerland have ⁎ Corresponding author. E-mail address: [email protected] (N. Panella-Riera). 1 The first and the second authors had an equal input to the paper.
http://dx.doi.org/10.1016/j.meatsci.2015.12.017 0309-1740/© 2015 Elsevier Ltd. All rights reserved.
already banned piglet castration without anaesthesia by law, and other countries such as The Netherlands and Germany have signed letters of intentions (Declaration of Noordwijk 2007 and Düsseldorf Declaration 2008, respectively) which aimed to avoid the need for piglet castration in the long term. In 2010, at the EU level, representatives of European farmers, meat industry, retailers, scientists, veterinarians and animal welfare NGOs, committed themselves to voluntarily end the practise of surgically castrating pigs in Europe by January 2018 by means of the European Declaration on alternatives to surgical castration of pigs (DG-SANCO, 2010). Consequently, entire male production is one of the alternatives and therefore, an increase in the production of entire males would result in a positive impact on production efficiency and carcass composition. But as a consequence, more tainted carcasses could enter the meat chain. A number of consumer studies have been carried out in order to ascertain the acceptability of pork from entire male pigs in different countries and using different methodologies as was reviewed in EFSA (2004) and more recently, by Font-i-Furnols (2012). The two main compounds responsible for boar taint are androstenone (Patterson, 1968) and skatole (Vold, 1970; Walstra & Maarse, 1970). The respective contributions of these two compounds vary depending on the characteristics of the meat evaluated, procedures for preparing the meat, the
138
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
ability of consumers to perceive androstenone among others, as has been reviewed by Font-i-Furnols (2012), and also it may vary depending on the habit of eating meat from castrated or entire male pigs. Therefore, it would be interesting to study the impact that meat from boars could have in the European market (consumer satisfaction), which is traditionally used to eating meat from castrated pigs, and identify if there is any difference comparing to other countries where the utilization of entire male pigs is more common. The present work had two main objectives: 1) to evaluate and update the results on sensory acceptability of meat from entire male pigs (as an alternative to the production of castrates), involving four European countries that produce a different proportion of castrated pigs in two consumer studies: Spain and United Kingdom, France, and Italy, and 2) to identify potential niche markets for meat with different boar taint levels. 2. Materials and methods 2.1. Consumer studies Two consumer studies were conducted in which each consumer assessed pork with various levels of androstenone and skatole. Participants in each study were: ▪ Consumer test 1. Spain (ES, n = 133, performed in Barcelona) and United Kingdom (UK, n = 146, performed in Reading); April, 2009. ▪ Consumer test 2. France (FR, n = 139, performed in Cain (n = 74) & Paris (n = 70)) and Italy (IT, n = 140, performed in Matelica (n = 70) & Reggio Emilia (n = 70)); April and May 2009. Considering the two consumer tests, a total of 558 consumers from the different European countries participated in this project. They were stratified by age (according to each country profile) and sex (approximately 50:50 ratio between men and women). Consumers were required to eat pork on a regular basis.
From each animal, the muscle Longissimus lumborum (with the subcutaneous fat) was taken from the 1st lumbar vertebra to the last rib, divided in two sections (min. 15 cm) and distributed in one of the two countries. 2.3. Androstenone and skatole measurements in fat Back fat samples from entire male pigs and females were collected on the day of slaughter at the level of the neck, vacuum packed and deep frozen until analysed for androstenone (AND) and skatole (SKA). The determination of SKA levels was performed using HPLC-FLD and the determination of AND levels using GC-MS (for ES and UK samples; Ampuero et al., 2011) or by HPLC-FLD (for FR and IT samples; Pauly, Spring, O'Doherty, Ampuero, and Bee, 2008). Results were expressed as μg/g, on pure fat basis. Meat samples were classified as FE (meat from gilts), Low boar taint (LBT), Medium boar taint (MBT) or High boar taint (HBT) depending on the sex and the levels of boar taint compounds (Blanch et al., 2012; Table 1, Fig. 1). In each consumer test 3 samples were evaluated in each session: samples from gilts (FE) as a reference, and two more samples with different levels of boar taint: in consumer test 1 (ES & UK) meat from LBT and MTB was used; and in consumer test 2 (FR & IT), meat from LBT and HBT. 2.4. Preparation of samples In each location, loins were defrosted at 2–4 °C during 24 h, and cut into 0.5 cm thick slices with 5 mm of subcutaneous fat (when it was possible). Each slice was cut in two pieces, and cooked using a cooking plate at 180 °C (which was greased with maize oil). A different cooking plate was used for each type of meat presented within a session. The meat was turned upside down regularly up to a core temperature of 80 °C measured with a flexible temperature probe (TESTO, Germany) and the meat was salted after cooking (1 g salt/200 g meat), reproducing home preparation. Samples were then immediately served to the consumers for evaluation.
2.2. Animals and samples 2.5. Sensory evaluation of samples The meat and fat samples used for the consumer studies were obtained from populations of pigs slaughtered according to commercial practise. They were collected from boars and gilts in commercial abattoirs from Spain and France. Carcass weights and genotypes were representative of current practise in these countries (carcass weight of 79.4 kg in Spain, and 84.0 kg in France; (2009)). The meat collected in Spain was used for the consumer studies carried out in ES and UK; and the meat collected in France was used for the consumer studies carried out in FR and IT. A minimum 20 loins from gilts and 20 loins from boars with Low boar taint (LBT) were collected in Spanish and French abattoirs. Furthermore, a minimum of 20 loins from boars with Medium boar taint (MBT) were collected in Spanish abattoirs and a minimum of 20 loins with High boar taint (HBT) were selected in French abattoirs (Table 1).
Sessions of 10–12 consumers were organized for meat evaluation. Each consumer assessed 3 pieces of meat. A 10 min interval was used between each of the three samples presented within a session. The order of presentation of samples was rotated using a partial Latin square design to avoid any first sample and carry-over effect and the identity of the samples was not given to consumers (Macfie, Bratchell, Greenhoff, & Vallis, 1989). Consumers were asked for hedonic questions (questions 1 to 4) and intensity evaluations (questions 5–11) as described in Table 2, using a modified nine-point intensity scale (from 1 to 9). The intermediate level (5) was not included to stimulate consumers to commit themselves and not to allow a neutral assessment which is not informative (Guerrero, 1999). Table 2 also shows the scale used to assess each attribute.
Table 1 Mean and standard error of the androstenone and skatole levels on the pure fat basis of the meat used for the consumer test. N of loins
Androstenone (ppm pure fat) [range]
Skatole (ppm pure fat) [range]
Consumer test 1: ES & UK Gilt (FE) Low boar taint (LBT) Medium boar taint (MBT)
18 18 18
b0.04 0.20 ± 0.07 [0.04–0.29] 1.07 ± 0.40 [0.58–2.28]
0.04 ± 0.02 [0.02–0.07] 0.06 ± 0.02 [0.02–0.08] 0.18 ± 0.07 [0.11–0.39]
Consumer test 2: FR & IT Gilt (FE) Low boar taint (LBT) High boar taint (HBT)
18 18 18
b0.2 b0.2 2.39 ± 1.07 [0.59–5.18]
b0.03 b0.03 0.11 ± 0.07 [0.02–0.28]
ES: Spain; UK: United Kingdom; FR: France; IT: Italy.
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
139
eat the pork with the fat), FREQ and GENMOD procedures were applied. The MIXED procedure was used to analyse the acceptability of odour and flavour by consumers: the model included the type of animal, country and their interaction as fixed effects, session as blocking variable and consumer as random effect. Interaction was kept in the model only if it was significant (P b 0.05). Differences were considered at P b 0.05 and tendencies at P b 0.10. In order to establish different clusters of consumers a hierarchical cluster analysis was performed with the CLUSTER procedure and the Ward method. The cluster analysis was conducted on the basis of the attribute ‘delicious’. For each cluster, the MIXED procedure was then used to analyse the hedonic attributes (delicious, odour, taste, and in-mouth feeling) and the intensity attributes (strength of odour and strength of taste, abnormal odour and abnormal taste) scored by the consumers. The model included the type of animal and country as fixed effects, session as blocking effect and consumer as random effect. The interaction type of animal and country was removed from the model because it was not significant (P N 0.05). Differences were also declared at P b 0.05. 3. Results 3.1. Description of consumers Table 3 shows the description of consumers according to the gender, age and educational level. 3.2. Habits of European consumers regarding pork consumption Fig. 1. Androstenone and skatole levels of the meat used in the two consumer studies.
2.6. Description of EU pork consumers At the end of the test, consumers also answered socio-demographic questions and frequency of consumption of different pork products, the most common purchasing place for fresh pork meat, if they were responsible for buying fresh pork at home, if they were responsible for cooking at home, and if they usually eat the pork with the fat.
2.7. Statistical analysis Data analyses were conducted using SAS Statistical Package (SAS Inst., Inc., Cary, NC, USA, version 9.2). To analyse the parameters regarding classification of consumers (age, gender, educational level, frequency of consumption of different pork products, the most common purchasing place for fresh pork meat, if they were responsible for buying fresh pork at home, if they were responsible for cooking at home, and if they usually
In general, over 90% of consumers ate fresh pork N 2 times/week (ES 95.5%; UK 97.3%; IT 92.9%) except for FR (34.8%) (Table 4). The most consumed product was dry cured ham in ES and IT, cooked ham in FR, and sliced bacon in UK (Fig. 2). In general, the percentage of respondents responsible for buying fresh pork in their household was 86.4% (Table 4). In all countries, women were more responsible for buying fresh pork than men, and they were mostly between 41 and 60 years old. Eighty-eight percent of respondents were partially responsible for cooking at home. Women were more responsible for cooking at home than men. France was an exception, where percentages were very similar 49.6% women and 50.4% men cooked at home. Considering all respondents, 43.5% were used to eat the pork with the fat in all the countries (36.4% of women and 53.1% of men). Considering the place of purchasing fresh pork meat, the supermarket was the most common one, followed by the butcher and finally the traditional market. Table 3 Description of consumers according to gender, age and educational level.
Table 2 Questionnaire and scoring scales for the evaluation of meat samples during eating. Questions Hedonic evaluations 1. How delicious do you find this product? 2. What do you think of the odour? 3. What do you think of the taste? 4. What do you think of the in-mouth feeling? Intensity evaluations 5. Strength of odour 6. Abnormal odour 7. Strength of taste 8. Salty taste 9. Meaty taste 10. Fatty taste 11. Abnormal taste a
All scoring scales were from 1 to 9, without 5.
Consumer test 1 n = 279
Scoring scalesa 1. Not good at all 9. Really delicious 1. Not pleasant at all 9. Very pleasant
1. Barely perceptible 9. Very strong
N Gender (%) Male Female Age (%) 18–25 26–40 41–60 N60 Educational level (%) Primary studies not completed Primary studies Secondary studies University studies
Consumer test 2 n = 279
ES
UK
FR
IT
133
146
139
140
Total
558
54.9 45.1
50.0 50.0
49.6 50.4
74.3 25.7
42.8 57.2
10.5 30.8 43.6 15.0
19.9 19.2 39.0 21.9
15.1 28.8 41.0 15.1
10.0 25.7 60.0 4.3
14.0 26.0 45.9 14.2
3.0 12.1 53.8 31.1
0.0 4.8 61.6 33.6
11.7 3.6 46.0 38.7
2.9 25.0 29.4 42.6
4.4 11.3 47.9 36.5
ES: Spain; UK: United Kingdom; FR: France; IT: Italy.
140
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
Table 4 Frequency (%) of consumption of fresh pork meat and behaviour of the consumers of each country participant in the study. Consumer test 1
Consumer test 2
ES
FR
UK
Table 5 Significance levels within a consumer test: consumer test 1 carried out in Spain and United Kingdom, and consumer test 2 carried out in France and Italy.
Total
IT
n 133 146 139 140 558 How many times a week do you eat fresh pork meat? 1 time per week 0.0 0.0 29.7 1.4 7.7 2 times per week 4.5 2.7 35.5 5.7 12.1 More than 2 times per week 95.5 97.3 34.8 92.9 80.2 Are you partially responsible of doing the shopping of fresh pork eaten in your household? Yes 78.9 81.5 93.5 91.4 86.4 Are you partially responsible for cooking in your household? Yes 80.5 82.9 97.8 91.4 88.2 When you eat the pork meat, do you eat it… without the fat 58.6 43.8 61.2 62.9 56.5 with the fat 41.4 56.2 38.8 37.1 43.5 Where do you buy fresh pork meat? (multiple choice question) At the butcher 58.6 21.9 15.8 14.3 27.2 At the supermarket 72.2 95.9 82 94.3 86.4 At the traditional market 42.1 4.8 0.7 1.4 11.8 ES: Spain; UK: United Kingdom; FR: France; IT: Italy.
3.3. Sensory evaluation The hedonic evaluation of in-mouth feeling and of the intensity of salty, meaty and fatty tastes did not result in any consistent difference between meat sample groups and therefore it is not presented.
3.3.1. Differences between meat sample groups, within experiments Table 5 shows significance levels within a consumer test: Consumer test 1 carried out in Spain and United Kingdom, and Consumer test 2 carried out in France and Italy. Results showed no significant interaction between the boar taint level and the country in each consumer test. In consumer test 1 (ES & UK), significant differences (P b 0.05) were observed for the hedonic attributes assessed by the consumers (delicious, odour and taste). In all the cases, the score given to the meat from FE and MBT was significantly higher (more favourable)
Hedonic evaluations Delicious Odour Taste Intensity evaluations Strength of odour Abnormal odour Strength of taste Abnormal taste
Consumer test 1 (ES & UK) n = 279
Consumer test 2 (FR & IT) n = 279
BT
C
BTxC
BT
C
BTxC
** *** ***
*** *** **
ns ns ns
† *** *
ns ns ns
ns † ns
*** ns * ns
*** ns † **
ns ns ns ns
ns ns ** *
ns ns ns ns
ns ns ns ns
BT: type of meat; C: country. ns: P N 0.10; +: P b 0.10; *: P b 0.05; **: P b 0.01; ***: P b 0.001.
than the one given to the meat from LBT (Table 6). With regard to the intensity evaluations, significant differences were observed between type of meat when considering strength of odour and strength of taste (P b 0.05) whereas no differences were observed when taking into account the attributes abnormal odour and abnormal taste. Results showed that the strongest odour was observed in MBT and FE meat, and the highest strength of taste was observed in MBT in comparison to LBT. In consumer test 2 (FR & IT), significant differences (P b 0.05) were observed for the hedonic odour and taste (Table 7). For all of them, HBT meat was rated worst (lower scores) than FE meat. Considering the intensity scores, significant differences (P b 0.05) were observed for the attributes strength of taste and abnormal taste. In both cases, HBT received the highest score (strongest taste and abnormal taste). 3.3.2. Consumer segmentation in consumer test 1: Spain and United Kingdom Due to the fact that the interaction between type of meat and country was not significant, results are presented considering consumers from the two countries together. A total of three clusters were identified
Fig. 2. Frequency (%) of consumption of different pork products by the consumers from each country who participated in the study.
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145 Table 6 Least square means (LSM) and standard error (S.E.) in consumer test 1 (ES & UK) of the scores given by consumers to meat from gilts (FE), from entire males with low levels of skatole and androstenone (LBT), and entire male pigs with medium levels of skatole and androstenone (MBT). FE
Hedonic evaluations Delicious Odour Taste Intensity evaluations Strength of odour Abnormal odour Strength of taste Abnormal taste 1
LBT
Sig1
MBT
LSM
S.E.
LSM
S.E.
LSM
S.E.
6.4a 6.5a 6.5a
0.11 0.10 0.11
5.9b 6.0b 6.0b
0.12 0.10 0.11
6.4a 6.5a 6.5a
0.12 0.10 0.11
0.0015 0.0003 0.0005
4.8a 2.6 5.7ab 2.6
0.12 0.11 0.10 0.11
4.3b 2.7 5.5b 2.7
0.12 0.11 0.10 0.11
4.8a 2.6 5.9a 2.7
0.12 0.11 0.11 0.11
0.0003 0.6171 0.0197 0.4422
Different letters within a line indicate significant differences (P b 0.05).
on the basis of ‘how delicious do you find this meat’. Fig. 3 shows least square means of the scores that consumers gave to each attribute (hedonic attributes: delicious, odour and taste; intensity attributes: strength of odour and strength of taste) and each type of meat (FE, LBT, MBT), for each cluster identified: ▪ The first cluster (n = 169, 62.1% of consumers; 57.4% from Barcelona and 42.6% from Reading) comprises respondents that gave high scores to all types of samples. These consumers were labelled as ‘Pork lovers’. ▪ The second cluster (n = 59, 21.7% of consumers; 30.5% from Barcelona and 69.5% from Reading) comprises respondents that like the boar taint and therefore, the higher the level of boar taint, the higher the score to the attribute. These consumers were considered as ‘Boar meat lovers’. ▪ The third cluster (n = 44, 16.2% of consumers; 25.0% from Barcelona and 75.0% from Reading) comprises respondents that did not like the boar taint and therefore, the higher the level of boar taint, the lower the score to the attribute ‘Delicious’. These consumers were labelled as ‘Reject boar tainted meat’. The three clusters did not display any relevant difference on the basis of demographic variables (age, gender and educational level) and consumers' behaviour such as frequency of consumption of fresh meat (Table 4), pork products (Fig. 2), habits related to doing the shopping, cooking at home, eating meat with or without fat and place of purchasing fresh pork (Table 4). 3.3.3. Consumer segmentation in consumer test 2: France and Italy As observed in consumer test 1, due to the fact that the interaction between type of meat and country was not significant, results are presented considering consumers from the two countries together. In this consumer test, a total of four clusters were identified on the basis of ‘how delicious do you find this meat’ (Fig. 4). Table 7 Least square means (LSM) and standard error (S.E.) in consumer test 2 (FR & IT) of the scores given by consumers to meat from gilts (FE), from entire males with low levels of skatole and androstenone (LBT), and entire male pigs with high boar taint (HBT). FE
Hedonic evaluations Delicious Odour Taste Intensity evaluations Strength of odour Abnormal odour Strength of taste Abnormal taste 1
LBT
Sig1
HBT
LSM
S.E.
LSM
S.E.
LSM
S.E.
5.7 5.9a 5.8a
0.11 0.11 0.11
5.6 5.9a 5.7ab
0.11 0.11 0.11
5.4 5.4b 5.4b
0.11 0.11 0.11
0.0518 0.0002 0.0153
4.4 2.3 4.8ab 2.2b
0.12 0.11 0.11 0.12
4.5 2.4 4.7b 2.5ab
0.12 0.11 0.11 0.12
4.5 2.5 5.1a 2.6a
0.12 0.11 0.11 0.12
0.3047 0.1537 0.0102 0.0309
Different letters within a line indicate significant differences (P b 0.05).
141
▪ The first cluster (n = 136, 49.5% of consumers; 57.4% from France and 42.6% from Italy) comprises respondents that gave high scores to all types of samples. These consumers were labelled as ‘Pork lovers’. ▪ The second cluster (n = 34, 12.4% of consumers; 23.5% from France and 76.5% from Italy) comprises respondents that liked the boar taint and therefore, the higher the level of boar taint, the higher the score to the attribute. These consumers were considered as ‘Boar meat lovers’. ▪ The third cluster (n = 55, 20.0% of consumers; 49.1% from France and 50.9% from Italy) comprises respondents that did not like the HBT meat and therefore, the higher the level of boar taint, the lower the score to the attribute ‘Delicious’. These consumers were labelled as ‘Reject boar tainted meat’. ▪ The fourth cluster (n = 50, 18.2% of consumers; 52.0% from France and 48.0% from Italy) comprises respondents that gave low scores to the meat with LBT. These consumers were labelled as ‘Reject low boar tainted meat’. Like in consumer test 1, the four clusters did not display any relevant difference on the basis of demographic variables (age, gender and educational level) and consumers' behaviour (Table 4 and Fig. 3). 4. Discussion In the present paper, consumer studies were carried out with the aim to ascertain consumer acceptability of boar meat with different levels of boar taint. Due to practical reasons, meat was collected in Spain and France, chemically analysed in terms of androstenone and skatole levels, and then distributed to two other countries: meat from Spain was used for the consumer test performed in Spain and United Kingdom, and meat from France was used for the consumer test carried out in France and Italy. It is interesting to highlight that, consumer test 1 was planned in two countries where, as pointed out by Fredriksen et al. (2009), consumers are used to the commercialization of meat from entire male pigs, whereas consumer test 2 was carried out in two countries where consumers are not used to this type of meat: in 2008, the percentage of production of castrated male pigs was estimated as 33.2% in ES, 2.1% in UK, 97.5% in FR and 100% in IT. It is also interesting to underline that in the two studies consumers could taste meat from entire male pigs with such low levels of boar taint comparable to the levels found in meat from gilts (Bonneau & Chevillon, 2012; type of meat: LBT – low boar taint – and FE ‐ meat from gilts). Additionally, each study had a third type of meat with medium levels (MBT in study 1) or high levels (HBT in study 2) of boar taint. 4.1. Methodology used As reviewed by Font-i-Furnols (2012), the methodologies used in the already published papers are very diverse and therefore, it is difficult to compare the results among them. Issues such as type of meat (loin slices, chops, cutlets, minced meat, fat, bacon, dry cured ham...), location of the test (hall or home test), cooking procedure (including cooking device, cooking time and temperature), type of meat samples (meat from gilts, castrated pigs and/or from entire male pigs), level of boar taint in the assessed meat, and the type of attributes and scales used during the consumer test are only some examples. As stated in the methodology section, in this paper we used meat from the Longissimus lumborum muscle from the 1st lumbar vertebra to the last rib. We served loin samples due to the fact that loin cuts are widely sold and consumed. On the day of the consumer test, the meat was cut into 0.5 cm thick slices with a maximum of 5 mm of subcutaneous fat. The acceptance of boar meat is assumed to be dependent on the sample preparation and presentation. In this study, meat samples were cooked on a hot plate and were served uncovered to the participants to simulate a meal at home. This kind of preparation is not very
142
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
Fig. 3. Least square means for hedonic traits and strength of odour/taste given by consumers of each of the three clusters observed in consumer test 1 (Spanish and English consumers).
common in previously published papers although it is very similar to the type of preparation that consumers usually do at home. Since AND and SKA are predominantly released during heating (Lunde et al., 2008), the intensity of these volatile compounds is expected to be higher during the preparation than at consumption. In our study, consumers were not involved in the cooking process. Therefore, the fact that the meat came from a very lean piece of meat, the thickness of the sample and the fact that it was served uncovered, could minimize the perception of boar taint. Regarding the androstenone and skatole levels of the samples used in this study, the highest level of boar taint was different in each consumer test: consumer test 1 had a group with medium levels whereas in consumer test 2 there were high levels of boar taint (Table 1 and Fig. 1). As can be seen, only very few pigs have high skatole concentrations which means that the variation in androstenone is the main background for classifying the pigs as high or medium boar taint level. 4.2. Overall acceptability Taking into account that different methodologies have been used in the already published consumer tests carried out in different countries (reviewed by Font-i-Furnols, 2012), results obtained in a given country cannot be extrapolated to another one. The present paper, however, provides the results from 4 countries using the same methodology. In terms of hedonic assessment, in consumer test 1 (ES & UK) meat with low levels of boar taint (LBT) was rated worse than meat with medium boar taint levels (MBT). Considering consumer test 2 (FR & IT),
meat with high levels of boar taint (HBT) was rated worse than meat from females (FE), and meat with low levels of boar taint (LBT) was scored in between. In the two countries where consumers are used to eating boar meat (ES & UK, consumer test 1) the meat with medium levels of boar taint was scored better than the meat with low levels of boar taint. This can be explained by the resulting lower boar taint levels in consumer test 1 compared with consumer test 2, but also can be explained by the habit of consumers to this type of meat that let the consumers perceive boar odour/flavour as normal as has been hypothesised in a previous study (Matthews et al., 2000) or may be also explained by differences in sensitivity to boar taint compounds. In the two countries where consumers are not used to eating meat from entire male pigs (FR & IT, consumer test 2), the higher the level of boar taint, the worst score given by the consumers. Another interesting result found in this study is that we observed significant differences on the hedonic scores of LBT and FE only in the case of the first consumer test (ES & UK), where the LBT meat was scored with significantly lower values (less delicious, and lower score of odour and taste) than meat from FE. These results contrast with the second consumer test (FR & IT) where no differences between LBT and FE were observed, and also with previous findings such as Bonneau and Chevillon (2012), who found that pork from entire males with very low levels of skatole and androstenone was well accepted as gilt pork by consumers evaluating it during eating. Differences between females and entire male pigs have been described by different authors. For example, Gispert et al. (2010) described differences between meat from females and meat from entire male pigs in terms
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
143
Fig. 4. Least square means for hedonic traits and strength of odour/taste given by consumers of the four clusters observed in consumer test 2 (French and Italian consumers).
of electrical conductivity, meat colour, but no differences in intramuscular fat were observed. In the literature, there are numerous studies that compare meat quality traits of entire male pigs and females but none of them distinguish entire male pigs with low levels and high levels of boar taint. Recently, Mörlein and Tholen (2015) focused their research on comparing entire male pigs with low levels and high levels of boar taint, but the authors of this study did not include females. One explanation of the different acceptability between meat from female and meat from entire male pigs with low levels of boar taint could be based on the intramuscular fat content or maybe tenderness of the meat, but in the present study, none of these attributes were analysed. Therefore, more research is needed to understand the different acceptability of these two types of meat.
4.3. Consumer segmentation based on the hedonic attribute deliciousness To the best of our knowledge, the cluster analyses with the aim to identify consumers' segmentation based on the question ‘How delicious do you find this meat (with different levels of boar taint)’ has not been applied yet. So far, consumer segmentation was identified on the basis of the sensitivity test described by (Weiler et al., 2000) based on smelling crystals of pure androstenone. To identify the existence of market segmentation, a hierarchical cluster analysis was conducted for each consumer test. As a result, a similar cluster solution was identified as best explaining data for each consumer study. Three groups of clusters were identified in consumer test 1 (ES & UK), and four in consumer test 2 (FR & IT). Interestingly, although the boar taint level of the most tainted type of meat (Table 1) was different in the two studies, clusters with similar behaviour were
observed: the first three clusters identified were labelled as Pork lovers, Boar meat lovers and Reject boar tainted meat: ▪ Cluster 1 (Pork lovers) rated the three types of meat with relatively high scores (all of them over 5 in a 9-point scale in which the middle point was not presented to the consumers). This cluster represents 62.1% of consumers in consumer test 1, and 49.5% in consumer test 2. In both cases, compared to the other clusters identified, this cluster comprises the higher percentage of consumers. Despite the fact that significant differences between types of meat have been observed in the different attributes in both studies, it is remarkable that all the scores were relatively high. This segment of consumers was found in the two consumer tests, despite the levels of boar taint in the type of meats assessed. In the case of consumer study 1 (ES & UK), scores given to the LBT meat were significantly lower (worst score, but still over 5 in the 9-point scale) than the meat from FE and MBT, which could explain the results when considering the overall acceptability. Salt addition during cooking, thickness of the loin and cooking methodology could have had an influence in these results, reducing the perception of boar taint to levels that are acceptable by consumers. Lack of differences in consumers' acceptability in meats from boars and gilts has also been reported (Aluwe et al., 2009; Smith, Ellis, Clark, & Innes, 1983). Considering the percentage of consumers represented in this cluster, 62.1% of consumers in the first consumer test and 49.5% of the consumers in the second one were indifferent towards boar taint. These findings could be related to the androstenone sensitivity since it is reported that about 40% of consumers are not able to perceive androstenone (Panella-Riera et al., 2010). Blanch et al. (2012) estimated the percentage of consumers that could reject tainted meat, and also
144
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145
reported a range of consumers classified as insensitive or low sensitive to androstenone in ES, UK and FR (no data was provided from IT): 52.5–79.2% in ES, 52.4–79.6% in UK and 43.1–73.6% in FR. In the two consumer tests, the percentage of consumers that were indifferent towards boar taint were close to the values associated to insensitive or low sensitive to androstenone. ▪ Cluster 2 (Boar meat lovers) rated the meat from gilt (FE) with lower scores, and the meat with MBT or HBT with higher scores. Meat with LBT was scored in-between in both cases. This cluster represents 21.7% of consumers in consumer test 1, and 12.4% in consumer test 2. The lower percentage of consumers that are part of this cluster in study 2 may be explained because the highest level of boar taint meat was higher in comparison to the levels in consumer study 1. In fact, it is reported that some of the sensitive consumers like the androstenone smell, and such preference could influence their acceptability of the meat (Blanch et al., 2012; Bonneau & Chevillon, 2012; Font-i-Furnols, Gispert, Diestre, & Oliver, 2003). Blanch et al. (2012) estimated the percentage of consumers that were highly sensitive to androstenone and that revealed that they liked the meat (middle and high sensitive + I like/neutral) in ES, UK and FR: 6.9% in ES, 17.0% in UK and 16.0% in FR. These percentages are close to the percentage of consumers represented in this cluster. ▪ In cluster 3 (Reject boar tainted meat) meat from gilt received the highest (best) score in both consumer tests, while meat with medium levels (MBT in study 1) or high levels (HBT in study 2) received the lowest (worst) score. This cluster represents 16.2% of consumers for the consumer test 1, and 20.0% in consumer test 2. These results are also in line with the results obtained in cluster 2. Since the levels of boar taint were higher in consumer test 2, and due to the fact that the consumers that tasted this meat were not used to boar meat, it was expected to find a higher percentage of consumers rejecting the boar tainted meat. It is interesting to highlight that the type of meat scored below 5 (in a 9-point scale) was HBT for consumer test 2, and MBT for consumer test 1. In both cases, the meat with LBT was scored in-between, but with scores around 5. Therefore, these consumers rejected meat with medium or high boar taint, but not the type of meat with low boar taint. The results of this cluster are in agreement with those reported by Babol, Squires, and Gullett (2002) and Lundström, Malmfors, Fjelkner-Modig, and Szatek (1982). Additionally, in consumer test 2 (FR & IT) there was a fourth cluster identified and labelled as consumers that Reject low boar tainted meat, representing 18.2% of the sample. In this one, the meat with low boar taint received the lowest (worst) score in all the attributes (between 4.3 and 5.1). The identification of this cluster could be expected if we take into account that it has been found in the two countries where consumers are not used to the meat from entire male pigs although it is strange that meat with HBT had high acceptability scores and similar to the acceptability observed for the FE meat. Considering the percentage of consumers being part of each cluster, a percentage of 16.2% (from cluster 3) in consumer study 1 (ES & UK) and 38.2% (from cluster 3 plus 4) in consumer study 2 (FR & IT) could react negatively when eating meat from entire male pigs. Again, to understand these differences two facts should be taken into account: (i) meat with boar taint used in FR and IT had higher boar taint levels than the meat used for ES and UK, and (ii) consumers from FR and IT were not used to eat meat from entire male pigs. 5. Conclusions and implications The present study shows that, when eating fresh pork (loin 0.5 cm thick with 5 mm back fat, cooked in a hot plate and served uncovered), different groups of consumers could be identified on the basis of ‘How delicious do you find this meat’. Among them, a group of consumers that like pork meat (49.5–62.1% of the participants), a group of
consumers that prefer meat with boar taint (12.4–21.7%) and a group of consumers that reject boar tainted meat (16.2–20.0%) have been identified. Results suggest that there is a group of consumers who may reject meat with boar taint, but there is also a niche for meat from medium and high levels of boar taint in the cities where the consumer tests were performed. Further research is needed to understand how to address this niche. The present research also identifies a need to develop tools to select and classify carcasses on the basis of boar taint level so that it can be properly addressed to the meat market. This situation can influence pork consumption so that it might guarantee that consumers do not purchase pork with inadequate sensory quality, according to their preferences. Acknowledgements This study was financially supported by the Health and Consumers Directorate-General, Animal Health and Welfare (DG-SANCO) of the European Commission (Contract No. SANCO/2008/D5/018). Complementary financial supports were provided at the national levels by INAPORC in France and IRTA in Spain. References Aluwe, M., Millet, S., Nijs, G., Tuyttens, F. A. M., Verheyden, K., De Brabander, H. F., ... Van Oeckel, M. J. (2009). Absence of an effect of dietary fibre or clinoptilolite on boar taint in entire male pigs fed practical diets. Meat Science, 82(3), 346–352. http://dx.doi.org/ 10.1016/j.meatsci.2009.02.001. Ampuero, S., Verkuylen, B., Dahlmans, H., Hortós, M., García-Regueiro, J. A., Dahl, E., ... Harlizius, B. (2011). Inter-laboratory comparison of methods to measure androstenone in pork fat. Animal, 5(10), 1–9. http://dx.doi.org/10.1017/S1751731111000553. Babol, J., Squires, E. J., & Gullett, E. A. (2002). Factors affecting the level of boar taint in entire male pigs as assessed by consumer sensory panel. Meat Science, 61(1), 33–40. http://dx.doi.org/10.1016/S0309-1740(01)00159-0. Blanch, M., Panella-Riera, N., Chevillon, P., Font-i-Furnols, M., Gil, M., Gil, J. M., ... Oliver, M. A. (2012). Impact of consumer's sensitivity to androstenone on acceptability of meat from entire male pigs in three European countries: France, Spain and United Kingdom. Meat Science, 90(3), 572–578. Bonneau, M., & Chevillon, P. (2012). Acceptability of entire male pork with various levels of androstenone and skatole by consumers according to their sensitivity to androstenone. Meat Science, 90(2), 330–337. http://dx.doi.org/10.1016/j.meatsci. 2011.07.019. DG-SANCO (2010). European Declaration on alternatives to surgical castration of pigs. Retrieved 5 August 2013, from http://www.alternativepig.eu/partnership/ declaration.html EFSA (2004). Report AHAW/04–087: Welfare aspects of the castration of piglets. Font-i-Furnols, M. (2012). Consumer studies on sensory acceptability of boar taint: A review. Meat Science, 92(4), 319–329. http://dx.doi.org/10.1016/j.meatsci.2012.05.009. Font-i-Furnols, M., Gispert, M., Diestre, A., & Oliver, M. A. (2003). Acceptability of boar meat by consumers depending on their age, gender, culinary habits, and sensitivity and appreciation of androstenone odour. Meat Science, 64(4), 433–440. http://dx. doi.org/10.1016/S0309-1740(02)00212-7. Fredriksen, B., Font-i-Furnols, M., Lundström, K., Migdal, W., Prunier, A., Tuyttens, F. A. M., & Bonneau, M. (2009). Practice on castration of piglets in Europe. Animal, 3(11), 1480–1487. http://dx.doi.org/10.1017/S1751731109004674. Gispert, M., Oliver, M.À., Velarde, A., Suarez, P., Pérez, J., & i Furnols, M. F. (2010). Carcass and meat quality characteristics of immunocastrated male, surgically castrated male, entire male and female pigs. Meat Science, 85(4), 664–670. Guerrero, L. (1999). Estudios de consumidores: análisis de los errores más habituales. In T. C. A. Almeida, G. Hough, M. H. Damasio, & M. A. A. d. Silva (Eds.), Avances en Análisis Sensorial (pp. 121–129). Sao Pablo: Ed. VArela. Lunde, K., Egelandsdal, B., Choinski, J., Mielnik, M., Flatten, A., & Kubberod, E. (2008). Marinating as a technology to shift sensory thresholds in ready-to-eat entire male pork meat. Meat Science, 80(4), 1264–1272. http://dx.doi.org/10.1016/j.meatsci. 2008.05.035. Lundström, K., Malmfors, B., Fjelkner-Modig, S., & Szatek, A. (1982). Consumer testing of boar meat in Sweden. Swedish Journal of Agricultural Research, 13, 39–46. Macfie, H. J., Bratchell, N., Greenhoff, K., & Vallis, L. V. (1989). Designs to balance the effect of order of presentation and first-order carry-over effects in hall tests. Journal of Sensory Studies, 4(2), 129–148. http://dx.doi.org/10.1111/j.1745-459X.1989. tb00463.x. Matthews, K. R., Homer, D. B., Punter, P., Béague, M. P., Gispert, M., Kempster, A. J., ... Bonneau, M. (2000). An international study on the importance of androstenone and skatole for boar taint: III. Consumer survey in seven European countries. Meat Science, 54(3), 271–283. Mörlein, D., & Tholen, E. (2015). Fatty acid composition of subcutaneous adipose tissue from entire male pigs with extremely divergent levels of boar taint compounds — An exploratory study. Meat Science, 99, 1–7. Panella-Riera, N., Blanch, M., Gonzàlez, J., Gil, M., Gispert, M., Font I Furnols, M., & Oliver, M. A. (2010, 18–19 March). Consumers' sensitivity to androstenone in Spain and UK.
N. Panella-Riera et al. / Meat Science 114 (2016) 137–145 Paper presented at the EAAP Working Group Meeting. Production and utilization of meat from entire male pigs, Bristol, UK. Patterson, R. L. S. (1968). 5α-Androst-16-ene-3-one:—Compound responsible for taint in boar fat. Journal of the Science of Food and Agriculture, 19(1), 31–38. http://dx.doi.org/ 10.1002/jsfa.2740190107. Pauly, C., Spring, P., O'Doherty, J. V., Ampuero, S., & Bee, G. (2008). Performances, meat quality and boar taint of castrates and entire male pigs fed a standard and a raw potato starch-enriched diet. Animal, 2(11), 1707–1715. http://dx.doi.org/10.1017/ s1751731108002826. PIGCAS (2009). Report on recommendations for research and policy support. Deliverable 4.1 of the EU project PIGCAS: Attitudes, practices and state of the art regarding piglet castration in Europe. Smith, W. C., Ellis, M., Clark, J. B. K., & Innes, N. (1983). A comparison of boars, gilts and castrates for bacon manufacture. 2. Curing characteristics, bacon yield and quality. Animal Production, 37, 11–15.
145
Vold, E. (1970). Fleischproduktionseigenschaften bei ebern und kastraten IV: organoleptische und gaschromatographische untersuchungen wasserdampfflüchtiger stoffe des rücken-speckes von ebern. Meldinger fra Norges Landbrukshoegskole, 49, 1–15. Walstra, P., & Maarse, G. (1970). Onderzoek gestachlengen van mannelijke mestvarkens. Researchgroep voor Vlees en Vleeswaren TNO, IVO-rapport C-147, Rapport, 2, 30. Weiler, U., Font-i-Furnols, M., Fischer, K., Kemmer, H., Oliver, M. A., Gispert, M., ... Claus, R. (2000). Influence of differences in sensitivity of Spanish and German consumers to perceive androstenone on the acceptance of boar meat differing in skatole and androstenone concentrations. Meat Science, 54(3), 297–304.