Caracterização Da Morcela De Arroz Tradicional De Monchique E Estudo Do Prazo De Validade Com Diferentes Tecnologias De Embalagem

CARACTERIZAÇÃO DA MORCELA DE ARROZ TRADICIONAL DE MONCHIQUE E ESTUDO DO PRAZO DE VALIDADE COM DIFERENTES TECNOLOGIAS DE EMBALAGEM

CHARACTERIZATION OF TRADITIONAL COOKED BLOOD AND RICE SAUSAGE FROM MONCHIQUE AND SHELF LIFE STUDY WITH DIFFERENT PACKAGING TECHNOLOGIES

TESE APRESENTADA PARA A OBTENÇÃO DO GRAU DE DOUTOR EM ENGENHARIA ALIMENTAR

JORGE ALBERTO DOS SANTOS GUIEIRO PEREIRA

Orientadora: Doutora Teresa de Jesus da Silva Matos Coorientadores: Doutora Lídia Adelina Pó Catalão Dionísio Doutor Luís Avelino da Silva Coutinho Patarata Júri: Presidente: Reitor da Universidade de Lisboa Vogais: Doutor Luís Avelino da Silva Coutinho Patarata Professor Auxiliar com agregação Escola de Ciências Agrárias e Veterinárias Universidade de Trás-os-Montes e Alto Douro; Doutora Maria Suzana Leitão Ferreira Dias Vicente Professora Auxiliar com agregação Instituto Superior de Agronomia Universidade de Lisboa; Doutor Miguel Nuno Geraldo Viegas Santos Elias Professor Auxiliar Escola de Ciências e Tecnologia Universidade de Évora; Doutor José António Oliveira e Silva Professor Auxiliar Escola de Ciências Agrárias e Veterinárias Universidade de Trás-os-Montes e Alto Douro; Doutora Marília Catarina Leal Fazeres Ferreira Professora Auxiliar Faculdade de Medicina Veterinária Universidade de Lisboa; Doutora Teresa de Jesus da Silva Matos Professora Auxiliar Instituto Superior de Agronomia Universidade de Lisboa. LISBOA 2015

CARACTERIZAÇÃO DA MORCELA DE ARROZ TRADICIONAL DE MONCHIQUE E ESTUDO DO PRAZO DE VALIDADE COM DIFERENTES TECNOLOGIAS DE EMBALAGEM

CHARACTERIZATION OF TRADITIONAL COOKED BLOOD AND RICE SAUSAGE FROM MONCHIQUE AND SHELF LIFE STUDY WITH DIFFERENT PACKAGING TECHNOLOGIES

TESE APRESENTADA PARA A OBTENÇÃO DO GRAU DE DOUTOR EM ENGENHARIA ALIMENTAR

JORGE ALBERTO DOS SANTOS GUIEIRO PEREIRA

To my mother

Resumo ______

RESUMO

O objetivo geral da tese foi a caracterização e o estudo do prazo de validade (PV) com diferentes tecnologias de embalagem da Morcela de Arroz (MA) tradicional de Monchique. O modelo de equações de transferência de calor em estado não estacionário utilizado previu a adequação da segurança do processo térmico da MA. No global, a MA cumpre com as recomendações para a produção segura de alimentos refrigerados embalados em vácuo (VP) ou em atmosfera modificada (MAP) com um PV superior a 10 dias, no respeitante ao crescimento/produção de toxina pelo C. botulinum não proteolítico. O léxico sensorial desenvolvido revelou que o traço distintivo da MA compreende uma maior nota de odor a cominhos e hortelã. O teor de humidade e aw decresceram de forma signiﬁcativa com o tempo de armazenamento sem embalagem (WP). O pH da MA decresceu (p < 0,05) na WP e MAP durante o armazenamento. As contagens totais de mesófilos e psicrotróficos aumentaram para cerca de 7 log UFC/g aos 44 dias. Na maior parte do tempo de armazenamento, as contagens de bactérias ácido láticas e de Enterobacteriaceae foram (p < 0,05) maiores na VP do que na WP e MAP. Pseudomonas spp. foram inibidas (p < 0,05) pela MAP após 8 dias de armazenamento comparativamente à VP e WP. Globalmente, as pontuações revelaram: MAP > VP > WP. O teste de consumo indica que a WP tem um PV de 11,6 dias (95% IC: 10,4-12,9) e a VP 27,8 dias (95% IC: 26,0-29,5). O controlo rigoroso da temperatura recomendada de armazenamento da MA fatiada (3 ± 1 ºC) foi suficiente para controlar o crescimento de L. monocytogenes. Quando a MA fatiada é armazenada à temperatura recomendada (3 ± 1 ºC) o PV deve ser estabelecido com base na deterioração sensorial, uma vez que a rejeição sensorial pelos consumidores ocorre antes do produto ser considerado perigoso devido ao possível crescimento de L. monocytogenes. Os consumidores indicaram que o PV da MA fatiada a 3 ± 1 ºC é de 4,4 (IC: 3,8-5,1); 8.1 (IC: 6,8-9,5) e 10,4 (IC: 8,8-12) dias embalada em aerobiose, VP e MAP, respetivamente.

Palavras-chave: Morcela de Arroz de Monchique; Microbiota; Características físicas, químicas e sensoriais; Tecnologias de embalagem; Prazo de validade

Abstract ______

TITLE: CHARACTERIZATION OF TRADITIONAL COOKED BLOOD AND RICE SAUSAGE FROM MONCHIQUE AND SHELF LIFE STUDY WITH DIFFERENT PACKAGING TECHNOLOGIES

ABSTRACT

The over-all aim of this thesis was the characterization of traditional cooked blood and rice sausage (MA) from Monchique and the shelf life (SL) study with different packaging technologies. Unsteady-state heat transfer equations model used can predict MA thermal process safety adequacy. Generally, MA complies with recommendations for safe production of vacuum packaged (VP) and modified atmosphere packaged (MAP) chilled foods with a SL longer than 10 days, considering growth/toxin production by non proteolytic C. botulinum. Sensorial lexicon reveals that MA distinctive trait comprises higher cumin and spearmint odour note. Moisture and aw signiﬁcantly decreased with storage time without packaging (WP). pH decreased (p < 0.05) in WP and MAP (80% CO2/20% N2) during storage. Mesophilic and psychrotrophic counts increased to around 7 log CFU/g at 44 days. Mostly of storage time, lactic acid bacteria and Enterobacteriaceae counts were (p < 0.05) higher in VP than in WP and MAP. Pseudomonas spp. were (p < 0.05) inhibited by MAP after 8 days of storage comparatively to VP and WP. Sensory parameters were (p < 0.05) affected by packaging method/storage time. Globally, scores were MAP > VP > WP. Consumer test indicates that MA SL is 11.6 (95% CI: 10.4-12.9) and 27.8 (95% CI: 26.0-29.5) days for WP and VP, respectively. Sliced MA SL stored at recommended temperature (3 ± 1 ºC) should be established based on sensory spoilage, once consumer sensory rejection occurs before product is considered hazardous due to L. monocytogenes possible growth. Consumers indicated that sliced MA SL at 3 ± 1 ºC is 4.4 (CI: 3.8-5.1); 8.1 (CI: 6.8-9.5) and 10.4 (CI: 8.8-12) days for aerobic packed, VP and MAP, respectively.

Key-Words: Blood and rice Monchique sausage; Microbiota; Physicochemical and sensory characteristics; Packaging technologies; Shelf life

Acknowledgements ______

AKNOWLEDGEMENTS

I would like to express my gratitude to those that, directly or indirectly, have contributed with their good will, encouragement and dedication to accomplish this work.

I would like to express my profound gratitude to my supervisor, Professor Teresa Matos, for providing me with valuable scientific inspiration and criticism, and for always finding time to discuss and to support me.

I am deeply grateful to my Co-supervisor, Professor Lídia Dionísio, for the valuable scientific guidance in the microbiology analysis, kindness and encouragement.

I would also to express my profound gratitude to my Co-supervisor, Professor Luís Patarata, for the valuable scientific guidance and criticism, in particular, for the sensory analysis and the statistical analysis of the data, kindness, enthusiasm and for finding always time to discuss and to support me.

Furthermore my thanks for the valuable scientific support in the thermal processing go to Professor Suzana Ferreira Dias, and to Professor Isabel Sousa for providing me with valuable scientific knowledge in the texture analysis, kindness and encouragement.

I would also to express my gratitude to Doctor Marina Martins, for the scientific knowledge in the phase of fungi identification, kindness and support.

I wish to thank to Professor Rui Silva, Director of the Faculty of Sciences and Technology of the University of Algarve, for allowing me to work in their microbiology laboratory.

Also thanks to Liseta, for the logistic support in the microbiology laboratory.

I would also to express my gratitude to the members of the sensory panel and consumers for their collaboration, so needed to attain the objectives of this work.

I am also very grateful to the Monchique producers of Morcela de Arroz that accepted to collaborate in this work and especially for open their doors and for offering their facilities and personnel for the production of the batches needed for this work.

I wish to thank to FCT - Fundação para a Ciência e a Tecnologia for the financial support of this work through the PhD grant SFRH/BD/63341/2009.

I wish to thank to Direção Regional de Agricultura e Pescas do Algarve and to Direção de Serviços de Alimentação e Veterinária da Região do Algarve for their support in providing technical data and legislation.

Finally, to my father and brothers for their unconditional support and encouragement, to my wife and daughter for a lot of love, care and patience. At last, to my beloved mother, a great woman, we all miss you.

iii

Scientific publications ______

SCIENTIFIC PUBLICATIONS

Pereira, J. A., Ferreira-Dias, S. and Matos, T. J. S. (2008). Application of unsteady- state heat transfer equations to Portuguese traditional meat products from Monchique region. In Proceedings of the 54th International Congress of Meat Science & Technology (ICoMST2008), Cape Town, South Africa, 10 to 15 of August, pp. 98. URL: http://www.icomst.helsinki.fi/ICoMST2008/CD Papers/General speakers+posters-3p papers/Session3/3B/3B.16.Pereira.pdf

Pereira, J. A., Dionísio, L. P. C, Patarata, L. e Matos, T. J. S. (2011). Morcela de arroz tradicional da região de Monchique; caracterização e otimização tecnológica. In Alimentos tradicionais - Engenharia dos Biossistemas - Cem Temas de Investigação no Centenário do ISA. CEER-Centro de Engenharia dos Biossistemas, Instituto Superior de Agronomia, 319-322, Edições Colibri, Lisboa, Portugal (ISBN 978-989-689-093-3).

Pereira, J. A. S. G., Dionísio, L. P. C., Patarata, L. A. S. C. and Matos, T. J. S. (2012). Antimicrobial technology packaging solution for a traditional blood sausage. In Book of Abstracts of the II International Conference on Antimicrobial Research (ICAR2012), Lisbon, Portugal, 21 to 23 of November, pp. 281. URL: http://www.formatex.info/icar2012/abstracts/htm/260.pdf

Pereira, J. A. S. G., Sousa, I., Patarata, L. A. S. C. and Matos, T. J. S. (2013). Effect of packaging technology on texture and colour of a Portuguese traditional blood sausage from Monchique region – Morcela de Arroz. In Proceedings of the International Conference on Food and Biosystems Engineering (I. C. FaBE 2013), Skiathos Island, Greece, 30 of May to 2 of June, [vol. 1] ISBN: 978-960-9510-10-3, pp. 69-70.

Scientific publications ______

Scientific manuscripts submitted for publication

Pereira, J. A., Dionísio, L., Patarata, L. and Matos, T. J. S. (2015). Effect of packaging technology on microbiological and sensory quality of a cooked blood sausage, Morcela de Arroz, from Monchique region of Portugal. Meat Science, 101, 33-41.

Pereira, J. A., Dionísio, L., Matos, T. J. S. and Patarata, L. (2015). Sensory lexicon development for a Portuguese cooked blood sausage - Morcela de Arroz de Monchique - to predict its usefulness for a geographical certification. Journal of Sensory Studies. doi:10.1111/joss.12136.

Pereira, J. A., Silva, P., Matos, T. J. S. and Patarata, L. (2015). Shelf life determination of sliced Portuguese traditional blood sausage - Morcela de Arroz de Monchique through microbiological challenge and consumer test. Journal of Food Science. doi: 10.1111/1750-3841.12782.

Pereira, J. A., Ferreira-Dias, S., Dionísio, L., Patarata, L. and Matos, T. J. S. (201x). Application of unsteady-state heat transfer equations to thermal process of Morcela de Arroz from Monchique region, a Portuguese traditional blood sausage. Journal of Food Quality.

Pereira, J. A., Dionísio, L., Patarata, L. and Matos, T. J. S. (201x). Multivariate shelf life modelling in a cooked blood and rice sausage during aerobic and vacuum package storage. (To be submitted for publication).

Financial support ______

FINANCIAL SUPPORT

This work received financial support from FCT - Fundação para a Ciência e a Tecnologia through PhD grant SFRH/BD/63341/2009.

vii

ÍNDEX

Acknowledgements iii Scientific publications v Financial support vii

CHAPTER 1 GENERAL INTRODUCTION 1 1.1 AIM AND RESEARCH APPROACHES OF THIS WORK 6 1.2 REFERENCES 7

CHAPTER 2 BIBLIOGRAPHIC REVIEW 13 2.1 MICROBIOTA OF COOKED MEAT SAUSAGES 15 2.2 FOOD PACKAGING 17 2.2.1 Packaging materials 19 2.2.2 Packaging gases 20 2.2.2.1 Effect on microorganisms 20 2.2.2.2 Effect on colour 22 2.3 THERMAL PROCESS OF MEAT PRODUCTS 23 2.3.1 Thermobacteriology 25 2.3.2 Foods REPFED 26 2.3.2.1 Spore former pathogens of concern in REPFEDs 26 2.3.2.1.1 Clostridium botulinum 27 2.3.2.1.2 Clostridium perfringens 29 2.3.2.1.3 Bacillus cereus 31 2.4 MICROBIOLOGICAL CHALLENGE TESTING 34 2.5 QUANTITATIVE DESCRIPTIVE ANALYSIS (QDA) 37 2.6 REFERENCES 39

CHAPTER 3 APPLICATION OF UNSTEADY-STATE HEAT TRANSFER EQUATIONS TO THERMAL PROCESS OF MORCELA DE ARROZ FROM MONCHIQUE REGION, A PORTUGUESE TRADITIONAL BLOOD SAUSAGE 67 3.1 ABSTRACT 69 3.2 INTRODUCTION 70 3.3 MATERIALS AND METHODS 72 3.3.1 Processing of the MA 72 3.3.2 Sampling 73 3.3.3 Cooking loss 73 3.3.4 Microbiological analysis 74 3.3.5 pH and water activity assays 74 3.3.6 Proximate composition of MA 74 3.3.7 Unsteady state heat transfer equations (USHTE) using Heisler charts 75 3.4 RESULTS AND DISCUSSION 76 3.4.1 MA characterization and processing 76 3.4.2 Temperature/time prediction using Heisler charts 81

3.4.3 Microbiological analysis of MA 82 3.5 CONCLUSIONS 83 3.6 ACKNOWLEDGEMENTS 84 3.7 REFERENCES 84

CHAPTER 4 SENSORY LEXICON DEVELOPMENT FOR A PORTUGUESE COOKED BLOOD SAUSAGE - MORCELA DE ARROZ DE MONCHIQUE 91 4.1 ABSTRACT 93 4.2 PRACTICAL APPLICATIONS 93 4.3 INTRODUCTION 94 4.4 MATERIALS AND METHODS 95 4.4.1 Formation of the panel 95 4.4.2 Samples of MA 96 4.4.3 Identification and selection of descriptors 98 4.4.4 Evaluation of the samples 99 4.4.5 Data analysis 99 4.5 RESULTS AND DISCUSSION 100 4.6 CONCLUSIONS 111 4.7 ACKNOWLEDGEMENTS 111 4.8 REFERENCES 111

CHAPTER 5 EFFECT OF PACKAGING TECHNOLOGY ON MICROBIOLOGICAL AND SENSORY QUALITY OF A COOKED BLOOD SAUSAGE, MORCELA DE ARROZ, FROM MONCHIQUE REGION OF PORTUGAL 115 5.1 ABSTRACT 117 5.2 INTRODUCTION 118 5.3 MATERIALS AND METHODS 119 5.3.1 Formulation and processing of the MA 119 5.3.2 Packaging of the MA 120 5.3.3 Sampling procedure 120 5.3.4 Physic-chemical analysis 121 5.3.4.1 Analysis of gas composition 121 5.3.4.2 pH, moisture and aw parameters 121 5.3.5 Microbiological analysis 122 5.3.6 Sensory analysis 122 5.3.6.1 Quantitative descriptive analysis (QDA) 122 5.3.6.1.1 Formation of the panel 122 5.3.6.1.2 Generation of vocabulary 123 5.3.6.1.3 Evaluation of the samples 124 5.3.7 Statistical analysis 124 5.4 RESULTS 125 5.4.1 Physic-chemical analysis 125 5.4.2 Microbiological analysis 127 5.4.3 Sensory analysis 130 5.5 DISCUSSION 133 5.6 CONCLUSIONS 137 5.7 ACKNOWLEDGEMENTS 137 5.8 REFERENCES 137 x

CHAPTER 6 MULTIVARIATE SHELF LIFE MODELLING IN A COOKED BLOOD AND RICE SAUSAGE DURING AEROBIC AND VACUUM PACKAGE STORAGE 141 6.1 ABSTRACT 143 6.2 PRACTICAL APPLICATIONS 144 6.3 INTRODUCTION 144 6.4 MATERIALS AND METHODS 147 6.4.1 Samples of MA 147 6.4.2 Lexicon development to sensory analysis 148 6.4.3 Consumer test 150 6.4.4 Sensory analysis - quantitative descriptive analysis 150 6.4.5 Microbiological analysis 151 6.4.6 Chemical indicators of spoilage 151 6.4.7 Data analysis 151 6.5 RESULTS AND DISCUSSION 152 6.6 CONCLUSIONS 162 6.7 ACKNOWLEDGEMENTS 163 6.8 REFERENCES 163

CHAPTER 7 SHELF LIFE DETERMINATION OF SLICED PORTUGUESE TRADITIONAL BLOOD SAUSAGE - MORCELA DE ARROZ DE MONCHIQUE THROUGH MICROBIOLOGICAL CHALLENGE AND CONSUMER TEST 169 7.1 ABSTRACT 171 7.2 PRACTICAL APPLICATIONS 172 7.3 INTRODUCTION 172 7.4 MATERIALS AND METHODS 173 7.4.1 Samples and experimental design 173 7.4.2 Microbiological and physicochemical analysis of fresh MA 175 7.4.3 Microbiological challenge test 176 7.4.4 Consumer test - sample preparation and analysis 177 7.4.5 Statistical analysis 178 7.5 RESULTS AND DISCUSSION 178 7.6 CONCLUSIONS 183 7.7 ACKNOWLEDGEMENTS 183 7.8 REFERENCES 184

CHAPTER 8 FINAL CONSIDERATIONS 189 8.1 CONCLUSIONS 191 8.2 REFERENCES 197

TABLES ÍNDEX

Table 3.1 - Thermal and physical properties considered for MA (Earle, 1988; Geankoplis, 1993; McCabe, Smith and Harriott, 1993). 76

Table 3.2 - Mean and standard deviation (sd) of physicochemical parameters and proximate composition of MA. 77

Table 3.3 - Thermal process parameters of MA (mean ± sd in four batches). 79

Table 3.4 - Experimental temperature and time and predicted temperature and time of MA in four batches. 81

Table 4.1 - List of terms identified by the panellists in the first open sessions and respective geometric mean obtained after analysis with a 6-point scale. 101

Table 4.2 - Factor loadings and communalities of the first three components after varimax normalized rotation for the initial solution (with 42 variables) and for the final solution (with the selected 14 variables). 103

Table 4.3 - Results of the logistic regression model for the origin of Morcela de Arroz (Monchique in relation to other origins) considering the scores PC1 and PC3 as predictors. 108

Table 4.4 - Descriptors selected to characterize Morcela de Arroz (MA), definitions and references used. 109

Table 4.5 - Sensory attributes of Morcela de Arroz from Monchique and other origins (Mean  Standard deviation) (Mean  Standard deviation). 110

Table 5.1 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on sensory parameters of Morcela de Arroz (MA). Results are presented as mean ± standard deviation (n = 9 for each packaging/storage time). 132

Table 5.2 - Selected Pearson correlations between microbial counts and sensorial evaluation of global freshness (r, p). 133

Table 6.1 - Procedure used for identification and selection of descriptors (adapted from ISO 11035, 1994). 149

xii

Table 6.2 - List of terms identified by the panellists in the first open sessions and respective geometric mean obtained after analysis with a 6-point scale. 153

Table 6.3 - Censoring of data and Weibull distribution parameters µ and σ and the estimated shelf life (50% of probability of consumers’ acceptance) for MA stored WP and VP at 4 ± 1 ºC. 156

Table 6.4 - Sensory, microbiological and chemical characteristic (mean ± standard deviation) of MA in the last storage time where samples are accepted and the first where the samples are rejected by the consumers, assessed by survival analysis on consumer data, in MA stored without package and vacuum packed. 157

Table 6.5 - Results of the logistic regression model for the acceptability of MA by the consumers considering the scores PC1 and PC2 as predictors. 161

Table 7.1 - Microbiological and physicochemical analysis of fresh MA (Mean ± standard deviation), n = 5. 178

Table 7.2 - Weibull distribution parameters µ and σ and the estimated shelf life (50% of probability of consumers’ acceptance) for sliced MA stored at 3 ± 1 ºC in different packaging conditions. 182

xiii

FIGURES ÍNDEX

Figure 3.1 - Thermal process of MA. 78

Figure 4.1 - Overview of the procedure used for identification and selection of descriptors (adapted from ISO 11035, 1994). 97

Figure 4.2 - Loadings for the PC1-PC2 (a) and PC1-PC3 (b) dimensions, after varimax normalized rotation, of the 14 variables selected. 105

Figure 4.3 - Scores of Monchique Morcela de Arroz (M) and Other origin Morcela de arroz (O) on PC1-PC2 (a) and PC1-PC3 (b) dimensions obtained from the 14 variables selected. 106

Figure 4.4 - Sensory profile of Morcela de Arroz from Monchique (full line) and other origins (dashed line). 110

Figure 5.1 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on moisture content (%). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-F, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 125

Figure 5.2 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on water activity (aw). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 126

Figure 5.3 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on pH. Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-C, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 126

xiv

Figure 5.4 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20 %N2) and of storage time (44 days) at 4 ± 1 ºC on mesophilic total viable counts (MTVC) (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 127

Figure 5.5 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on psychrotrophic total viable counts (PTVC) (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 128

Figure 5.6 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on lactic acid bacteria (LAB) (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 128

Figure 5.7 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on Enterobacteriaceae (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-c, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-C, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 129

Figure 5.8 - Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on Pseudomonas spp. (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-c, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-C, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 130

Figure 6.1 - Hedonic evaluation of MA WP (dark grey) and VP (light grey) for 0, 8, 15, 22, 30, 44 days of storage at 4 ± 1 ºC (n = 102 consumers). VP at day zero does not exist; for comparison purposes fresh unpacked samples were used. A, B, C - bars with different letters are different (p < 0.05) between storage times for each package. 154 xv

Figure 6.2 - Loadings of the variables (upper part) and scores of the samples WP and VP with different storage times (lower part) for the PC1-PC2 dimensions, after varimax rotation, of the 22 variables used to study the spoilage of MA. 160

Figure 7.1 - Experimental design for shelf life determination based on a microbiological challenge test (MCT) and a consumer test. 174

Figure 7.2 - Counts of L. monocytogenes (Log CFU/g) in MA slices packaged in aerobic condition (AP), vacuum packaging (VP) and modified atmosphere packaging (MAP, 80% CO2 and 20% N2) stored at 3 ± 1 ºC. Results are presented as mean ± sd (n = 3). 179

Figure 7.3 - Counts of L. monocytogenes (Log CFU/g) in MA slices packaged in aerobic condition (AP), vacuum packaging (VP) and modified atmosphere packaging(MAP, 80% CO2 and 20 % N2) stored at 7 ± 1 ºC. Results are presented as mean ± sd (n = 3). a-b, means with different lowercase letters are different (p < 0.05) (between packaging technologies at the indicated storage time). A-E, means with different uppercase letters are different (p < 0.05) (between storage periods for each packaging condition) (p < 0.05). 179

xvi

CHAPTER 1

General Introduction

CHAPTER 1 ______

1. GENERAL INTRODUCTION

In many European countries the importance of traditional foods has been increased (Röhr et al., 2005). Traditional foods, being part of a gastronomy heritage (Toldrá and Navarro, 2000), are inevitably associated with tourism across Europe, traduced in a renewed interest of consumers in typical regional food. Produced in small scale units, which rely on traditional manufacturing processes, with a limited degree of mechanization, is strongly identiﬁed with a place or region of origin (Talon et al., 2007).

Meat products from Serra de Monchique (south of Portugal), are part of a historic and cultural patrimony that is important to preserve and about which little information exists. Information and subsequent valorisation of these products has occurred essentially through local annual fairs and restaurants, resulting in a growing economic importance to the region. Nowadays commercialization is traduced by direct sale to consumers, retailers and wholesalers. In this context, in the last decade, and due to legal enforcement (Portuguese Law n.º 41/2001; Portuguese Law n.º 110/2002 and European Regulation No 2074/2005) several production units with a semi-industrial dimension have emerged (Portuguese Law nº 209, 2008). At the present time there are four authorized meat plants according to the Regime de Exercício de Atividade Industrial1 (REAI) (Portuguese Law nº 209, 2008), which are classified as Tipo 3 - Atividade Produtiva Similar2. Some other producers are considering applying for the authorization process (REAI, Tipo 3 - Atividade Produtiva Local3). From the range of meat products produced in this region, namely chouriço, morcela and farinheira de milho, Morcela de Arroz is one of the most important products. From the third of September of 2012, the Portuguese Law nº 209/2008 was replaced by the Portuguese Law nº 169/2012 which approves the Sistema da Indústria Responsável4 (SIR) in which Tipo 3 - Atividade Produtiva Similar2 and Tipo 3 - Atividade Produtiva Local3 are both included in Tipo 3. The Atividade Produtiva Local3 comprises a contracted electrical potency < 15kVA and a thermal potency ≤ 4 x 10 5 kJ/h and no more than 5 workers, with an annual production limit of 2000 kg.

1 Regulations for the Exercise of Industrial Activity. 2 Type 3 - Similar Productive Activity. 3 Type 3 - Local Productive Activity. 4 Responsible Industry System.

3 CHAPTER 1 ______

Morcela de Arroz or molho is a ready-to-eat cooked blood and rice sausage, it is made with minced meats (loin, shoulder, neck and bacon), rice, pork blood, grinded onion, spearmint, salt and ground spices (cumin, allspice, clove and black pepper), all stuffed in natural casings (pork stomach, pork or bovine intestine), tied with thread and then boiled in the molho de moura, a sauce prepared with water, whole onions, whole garlic, salt and laurel leaves (Pereira, Ferreira-Dias and Matos, 2008). In this traditional blood and rice sausage, like in Cavourmas in Greece (Arvanitoyannis et al., 2000), meat is added and does not undergo a drying or smoking processes. Is different from other traditional blood sausages namely the cooked Spanish Morcilla de Burgos (Santos et al., 2003) in which no meat is added or Spanish Morcilla that undergoes a ripening and smoking process before cooking (Cachaldora et al., 2013) and also from Portuguese Morcela de assar, which only includes fatty trimmings, blood and spices and undergoes a short smoking period after the cooking process (Roseiro et al., 1998) or Spanish onion chorizo (Castaño et al., 2002) with great similarity with pumpkin black pudding which are submitted to a drying process.

Traditional meat products are seen by the consumer as safe products however, considering that fresh meats are frequently contaminated (Dickson and Anderson, 1992), there is a real possibility that pathogenic microorganisms pass through the barriers used during normal processing and being in the final product. In most of these products, safety is assured controlling/preventing the growth of pathogens during process and reducing contamination/colonization to the lowest possible level (Barbuti and Parolari, 2002). Deficient hygienic conditions during process or an insufficient thermal treatment may compromise product food safety (Oteiza et al., 2006). Product temperature profile determination is crucial to guarantee product safety and to optimize thermal process (Mckenna et al., 2006). However, post process recontamination through contaminated surfaces for instance, is a possibility. Subsequent growth during storage of pathogenic may also occur as a consequence of incorrect temperatures application (Ghazala et al., 1995). Many pathogenic can grow at temperature of 3-12 ºC and produce toxins without visible deterioration (Smith et al., 1990; Mossel, Moreno and Struijk, 2003). In order to promote food safety (European Regulation No 178/2002) is crucial to study the microbiota of these meat products and its processing.

4 CHAPTER 1 ______

The EC is involved in the protection of traditional foods quality from specific regions to promote rural areas and support local population. Special labels, namely Protected Designation of Origin (PDO) and Protected Geographical Indication (PGI), have been created in the EC (European Regulation No 2081/1992; European Regulation No 510/2006). In this context, producers of Monchique are interested to apply for one of these special labels certification and, to do so, sensorial characterization and physical and chemical determinations of traditional MA from Monchique region are imperative.

Morcela de Arroz is produced without addition of chemical preservatives and without the use of any packaging technology. Distribution and storage are performed in a very short period (one week) under very limited conditions (0-5 ºC), thus limiting the commercialization period and compromising the production by the local producers, whose are very interested in the extension of the shelf life of this product. The use of packaging technology can extend shelf life of meat products by improving safety and stability (Santos et al., 2005). Vacuum packaging (VP), being cost-effective and easy to apply (McMillin, 2008), has been so far the most widely used packaging technique for different types of meat products (Korkeala et al., 1985; Borch, Kant-Muermans and Blixt, 1996; Samelis, Kakouri and Rementzis, 2000; Ansorena and Astiasarán, 2004; Summo, Caponio and Pasqualone, 2006). However, it presents some disadvantages, such as product deformation and exudate forming. To overcome this problem modified atmosphere packaging (MAP) has become popular among cooked (Hintlian and Hotchkiss, 1987; Pexara, Metaxopoulos and Drosinos, 2002; Gokoglu et al., 2010) and cured meat products (Viallon et al., 1996, Rubio et al., 2007a, 2007b; Gök, Obuz, and

Akkaya, 2008; Summo et al., 2010). CO2/N2 atmospheres are referred as particularly adequate for the preservation of cooked and cured meat products, mainly because of the strong inhibition of growth of the great majority of microorganisms (Matos et al., 2004, 2005; Santos et al., 2005; Summo et al., 2010). However, studies performed on shelf life of meat products packed in VP versus MAP reported by several authors are not consensual about the longer shelf life of MAP products (Blinckstad and Molin, 1983; Samelis and Georgiadou, 2000; García-Esteban, Ansorena and Astiasarán, 2004; Cilla et al., 2006; Rubio et al., 2006; Valencia, Ansorena and Astiasarán, 2006; Rubio et al., 2007a; Parra et al., 2010) and there is a complete lack of information for Morcela de Arroz.

5 CHAPTER 1 ______

1.1 AIM AND RESEARCH APPROACHES OF THIS WORK

The over-all aim of this thesis was the characterization of traditional cooked blood and rice sausage from Monchique region and shelf life study with different packaging technologies.

In this context, processing technology and product profile temperature were studied, unsteady-state heat transfer equations (USHTE) were used to predict temperature profile inside the product during thermal processing, the cooking loss and proximate composition were determined. A trained sensorial panel was established, the sensorial characteristics of this traditional cooked blood and rice sausage were determined and consumer tests performed. Microbiota was characterized, several physicochemical parameters were determined and different packaging technologies, vacuum (VP), modified atmosphere packaging (MAP) using 80% CO2/20% N2 and without packaging (WP), were employed and compared in order to extend the shelf life of the product. The following sub-objectives and approaches were performed to reach this aim:

(i) Characterization of processing technology, Morcela de Arroz profile temperature, cooking loss, proximate composition, adequacy assess of unsteady-state heat transfer equations (USHTE) using Heisler charts to predict temperature in the center of the product and the time needed to attain the necessary temperature were determined. The thermal treatment traditionally applied to this cooked blood sausage was evaluated to study if it complies with the recommendations for the safe production of refrigerated processed foods of extended durability (REPFEDs) (Chapter 3).

(ii) Build up terminology to develop a quantitative descriptive analysis for Morcela de Arroz from Monchique region in comparison to other blood and rice sausages, in order to improve its sensory description and to support the requirement for one of the EC special labels: PDO or PGI (Chapter 4).

(iii) Evaluation of the microbiological, physicochemical and sensory characteristics of Morcela de Arroz packaged in VP, MAP (80% CO2 and 20% N2) and WP stored at 4 ± 1 ºC in order to investigate the best packaging solution to extend the shelf life period (Chapter 5).

6 CHAPTER 1 ______

(iv) Model the spoilage in Morcela de Arroz WP and VP stored at 4 ± 1 ºC by a multivariate approach, using quantitative descriptive analysis, microbiological and chemical indicators of spoilage and the consumer acceptability, in order to determine MA shelf life and an attempt to find spoilage indicators that might be used to predict shelf life (Chapter 6).

(v) Determination of sliced Morcela de Arroz shelf life considering both the behaviour of Listeria monocytogenes through a microbiological challenge test and the consumer acceptability of stored MA: vacuum packed (VP), modified atmosphere packed (MAP) using 80% CO2/20% N2 and aerobic packed (AP) (Chapter 7).

1.2 REFERENCES

Ansorena, D. and Astiasarán. I. (2004). Effect of storage and packaging on fatty acid composition and oxidation in dry fermented sausages made with added olive oil and antioxidants. Meat Science, 67, 237-244.

Arvanitoyannis, I., Bloukas, J. G., Pappa, I. and Psomiadou, E. (2000). Multivariate data analysis of Cavourmas- a Greek cooked meat product. Meat Science, 54, 71-75.

Barbuti, S. and Parolari, G. (2002). Validation of manufacturing process to control pathogenic bacteria in typical dry fermented products. Meat Science, 62, 323-329.

Blinckstad, E. and Molin, G. (1983). The microbial flora of smoked pork loin and frankfurter sausage stored in different gas atmospheres at 4 ºC. Journal of Applied Bacteriology, 54, 45-56.

Borch, E., Kant-Muermans, M. L. and Blixt, Y. (1996). Bacterial spoilage of meat and cured meat products. International Journal of Food Microbiology, 33, 103-120.

Cachaldora, A., García,. G., Lorenzo, J. M. and García-Fontán, M. C. (2013). Effect of modiﬁed atmosphere and vacuum packaging on some quality characteristics and the shelf life of morcilla, a typical cooked blood sausage. Meat Science, 93, 220-225.

Castaño, A., García-Fontán, M. C., Fresno, J. M., Tornadijo, M. E. and Carballo, J. (2002). Survival of Enterobacteriaceae during processing of Chorizo de cebolla, a Spanish fermented sausage. Food Control, 13, 107-115.

7 CHAPTER 1 ______

Cilla, I., Martínez, L., Beltrán, J. S. and Roncalés, P. (2006). Dry-cured ham quality and acceptability as affected by the preservation system used for retail sale. Meat Science, 73, 581-589.

Dickson, J. S. and Anderson, M. E. (1992). Microbiological decontamination of food animal carcasses by washing and sanitizing systems. A review. Journal of Food Protection, 55, 133-140.

European Regulation No 178 (2002). Regulation of the European Parliament and of the Council of 28th January 2002. Official Journal of the European Communities.2002, L31, 1-24.

European Regulation No 510 (2006). Council Regulation of 20th March 2006. Official Journal of the European Union.2006, L93, 12-25.

European Regulation No 2073 (2005). Commission Regulation of 15th November 2005. Official Journal of the European Union.2005, L338, 1-26.

European Regulation No 2074 (2005). Commission Regulation of 5th December 2005. Official Journal of the European Union.2005, L338, 27-59.

European Regulation No 2081 (1992). Council Regulation of 14th July 1992. Official Journal of the European Communities.1992, L208, 1-8.

García-Esteban, M., Ansorena, D. and Astiasarán, I. (2004). Comparison of modified atmosphere packaging and vacuum packaging for long period storage of dry-cured ham: effects on colour, texture and microbiological quality. Meat Science, 67, 57-63.

Ghazala, S., Ramaswamy, H. S., Smith, J. P. and Simpson, M. V. (1995). Thermal process simulations for sous vide processing of fish and meat foods. Food Research International, 28, 117-122.

Gök, V., Obuz, E. and Akkaya, L. (2008). Effects of packaging method and storage time on the chemical, microbiological, and sensory properties of Turkish pastirma - A dry cured beef product. Meat Science, 80, 335-344.

Gokoglu, N., Pinar, Y., Uran, H. and Topuz, O. S. (2010). The effect of modified atmosphere packaging on the quality and shelf life of frankfurter type-sausages. Journal of Food Quality, 33, 367-380.

Hintlian, C. B. and Hotchkiss, J. H. (1987). Comparative growth of spoilage and pathogenic organisms on modified atmosphere packaged cooked beef. Journal of Food Protection, 50, 218-223.

8 CHAPTER 1 ______

Korkeala, H., Lindroth, S., Suihko, M., Kuhmonen, A. and Penttilä, P-L. (1985). Microbiological and sensory quality changes in blood pancakes and cooked ring sausage during storage. International Journal of Food Microbiology, 2, 279-292.

Matos, T. J. S., Barreto, A. S. F. H. and Bernardo, F. M. A. (2005). Effect of shelf life period in modified atmosphere package and of processing technology on microflora of Portuguese smoked dry sausages. Revista Portuguesa de Zootecnia, 2, 15-35.

Matos, T. J. S., Bent, B. J., Barreto, A. S. F. H. and Hojberg, O. (2004). Identification of important spoilage bacteria and fungi from two types of Portuguese smoked dry sausages after shelf-life period in modified atmosphere package. In Proceedings of the 50th International Congress of Meat Science and Technology, 8-13th August, Helsinki, Finland.

Mckenna, B. M., Lyng, J., Brunton, N. and Shirsat, N. (2006). Advances in radio frequency and ohmic heating of meats. Journal of Food Engineering, 77, 215-229.

McMillin, K. W. (2008). Where is MAP going? A review and future potential of modified atmosphere packaging for meat. Meat Science, 80, 43-65.

Mossel, D. A. A., Moreno, B. and Struijk, C. B. (2003). Microbiologia de los alimentos: fundamentos ecológicos para garantizar y comprobar la integridad (inocuidad y calidad) microbiológica de los alimentos. Editorial Acribia, SA, Zaragoza, Spain.

Oteiza, J. M., Chinen, I., Miliwebsky, E. and Rivas, M. (2006). Isolation and characterization of shiga toxin-producing Escherichia coli from precooked sausages (morcillas). Food Microbiology, 23, 283-288.

Parra, V., Viguera, J., Sánchez, J., Peinado, J., Espárrago, F., Gutierrez, J. I. and Andrés, A. I. (2010). Modified atmosphere packaging and vacuum packaging for long period chilled storage of dry-cured Iberian ham. Meat Science, 84, 760-768.

Pexara, E. S., Metaxopoulos, J. and Drosinos, E. H. (2002). Evaluation of shelf life of cured, cooked, sliced turkey fillets and cooked pork sausages-“piroski”-stored under vacuum and modified atmosphere at + 4 and +10 ºC. Meat Science, 62, 33-43.

Portuguese Law n.º 41 (2001). De 9 de fevereiro. Diário da República n.º 34. I Série, 724-727.

9 CHAPTER 1 ______

Portuguese Law n.º 110 (2002). De 16 de abril. Diário da República n.º 89. I Série, 3703-3711.

Portuguese Law nº 209 (2008). De 29 de outubro. Diário da República n.º 210. I Série, 7581-7613.

Portuguese Law nº 169 (2012). De 1 de agosto. Diário da República n.º 148. I Série, 3969-4007.

Röhr, A., Lüddecke, K., Drusch, S., Müller, M. J. and Alvensleben, R. V. (2005). Food quality and safety-consumer perception and public health concern. Food Control, 16, 649-655.

Roseiro, L. C., Santos, C., Almeida, J. and Vieira, J. A. (1998). Influence of packaging and storage temperature on cured pork blood sausages shelf life. In Proceedings of the 44th International Congress of Meat Science and Technology, 30th August-4th September, Barcelona, Spain.

Rubio, B., Martínez, B., González-Fernández, C., García-Cachán, M. D., Rovira, J. and Jaime, I. (2006). Influence of storage period and packaging method on sliced dry cured beef “Cecina de Leon”: Effects on microbiological, physicochemical and sensory quality. Meat Science, 74, 710-717.

Rubio, B., Martínez, B., González-Fernández, C., García-Cachán, M. D., Rovira, J. and Jaime, I. (2007a). Effect of modified atmosphere packaging on the microbiological and sensory quality on a dry cured beef product: “Cecina de Leon”. Meat Science, 75, 515-522.

Rubio, B., Martínez, B., Sánchez, M. F., García-Cachán, M. D., Rovira, J. and Jaime, I. (2007b). Study of the shelf life of a dry fermented sausage “salchichon” made from raw material enriched in monounsaturated and polyunsaturated fatty acids and stored under modified atmospheres. Meat Science, 76, 128-137.

Samelis, J. and Georgiadou, K. G. (2000). The microbial association of Greek taverna sausage stored at 4 and 10 ºC in air, vacuum or 100% carbon dioxide, and its spoilage potential. Journal of Applied Microbiology, 88, 58-68.

Samelis, J., Kakouri, A. and Rementzis, J. (2000). Selective effect of the product type and the packaging conditions on the species of lactic acid bacteria dominating the spoilage microbial association of cooked meats at 4ºC. Food Microbiology, 17, 329- 340.

Santos E. M., Diez, A. M., González-Fernández, C., Jaime, I. and Rovira, J. (2005). Microbiological and sensory changes in “Morcilla de Burgos” preserved in air, vacuum and modified atmosphere packaging. Meat Science, 71, 249-255.

10 CHAPTER 1 ______

Santos, E. M., González-Fernández, C., Jaime, I. and Rovira, J. (2003). Physichochemical and sensory characterization of “Morcilla de Burgos”, a traditional Spanish blood sausage. Meat Science, 65, 893-898.

Smith, J. P., Toupin, C., Gagnon, B., Voyer, R., Fiset, P. P. and Simpson, M. V. (1990). A hazard analysis critical control point approach (HACCP) to ensure the microbiological safety of sous vide processed meat/pasta product. Food Microbiology, 7, 177-198.

Summo, C., Caponio, F. and Pasqualone, A. (2006). Effect of vacuum-packaging storage on the quality level of ripened sausages. Meat Science, 74, 249-254.

Summo, C., Caponio, F., Tricarico, F., Pasqualone, A. and Gomes, T. (2010). Evolution of the volatile compounds of ripened sausages as a function of both storage time and composition of packaging atmosphere. Meat Science, 86, 839-844.

Talon, R., Lebert, I., Lebert, A., Leroy, S., Garriga, M., Aymerich, T., Drosinos, E. H., Zanardi, E., Ianieri, A., Fraqueza, M. J., Patarata, L. and Laukova, A. (2007). Traditional dry fermented sausages produced in small-scale processing units in Mediterranean countries and Slovakia.1: Microbial ecosystems of processing environments. Meat Science, 77, 570–579.

Toldrá, F. and Navarro J. L. (2000). Improved traditional foods for the next century. Food Research International, 33, 145.

Valencia, I., Ansorena, D. and Astiasarán, I. (2006). Stability of linseed oil and antioxidants containing dry fermented sausages: A study of the lipid fraction during different storage conditions. Meat Science, 73, 269-277.

Viallon, C., Berdagué, J. L., Montel, M. C., Talon, R., Martins, J. F., Kondjoyan, N. and Denoyer, C. (1996). The effect of stage of ripening and packaging on volatile content and flavor of dry sausage. Food Research. International, 29, 667-674.

CHAPTER 2

BIBLIOGRAPHIC REVIEW

CHAPTER 2 ______

2. BIBLIOGRAPHIC REVIEW

2.1 MICROBIOTA OF COOKED MEAT SAUSAGES

Microbiological quality control of meat products is essential to satisfy the criteria of food safety, of hygienic quality, of organoleptic characteristics and of adequate shelf life at an acceptable cost to the consumer. It is fundamental to identify the spoilage microorganisms that will persist and/or proliferate during the storage period under packaging conditions (vacuum or modified atmosphere packaging) and which might compromise shelf life and product safety during distribution and before consumption. Shelf life is defined as the period of time for which a product remains safe and meets its quality specifications under expected storage and use conditions (ECFF, 2006). The end of shelf life of a meat product has generally been deﬁned by the formation of unacceptable off-odours/ﬂavours and/or appearance/consistency and by the maximum acceptable microbial counts (Stolzenbach, Leisner and Byrne, 2009). However, in some cases, high microbial counts may not cause evident spoilage (Korkeala et al., 1987; Samelis et al., 1998; Samelis, 2006). Microbial ecosystems of processing environments affect the type and extent of the initial contamination of spoilage and pathogenic microorganisms in meat products (Lebert et al., 2007; Talon et al., 2007; Gounadaki et al., 2008). Additionally, microbial counts during shelf life depend of the initial levels and of the extent of growth, which are affected by the type of product and the storage conditions (Borch, Kant-Muermans and Blixt, 1996; Samelis and Georgiadou, 2000). Spoilage potential is also dependent of the microbial growth rate and of the specific metabolic activity of the different species growing in meat products as well of microorganisms interactions under certain storage conditions (Broda et al., 1996; Samelis, Kakouri and Rementzis, 2000; Liu, Guo and Li, 2006).

Many studies have dealt with the microbiology of cooked meat products, identifying and characterizing the bacterial population of these products (Kotzekidou and Bloukas, 1996; Samelis and Georgiadou, 2000; Vermeiren et al., 2006; Han et al., 2010). Lactic acid bacteria (LAB) were reported by many authors as the major spoilage bacteria in cooked meat products ecosystems (Blickstad and Molin, 1983; Korkeala et al., 1985; Mäkelä and Korkeala, 1987; Korkeala, Suortti and Mäkelä, 1988; Korkeala

15 CHAPTER 2 ______and Mäkelä, 1989; Dykes, Cloete and von Holy, 1991; Mäkelä et al., 1992; Borch, Kant-Muermans and Blixt, 1996; Korkeala and Björkroth, 1997; Koort et al., 2006; Samelis, 2006; Vermeiren et al., 2006; Chenoll et al., 2007; Vasilopoulos et al., 2008; Laursen et al., 2009; Audenaert et al., 2010; Matamoros et al., 2010) causing defects such as off-ﬂavours, souring, slime formation, colour alterations, gas production and pH decrease. The rate of spoilage of these meat products can reduce shelf life and substantial financial losses to small scale producers.

Blood sausages are among cooked meat products, however it have been less studied then fermented meat products (Santos et al., 2005a, 2005b; Oteiza et al., 2006; Diez et al. 2008a, 2008b, 2009a, 2009b; Diez, Jaime and Rovira, 2009; Cachaldora et al., 2013). Santos et al. (2005a) in morcilla, a Spanish blood sausage, found that with the application of vacuum package (VP) or modified atmosphere package (MAP) typical spoilage microbiota of aerobic storage, predominantly Pseudomonas, was replaced by LAB initially deposited after heat processing, namely during cooling and product manipulation. Santos et al. (2005b), also found that heterofermentative LAB contributed actively for the spoilage of the product from different producers, especially in VP and MAP.

Blood sausages are products of difficult preservation due to their rich nutrient composition, high pH (about 6.0) and high water activity reaching values up to 0.984 (Santos et al., 2003). Although, the high temperature and cooking time, boiled in water at 95–96 ºC for around 1 h (Santos et al., 2005a), eliminates the vegetative forms of microorganisms. Nevertheless, product handling during cooling leads to post-cooking contamination by other bacteria on product surface (Mäkelä and Korkeala, 1987; Santos et al., 2005a). Chilled foods are very sensitive to microbiological contamination, growth and toxin development. Although, each food product harbours its own speciﬁc and characteristic microbiota that is a function of raw material, bacterial populations and processing effects, preservation and storage conditions, pathogens, such as Listeria monocytogenes and spore formers as psychrotrophic Clostridium botulinum and Bacillus cereus, are of particular relevance (ECFF, 2006).

16 CHAPTER 2 ______

2.2 FOOD PACKAGING

Modern food packaging objectives are beyond the preservation and protection of food from environmental factors: chemical, physical and biological up to the point of consumption (Lee, 2010). Traditionally, food packaging focused on retarding spoilage, extending shelf life and preserving quality of food (Brody et al., 2008).

In the modern food chain system, packaging constitutes an efficient tool for keeping quality of food, protecting against deteriorative effects: dehydration, lipid oxidation, colour alteration, off-ﬂavour and off-odour development, nutrient loss, texture changes, pathogenicity and other measurable factors (Kerry, O’Grady and Hogan, 2006; McMillin, 2008). Also constitutes a way of increasing product value, promoting sales (Lee, 2010) and imparting obligatory information to the modern consumer (European Regulation No 1169/2011). Consumer looks for food products of upgraded sensory quality, with increased functional and nutritional properties, combined with a traditional and wholesome image, as well as guaranteed safety, with no additives, minimal “technological” interventions, extended shelf life and use convenience (Eilert, 2005; Nychas et al., 2008; Vandendriessche, 2008; Duizer, Robertson and Han, 2009). The display of meat products in plastic materials allows consumer evaluation of the product in a hygienic, attractive and convenient package (McMillin, 2008).

In this sense, different packaging technologies have being increasingly used in the meat industry, namely vacuum packaging (VP) and modified atmosphere packaging (MAP) (McMillin, 2008; Ruiz-Capillas and Jiménez-Colmenero, 2010; Zhou, Xu and Liu, 2010; Doulgeraki et al., 2012) to extend the shelf life by improving its safety and stability; evidencing that the modiﬁcation of the extrinsic parameters of the product ecosystem, like temperature and gaseous atmosphere, is a common practice to extend shelf life.

VP being cost effective and easy to apply (McMillin, 2008), has been so far the most widely used packaging technique in different types of meat products (Korkeala et al., 1985; Kemp et al., 1988; von Holy, Cloete and Holzapfel, 1991; Kotzekidou and Bloukas, 1996; Borch, Kant-Muermans and Blixt, 1996; Korkeala and Björkroth, 1997; Ahn et al., 2000; Samelis, Kakouri and Rementzis, 2000; Bredholt, Nesbakken and Holck, 2001; Ansorena and Astiasarán, 2004; Summo, Caponio and Pasqualone,

17 CHAPTER 2 ______

2006). However, it presents some disadvantages, such as product deformation and exudate forming. To overcome this problem, MAP application to meat products stored at refrigeration temperatures has greatly increased (Penney, Hagyard and Bell, 1993; Viana, Gomide and Vanetti, 2005; Vasilopoulos et al., 2010; Curiel et al., 2011; Karabagias, Badeka and Kontominas, 2011) and became popular among cooked (Blickstad and Molin, 1983; Hintlian and Hotchkiss, 1987; Jackson et al.,1992; Devlieghere, Van Belle and Debevere, 1998; Sørheim, Nissen and Nesbakken, 1999; Pexara, Metaxopoulos and Drosinos, 2002; De Santos et al., 2007; Gokoglu et al., 2010; Mancini et al., 2010) and cured meat products (Viallon et al., 1996; Fernández- Fernández, Vázquez-Odériz and Romero-Rodríguez, 2002; García-Esteban, Ansorena and Astiasarán, 2004; Matos et al., 2005; Cilla et al., 2006; Rubio et al., 2007a, 2007b; Gök, Obuz and Akkaya, 2008; Parra et al., 2010; Summo et al., 2010).

Any successful type of MAP application requires the remove and change of normal composition of atmospheric air, and the gases surrounding the product must contain at least 20% (v/v) CO2 (Matos et al., 2005). It also requires a barrier of either of moisture and gas permeation through packaging materials to maintain a constant package environment during storage (McMillin, 2008; Zhou, Xu and Liu, 2010). Additionally, storage temperature (Portuguese Law nº 207/2008) must be controlled to insure the effectiveness of the gas mixture on microbial growth inhibition. When CO2 is used in MAP it is absorbed by water and lipid portions of meat until saturation or equilibrium is reached (Jakobsen and Bertelsen, 2002; Sivertsvik and Jensen, 2005).

Product pH and decreased temperatures increases CO2 absorption and its effectiveness as a bacteriostatic (Carr and Marchello, 1986). The full preservative effect of CO2 is only achieved when an excess of CO2 above saturation level is reached (Gill and

Penney, 1988). To allow the proper initial ratio of solid to gaseous CO2, which depends on product CO2 solubility proﬁle, temperature and package combination, a headspace gas must be approximately 1.5-2 times the meat volume (McMillian, 2008; Al-Nehlawi et al., 2013) to prevent package collapse. The effectiveness of MAP is also dependent of the barrier capacity of the package materials to the MAP gases.

18 CHAPTER 2 ______

2.2.1 Packaging materials

The quality of packaged foods is greatly determined by the properties of packaging materials, especially by the control of gas permeability. The quantity of gas transmission through a packaging material depends on several factors, namely type, area, thickness and gas permeability of the ﬁlm, differences in the partial pressure on both sides of the ﬁlm, storage temperature and relative humidity (Lee, Yam and Piergiovanni, 2008). Additionally, the increasing demand for transparency by the consumer, determines a package combining good barrier properties with product visibility and, polymer-based barriers are well suited (Lange and Wyser, 2003). High barrier packaging materials can significantly reduce adsorption, desorption and diffusion of gases and liquids and prevents the penetration of other molecules such as

O2, pressurized gas or liquid and water vapour, which are generally undesirable for food preservation, compromising shelf life (Brody et al., 2008). Barrier films, with O2 permeability less than 100 cm3/m2.24 h.atm (23 °C and 0% of relative humidity) are generally used for VP or MAP in meat industries (Lee, 2010). High-barrier films contain CO2 longer than low-barrier films, but with a cost increase (Robertson, 1993). PA (polyamide), PETP (polyethylene terephthalate) and PVC (polyvinyl chloride) have been commonly used as a barrier layer in high barrier films (Lee, 2010). A widely used method of creating an effective barrier layer is lamination or co-extrusion with a high-barrier polymer, such as ethylene vinyl alcohol (EVOH), which is embedded in the multilayered structure (Lange and Wyser, 2003) allowing to achieve an O2 3 2 permeability normally less than 1 cm /m .24 h.atm (Lee, 2010). The O2 permeability has a significant influence on colour stability and on spoilage microbiota of processed meat products during storage. Grini, Sørheim and Nissen (1992) reported no effect of illuminated storage on the surface colour of sliced bologna when the samples were 3 2 packed in films with O2 permeability lower than 10 cm /m .24 h.atm. In cooked 3 2 sausages packaged in a film with low O2 permeability (19 cm /m .24 h.atm), the growth of Brochothrix thermosphacta was similar to that of lactic acid bacteria. Using 3 2 films with higher O2 permeability (70 and 150 cm /m .24 h.atm) growth of B. thermosphacta decreased (Cayré, Garroa and Vignolob, 2005). The Combitherm XX, 115 µm thickness film, is a polyamide/ethylene-vinyl-alcohol/polyethylene multilayer

19 CHAPTER 2 ______

complex film, indicated for MAP of cooked meat products, with an O2 transmission rate of 3.0 cm3/m2.24 h.atm at 23 ºC and 75% of relative humidity (RH), carbon dioxide transmission rate of 13.5 cm3/m2.24 h.atm at 23 ºC and 75% RH and a water vapour transmission rate of 1.0 g/m2.24 h at 23 ºC and 85% RH (Wipak Walsrode GmbH & Co., Walsrode, Germany). The properties of this packaging film allow a good control of both O2 and CO2 permeability, being a good technological solution for MAP and VP. Depending on the initial quality of the product, storage temperature and packaging material, shelf life of cooked meat products can be extend by two- to four- fold with MAP (Hotchkiss and Langston, 1995).

2.2.2 Packaging gases

CO2/N2 atmospheres are referred as particularly adequate for the preservation of cooked and cured meat products, mainly because of the strong inhibition of growth of the great majority of microorganisms (Hintlian and Hotchkiss, 1987; Matos et al.,

2005; Santos et al., 2005a; Summo et al., 2010). CO2 is the most important component in the gas mixtures related with its antimicrobial activity (Devlieghere, Debevere and

Van Impe, 1998). N2 is used as a filler gas, an inert gas that is not reactive with meat pigments or absorbed by the meat products, with low solubility in water and fats, without antimicrobial effect (Gariepy et al., 1986) which maintains package integrity by its presence in the headspace (Zhou, Xu and Liu, 2010).

2.2.2.1 Effect on microorganisms

It is accepted that CO2 increases the lag phase and generation time of microorganisms, even those that can growth in the absence of oxygen. This inhibitory effect is due to the reduction of pH, inhibition of certain enzymes involved in energy production and damage of the cell membrane. The proportion of CO2/N2 in MAP is variable with the product and the desired shelf life (Mullan and McDowell, 2003). Among the several studies made with blood sausages shelf life, Santos et al. (2005a)

20 CHAPTER 2 ______

concluded that the combination of 80% CO2 and 20% N2 resulted in an improved control of spoilage microbiota and shelf life extension.

Although the optimum concentration of CO2 for maximum shelf life has not been clearly established, literature indicates that, generally, higher amounts of initial CO2 give longer shelf life and, levels of 20-60% CO2 are required for effectiveness against aerobic spoilage organisms by penetrating membranes and lowering intracellular pH (Smith, Ramaswamy and Simpson, 1990). Additionally, the bacteriostatic effect of

CO2 seems to be more effective when bacteria are in the lag phase of growth rather than when they are in the exponential phase, suggesting that the earlier the application of CO2 to the food, the longer its potential to extend shelf life (Gill and Tan, 1980;

Silliker, 1981). Besides individual bacteria vary in sensitivity to CO2 (Farber, 1991),

Gram-negative bacteria are referred to be, in general, more sensitive to CO2 than Gram-positive bacteria because most of these are facultative or strict anaerobes (Gill and Tan, 1980). However, several studies performed on shelf life of meat products packed in VP versus MAP are not consensual about the longer shelf life of MAP meat products (Bell et al., 1996; Samelis and Georgiadou, 2000; Pexara, Metaxopoulos and Drosinos, 2002; García-Esteban, Ansorena and Astiasarán, 2004; Valencia, Ansorena and Astiasarán, 2006; Cilla et al., 2006; Rubio et al., 2006, 2007a, 2008; Parra et al., 2010; Gómez and Lorenzo, 2012; Cachaldora et al., 2013 ).

Cachaldora et al. (2013) in morcilla,, a cooked blood sausage that undergoes a period of 21 days ripening and 5 days smoking before cooking, found no significant differences in shelf life (evaluation based on pH, water activity, colour, thiobarbituric acid reactive substances - TBARS formation and microbial counts) between VP and

MAP (using three different gas mixtures: 15 O2/35 N2/50 CO2, 60 N2/40 CO2 and 40

N2/60 CO2), but samples packaged in MAP using high CO2 showed the lowest values of TBARS at the end of storage (8 weeks). Also Santos et al. (2005a), in a cooked blood sausage, recommended the application of MAP using high CO2 (80 %) to preserve the product for longer periods comparatively to VP or to MAP with lower

CO2. The high CO2 was responsible for the inhibition of growth of the great majority of the microorganisms, without aﬀecting sensory properties, colour alteration and exudate increase. The authors found that, besides the CO2 inhibitory eﬀect on Pseudomonas, enterobacteria and moulds growth, it also delayed the development of spoilage lactic acid bacteria (LAB), resulting in better sensorial scores, allowing a shelf life of 42

21 CHAPTER 2 ______days. The results obtained in other cooked meat products (Blickstad and Molin, 1983; Gokoglu et al., 2010) are in accordance with the results described by Santos et al.

(2005a) about the inhibitory eﬀect of CO2.

2.2.2.2 Effect on colour

Literature reports that higher concentrations of CO2 may increase the deterioration of the bright red oxymyoglobin pigments to brown metmyoglobin. However, numerous factors affect the appearance of cooked meat, including muscle pH, content of meat pigment and inherent meat quality (Lien et al., 2002a). Finally, the exact colour of cooked meat depends on the cooking temperatures employed and the extent of protein denaturation (Gašperlin, Žlender and Abram, 2001). In cooked meats, metmyoglobin (brown) formation occurs as a consequence of oxymyoglobin denaturation during cooking process (Lien et al., 2002a). Cooking increases the rate and extent of oxidation (Sen et al., 2011) and cooked blood meat products are especially susceptible to oxidation because of high iron content (Chen et al., 1984; Igene et al., 1979). Consequently, for cooked meat products, flavour deterioration due to oxidation seems to be the main problem, being undesirable the addition of O2, because may enhance oxidative rancidity in these products (Robertson, 1993). Killinger et al. (2000) and Mancini et al. (2010) refers that the redox state of myoglobin prior to cooking plays a crucial role in determining the colour of cooked beef steaks and patties, because it determines the resistance against heat induced denaturation, with deoxymyoglobin being more resistant and metmyoglobin less resistant. It has been shown that redness (a* values) and yellowness (b* values) decreases for pork as endpoint temperatures increases (Lien et al., 2002a), once heating is accompanied by the denaturation of myoglobin and by the loss of its characteristic bright red colour. Rising lightness (L*) values positively correlate with the development of cooked colour in red meats, as endpoint temperature increases (Lien et al., 2002a, 2002b; Zhang, Lyng and Brunton, 2004).

22 CHAPTER 2 ______

2.3 THERMAL PROCESS OF MEAT PRODUCTS

The most common processing method to assure the microbiological safety of cooked foods is still heat application (Juneja and Marmer, 1999; Juneja, 2000; Juneja, and Marks, 2003; Juneja, Marks and Huang, 2003).

Traditionally, cooking was used as a multipurpose heat-processing technique for meat products allowing: i) to acquire the desired sensory characteristics, namely specific flavour, colour and texture, ii) to improve digestibility, iii) to inactivate enzymes and toxins, iv) to reduce the most resistant microorganisms of public health concern under normal conditions of distribution and storage, v) to extend shelf life by inactivating spoilage microorganisms and, vi) to reduce or eliminate the addition of chemical preservatives (Incze et al., 1999; NACMCF, 2006; Ramaswamy and Marcotte, 2006; Awuah, Ramaswamy and Economides, 2007). Although cooking has not traditionally been referred as pasteurization, a mild heat treatment (55-90 ºC) used for food preservation which inactivates vegetative forms of pathogenic and spoilage microorganisms, it is clearly capable of achieving that effect (NACMCF, 2006). In high-acid foods (pH<4.6), pasteurization intends to avoid spoilage and economic losses, although, in low-acid chilled foods (pH> 4.6), the main goal of pasteurization is the reduction of pathogens responsible for food borne illness and human diseases (Silva and Gibbs, 2010).

Product temperature profile determination on real thermal process conditions is crucial to guarantee product safety and to optimize thermal process (Mckenna et al., 2006). Effective thermal treatment of cooked ready-to-eat (RTE) meat products is one critical control point needed to ensure microbiological safety and protection against food borne diseases (Juneja, Eblen and Ransom, 2001; Borch and Arinder, 2002; Juneja, 2003). To be achieved, the binomial time/temperature parameters are critical and the cooking process should be designed to allow the required heat point in the product according to the thermal inactivation kinetics of the specific food product most heat resistant pathogen of concern (Incze et al., 1999). The critical process time/temperature parameters in cooking are influenced by a variety of factors, mainly product formulation (water activity, pH, composition and consistency) and equipment design (Holdsworth, 2004; Ramaswamy and Marcotte, 2006). In most models for

23 CHAPTER 2 ______inactivation of pathogens by heat, process validation passes through the determination of the coldest spot in the product and the development of data to demonstrate that the required binomial time/temperature parameters are met at this point (Shaw, 2004; NACMCF, 2006). Some cooking processes are applied to products prior to packaging (cook-chill or cook-package), is an assemble-pasteurize-cool-package-store or pasteurize-package-cool-store process whereas others, such as sous vide processes (package-cook), are applied to packaged product, is an assemble-package-pasteurize- cool-store process (Juneja and Snyder, 2007). When the cooking process is applied prior to packaging, prevention of post-heat process contamination with pathogens is essential for the cooking process to be considered equivalent to pasteurization (NACMCF, 2006).

The application of the recommendations of ECFF (European Chilled Food Federation) (2006) for the production of low-acid cooked-chilled foods, namely HACCP (Hazard Analysis and Critical Control Points) principles, Good Hygienic Practices, Good Manufacturing Practices and a Cooked High Care Area (CHCA), reduces the risk of post-heat treatment contamination with the most frequent and heat resistant facultative psychrotrophic vegetative pathogen of concern in RTE foods, Listeria monocytogenes (Doyle et al., 2001; Selby et al., 2006; Walls, 2006; Gandhi and Chikindas, 2007; Huang, 2007). A CHCA is a chilled area designed for a high standard of hygiene where practices relating to personnel, ingredients, equipment and environment are managed to minimise microbiological contamination. This area comprises only cooked ingredients and materials that have been decontaminated, i.e., equivalent to a time/temperature of 2 min/70 ºC for achieving a 6 log cycles reduction of Listeria monocytogenes (ICMSF, 2002; ECFF, 2006). Others, considered prime hazards in refrigerated RTE meat products, includes non-spore forming, psychrotrophic, facultative pathogens such as Yersinia enterocolitica and Aeromonas hydrophila, which can grow at refrigeration temperatures under anaerobic conditions (Hudson, Mott and Penny, 1994; Marth, 1998; Daskalov, 2006) and non-spore forming, mesophilic, facultative anaerobes such as Salmonella spp., Staphylococcus aureus and enteropthogenic strains of Escherichia coli, which can grow at storage temperature abuse (Juneja and Snyder, 2007; Nørrung and Buncic, 2008). Additionally, these pathogens may pose a health risk in a post-heat process contamination or if the food raw ingredients used are of poor microbiological quality

24 CHAPTER 2 ______

(Newell et al., 2010) and/or pasteurization is inadequate to eliminate this initial high microbial load.

2.3.1 Thermobacteriology

In the study of thermal inactivation of microorganisms two parameters defines their thermal resistance: the D-value of the process (the time in min which is required to reduce the microbial population by tenfold, 1 log cycle, at a constant temperature) which provides a quantitative index for heat resistance of microbial cells or spores and, the z-value, (the number of degrees Celsius required to reduce the D-value by a factor of 10, 1 log cycle) which is used for comparing processes at various temperatures (Singh, 2007).

Despite the possibility that pathogens may synthesize heat shock proteins that can induce a thermo tolerance response (Kalpana and Elsa, 1996; Juneja, Klein and Marmer, 1998), thermo bacteriology data reported in literature indicates that a normal pasteurization process eliminates pathogenic vegetative bacteria of public health concern in different meat product matrixes: in ground morcilla, Escherichia coli O157:H7 presented D-values of 5.5, 2.1 and 0.60 min at 54, 58 and 60 ºC, respectively and a z-value of 7.4 ºC (Oteiza, Giannuzzi and Califano, 2003), in ground pork, Salmonella spp. and Listeria monocytogens showed D-values of 0.083 and 0.085 min at 70 ºC, respectively and a z-value of 5.9 ºC (Murphy et al., 2004), in breaded pork patties, Escherichia coli O157:H7, Salmonella spp. and Listeria monocytogens presented D-values of 0.08, 0.29 and 0.43 min at 70 ºC and z-values of 5.4, 6.2 and 5.9 ºC, respectively (Osaili et al., 2007), in a commercial bologna batter, Salmonella enterica serovar typhi and Listeria monocytogens presented D-values of 0.04 and 0.08 min at 70 ºC and z-values of 5.72 and 7.04 ºC, respectively (Lamjed et al., 2006). For Yersinia enterocolitica and Aeromonas hydrophila, bibliography indicates that, generally, a few seconds at 60 ºC is sufficient to achieve one decimal reduction in their population (Silva and Gibbs, 2010). Psychrotrophic spoilage microorganisms that normally occur in chilled low-acid meat products during storage, mainly due to post-

25 CHAPTER 2 ______heat processing contamination, such as LAB, Pseudomonas spp., moulds and yeasts, are very heat sensitive.

These findings through thermobacteriology, with the optimization of the pasteurization processes, associated with improved hygienic practices, HACPP methodology, application of modified low-oxygen atmosphere/vacuum packaging and a cold storage/distribution chain made possible the production of mild heated chilled foods according to consumer demand for high-quality convenience foods, with minimal preparation time and low levels or total absence of chemical preservatives.

2.3.2 Foods REPFED

Foods known as ready-to-eat, ready meals, cook-chill foods, sous vide foods and refrigerated processed foods of extended durability (REPFEDs), safety and quality relies on a combination of a minimal heat treatment (maximum: 70-95 ºC), intended to minimize loss of sensory quality, a storage/distribution at a chilled temperature (≤ 8 ºC) under a low-oxygen modified atmosphere/vacuum and a restricted shelf life (from few days to several weeks), depending on the food and the severity of the heat treatment (Peck and Stringer, 2005).

2.3.2.1 Spore former pathogens of concern in REPFEDs

At present time, the most significant threat of foodborne illness in REPFEDs consumption is associated with the contamination by the spore former pathogens Clostridium botulinum, Clostridium perfringens and Bacillus cereus (Juneja and Snyder, 2007) as a consequence of their capacity to survive to the heat process usually given to these products and their posterior germination, outgrow and multiplication during cooling, storage and distribution. Therefore, a new strategy for the design of pasteurization protocol based on the heat resistance of the spores of these species are now being proposed for targeting the production of safe and stable low-acid chilled meat products (ECFF, 2006; FSAI, 2006; FSA, 2008). The microbial spore is a highly

26 CHAPTER 2 ______resistant dehydrated form of a dormant cell, with extreme longevity, produced under conditions of environmental stress and, as a result of quorum sensing, is the most heat resistant microbial form, being much more heat resistant than the respective vegetative form (Goul, 2006).

2.3.2.1.1 Clostridium botulinum

Although commercially REPFEDs have an excellent safety record in Europe related to foodborne botulism (Carlin et al., 2000), it remains the greatest threat associated to this type of foods. Clostridium botulinum is a heterogeneous species of four groups of Gram-positive spore-forming anaerobic bacteria that produce a botulinum neurotoxin (Juneja and Snyder, 2007). The strains causing human botulism belong to groups I (proteolytic, produces one or two of neurotoxins types A, B and F) and II (non proteolytic, produces neurotoxin types B, E or F) (Dahlsten et al., 2008). The proteolytic strains of Group I are mesophilic. Minimum temperature required to grow and produce neurotoxin is within the range of 10-12 ºC (Lynt, Kautter and Solomon, 1982; Peck and Stringer, 2005) and thus will not occur in appropriated cold distributed foods. Food botulism outbreaks due to Group I are rare and often related to insufficiently processed home preserved foods such as canned vegetables and cured meats (Peck, 2006; Dahlsten et al., 2008). Besides, the high heat resistance of proteolytic Clostridium botulinum spores is the major concern, and constitutes the reference microorganism for safe production of low-acid canned foods, in which a minimum heat treatment, or equivalent, of 3 min at 121.1 ºC is required (Awuah, Ramaswamy and Economides, 2007). However, the psychrotrophic spores of the non proteolytic strains can germinate, grow and produce neurotoxin at temperatures as low as 3 ºC (Graham et al., 1997), being a safety risk in modern RTE foods. Information on distribution of clostridia in food, especially in REPFEDs, is scarce, contamination is essentially derived from soil and, if food is correctly mishandled, the occurrence of Clostridium species is likely to be very low (Del Torre et al., 2004; Stringer, Webb and Peck, 2009). Psychrotrophic C. botulinum spores contamination levels in raw foods are typically not very high, varying from less than 1 spore/kg to 102 spores/kg (Carlin et al., 2004; Lindström, Kiviniemi and Korkeala, 2006). Although the low incidence of

27 CHAPTER 2 ______this intoxication, the mortality rate is high if not rapidly treated. Botulism outbreaks associated with non proteolytic C. botulinum have occurred most frequently with processed ﬁsh, namely salted, dried, hot smoked-vacuum packed and fermented marine products (Lindström, Kiviniemi and Korkeala, 2006), some involving temperature abuse, being European outbreaks most associated to type B neurotoxin (Peck, 2006). The heat resistance of psychrotrophic strains of C. botulinum spores is affected by the heating medium (food), pH, water activity, protein and fat content, variation between strains and presence of lysozymes in the recovery medium of heat-injured spores (Lindström, Kiviniemi and Korkeala, 2006; Peck, 2006). Psychrotrophic strains of C. Botulinum forms spores of moderate heat resistance, heating spores at 80-85 ºC for 5- 10 min inactivates the spore germination system (Lund and Peck, 1994). The D-values reported in the literature through several review articles, are scarce and often refers to fish and other seafood, because of its associated high risk of botulism: 0.43≤ D90.0ºC ≤8.2 min and 8.3≤ z ≤9.8 ºC, for types B and E (Lindström, Kiviniemi and Korkeala, 2006; Silva and Gibbs, 2010). Crab meat presented the highest heat resistance (using a mix of three strains of type B): D94.4ºC = 2.9 min and z-value = 8.6 ºC (Peterson et al.,

1997). Values reported for pork and pea broth: 1.5≤ D82.2ºC ≤ 32.3 min, 6.5≤ z ≤16.5

ºC (type B) and D82.2ºC = 0.3 min, z = 0.3 ºC (type E) (Scott and Bernard, 1982) and with added lysozyme: 28.2 ≤ D82.2ºC ≤ 2224 min and D82.2ºC = 24.2 min, for types B and E, respectively (Scott and Bernard, 1985). For turkey slurry with added lysozyme:

D90.0ºC = 0.8 min, z = 9.4 ºC (type B); D85.0ºC =1.2 min, z = 9.9 ºC (type E) (Juneja et al., 1995) and for turkey slurry with 1% salt (wt/vol) and added lysozyme; D90.0ºC = 1.1 min, D85.0ºC = 7.8 min, z = 10.1 ºC (type B) (Juneja and Eblen, 1995). Fortunately, just a small fraction of the heated-injured spores (0.1-1%) is permeable to lysozyme and in lysozyme media grow presented bi-phasic survival curves (Peck and Stringer, 2005). On targeting C. botulinum spores in pasteurization, the range of temperatures used can overcame the problem by originating resistant “shock proteins” and damaged cells (Silva and Gibbs, 2010). The water activity of the heating medium has a considerable impact on thermal destruction of C. Botulinum spores and the use of moist heat (relative humidity > 70%) in thermal processing of unpackaged foods should be considered for preventing botulism (Lindström, Kiviniemi and Korkeala, 2006). Because of the toxicity of non proteolityc C. botulinum, FSA (2008) present recommendations for the safe production of vacuum and MAP chilled foods with a shelf life longer than 10 days, with respect on growth and on toxin production by non

28 CHAPTER 2 ______proteolityc C. botulinum includes, besides refrigeration temperature of 3 ≤ 8 ºC which should be maintained in the entire cold chain, controlling factors used singly or in combination: i) a minimal thermal treatment of 10 min/90 ºC or equivalent lethality time/temperature combinations (ACMSF, 1992) achieved at the slowest heating point in the product, to ensure a 6D reduction of the initial number of spores by a factor of 106; ii) a pH ≤ 5 throughout the food and throughout all components of complex foods; iii) a minimum salt level of 3.5% in the aqueous phase throughout the food and throughout all components of complex foods; iv) an aw ≤ 0.97 throughout the food and throughout all components of complex foods; v) a combination of heat and preservative factors which can be shown consistently to prevent growth and toxin production by non proteolytic C. botulinum. Additionally, a good antitoxigenic effect has been demonstrated through the application of MAP using 65-100% CO2 in growth medium (peptone, yeast, glucose and starch broth), fresh fish and cooked turkey (Baker, Genigeorgis and Garcia, 1990; Reddy et al., 1997; Gibson et al., 2000; Lawlor et al., 2000), allowing the control of psychrotrophic strains of C. botulinum in

REPFEDs. Packaging under air or a similar O2 atmosphere does not ensure to slow/prevent toxin formation by non proteolytic C. botulinum and should be considered to present a similar botulism risk as under VP or low-oxygen MAP (Peck et al., 2008).

2.3.2.1.2 Clostridium perfringens

Another Gram-positive spore former bacteria of particular concern in REPFEDs is Clostridium perfringens, an anaerobe relatively aerotolerant (Mossel, Moreno and Struijk, 2003) that produces a heat labile enterotoxin (heating at 60 °C/5 min destroys its biological activity) during sporulation in the small intestine. C. perfringens is capable of extreme rapid growth in meat systems within a pH and temperature range of 5.0-9.0 and 6-52.3 ºC, respectively (Juneja et al., 1999). Clostridium perfringens vegetative cells are much less heat resistant than the spores, for example: D60.0ºC = 8.5,

D65.0ºC = 0.8 min, z= 7.7 ºC, spores: D90.0ºC = 30.6 D95.0ºC = 9.7, D100.0ºC = 1.9 min, z= 8.3 ºC, using a cocktail of three strains (two strains involved in food poisoning incident and one strain from reference culture) in pork luncheon roll (Byrne, Dunne and Bolton,

29 CHAPTER 2 ______

2006), and these are more heat resistant than those of psychrotrophic strains of C. botulinum, indicating that is difficult to destroy these spores by normal cooking temperatures without compromising the quality and sensory properties of foods. Additionally, C. perfringens exhibits one of the fastest growth rates of food associated bacteria, with a mean generation time of few minutes at relatively high temperatures (e.g. 37-45 ºC), being the major contributing factor leading to food poisoning. A generation time of 7.1 min was found in autoclaved ground beef held at 41 ºC (Willardsen, Busta and Allen, 1979), 39.4 and 300 min in sous vide turkey at 28 ºC and 15 ºC, respectively (Juneja and Marmer, 1996), indicating that rapid cooling of foods is critical to restrict spores germination and prevent the most common food borne illness caused by a spore former in cooked meats. Juneja, Snyder and Cygnarowicz-Provost (1994) indicate a maximum cooling time from 60 to 7 ºC of 15 h, to prevent germination and growth of Clostridium perfringens in cooked beef. Márquez-González et al. (2012) found that cooking at a heating rate of 4 ºC/h originated growth of C. perfringens in ground pork inoculated with C. perfringens spores in the product temperature range of 44-66 ºC, and during cooling under temperature abuse conditions (54.4 to 7.2 ºC within 20 h) resulted in a 2.8 log CFU/g increase in C. perfringens. According to EFSA (2005a), almost all outbreaks resulted from slow cooling of foods or holding without refrigeration, which allowed the multiplication of C. perfringens to 106-107/g and an infective dose of 108 vegetative cells with enterotoxin production. The effective prevention of food borne diseases caused by this spore former passes through appropriate cooking, wet heat is reported to be effective against C. perfringens spores germination (Wang et al., 2012), cooling rapidly through the temperature range of 55-15 °C, holding foods at temperatures <10-12 °C, and re-heating the product to an internal temperature of 72 °C before consumption (EFSA, 2005a). FSAI (2006) recommends for chilling of heat processed food best practice: beginning of chilling process in ≤ 90 min after the completion of heat processing, a maximum period of 120 min for cooling in the range of 55 to ≤10 ºC and 60 min from 10 ºC to ≤ 3 °C, followed by chilled storage at ≤ 3 °C. FDA (2009) indicates that cooked potential hazardous foods should be cooled from 57 to 21 ºC within 2 hours and within a total of 6 hours from 57 to ≤ 5 ºC.

30 CHAPTER 2 ______

2.3.2.1.3 Bacillus cereus

Bacillus cereus is a Gram-positive facultative anaerobe spore forming bacterium, with ubiquitous distribution, easily differenced from other spore formers by their inability to catabolize mannitol and by their production of lecithinase (Mossel, Moreno and Struijk, 2003). When propagated in solid surfaces, B. cereus can produce differentiated swarm cells in a wide range of growth conditions (Senesi et al., 2010).Widely distributed in nature, is commonly found in soil, plants and consequently in many raw and cooked foods: pulses and cereals, vegetables, pasta, milk and milk products, seasonings, herbs and spices, cooked and fried rice, meat and cooked meat products (Blakey and Priesti, 1980; Johnson, 1984; Kramer and Gilbert, 1989; Te Giffel et al., 1996; Valero et al., 2002; Ouoba, Thorsen and Varnam, 2008; Fekete, 2009; Chon et al., 2012; Dong, 2012). This organism, comparatively to C. perfringens, easily forms spores in foods, therefore being more probable that B. cereus spores resists at a heat treatment and subsequent spore germination/outgrowth originates food borne illness and spoilage in REPFEDS (Granum and Lund, 1997; Mossel, Moreno and Struijk, 2003). B. cereus is the etiological agent of two different clinical food poisoning syndromes in humans: emetic syndrome (vomiting) intoxication and a diarrhoeal syndrome infection. Emetic syndrome is due to the ingestion of a preformed extreme heat stable toxin (Granum and Lund, 1997), a cyclic peptide (cereulide) (Johnson, 1984; Agata, Ohta and Yokoyama, 2002; Rajkovic et al., 2008), in the food (mostly associated to starch rich foods, specially rice and rice products) (Kramer and Gilbert, 1989; Carlin et al., 2000; Rajkovic et al., 2006; Abee et al., 2011), resembling staphyloenterotoxicosis, is presumed to be unable to grow and produce the toxin below 10 °C or in the absence of oxygen (EFSA, 2005b). Diarrhoeal syndrome infection, considered more common (Nauta et al., 2003; Fekete, 2009; Logan, 2011), due to the ingestion of spores/vegetative cells present in the food (Clavel et al., 2004) (mostly associated to proteinaceous foods) which produce heat label enterotoxins: haemolysin, non-haemolyitic and cytoxin (Drobniewski, 1993; Lund and Granum, 1996; Ceupens et al., 2012). Although resembling the symptoms of C. perfringens infection, the non-haemolyitic enteroxin is considered much more potent than that produced by C. perfringens. Both syndromes requires ingestion of high numbers of B. cereus with food, being the reference dose reported of >104 cells g-1 food to be considered hazardous (Notermans and Batt, 1998; Arnesen, Fagerlund and Granum,

31 CHAPTER 2 ______

2008). Although, emetic toxin producing B. cereus strains have been identified to be distinct from non-emetic ones, because the former are not able to degrade starch or ferment salicin, these strains do not have genes encoding haemolysin and presents just weak or no haemolysis (van der Voort and Abee, 2012). Some strains of B. cereus are reported to produce both emetic and diarrhoeal enterotoxin(s) (Kim et al., 2012; Oh, Ham and Cox, 2012). The fact that both illnesses are, in general, self-limiting (24 h) contributes to the underestimation of the important role of this ubiquitous organism in many foods borne illness (Logan, 2011). In 2009 a total of 124 foodborne outbreaks attributed to Bacillus spp. were reported by 11 EU Member States, representing 2.2% of all foodborne outbreaks reported by Member States (FSAI, 2011). B. cereus is reported to have a generation time of 11-27 min, capable of growth in the pH range of 4.9-9.3 and an optimum growth temperature of 28-35 ºC with a minimum of 4-5 ºC and maximum 55 ºC (Nothermans and Batt, 1998; Borge et al., 2001; Afchain et al., 2008), revealing a great adaptability (Guinebretière et al., 2008; Carlin et al., 2010). However, strains capable to multiply below 7 °C and strains able to multiply above 45 °C, are not so common (EFSA, 2005b). Typical mesophilic strains do not normally grow at temperatures below 10 ºC, however psychrotrophic strains of this organism, although seem to be less heat resistant (Guinebretière et al., 2008), involved in food poisoning, have been isolated from foods stored at refrigeration temperatures (4-7 ºC), namely in raw and pasteurized milk (Meer et al., 1991; Dufrenne et al., 1995; Te Giffel et al., 1997; Lin et al., 1998) and in cooked chilled foods containing vegetables (Fernández et al., 1999; Carlin et al., 2000; Choma et al., 2000; Del Torre, Corte and Stecchini, 2001) and meat products (meat stew, meat and rice dishes) (Borge et al., 2001) being of special interest in the modern RTE refrigerated food industry (Nauta et al., 2003; Afchain et al., 2008). The dimension of the safety risk that B. cereus may pose within a REPFED will depend on many factors, namely spore prevalence/concentration in raw materials, heat treatment, heat resistance of spores, product formulation and supply chain storage temperatures/times (Membré, Yu and Blackburn, 2008), being the majority of foodborne cases associated to heat treated foods subjected to temperature abuse during storage and handling (FSAI, 2011). Carlin et al. (2006) reported that emetic strains (17 stains) of B. cereus, presenting spores with higher heat resistance comparatively to diarrheal type, representing a special risk in heat processed foods or in preheated foods that are kept warm, but should not pose a risk in properly refrigerated foods, once no growth was observed at 4-7 ºC. However, non-emetic

32 CHAPTER 2 ______producing strains (28 strains associated to diarrhoeal syndrome and 22 from food- environment strains) were able to grow at 4-7 ºC (Carlin et al., 2006). Cronin and Wilklinson (2009) found no growth using a mesophilic strain (diarrhoeal syndrome) in cooked rice at 4 ºC, however, growth to hazardous levels was observed at 10 ºC after 6 days without detectable aromatic or visual deterioration in the rice. The authors concluded that rice containing 1.0x102 CFUg-1 of viable B. cereus spores post-cooking can become unsafe if stored under inappropriate refrigeration temperature either by the producer or the consumer. In another study using pasta penne inoculated with two emetic strains (ca. 105 CFU g-1) involved in lethal food poisoning, no cereulide production was observed at 4 ºC during 7 days, however, at 8 ºC after 2 days cereulide was produced in small amounts by one strain and both strains died off during the long storage at 4 and 8 ºC, confirming that emetic toxin production in food is controlled by multiple factors, being temperature a key parameter (Delbrassinne et al., 2011). MAP with 40% of CO2 also seems to inhibit both growth and haemolytic enterotoxin production (Samapundo et al., 2011). Additionally, several studies (Gonzalez et al., 1999; Gounina-Allouane, Broussolle and Carlin, 2008; Planchon et al., 2011), confirmed that low sporulation temperature has a detrimental effect on B. cereus wet- heat resistance, being effective on killing B. cereus spores (Coleman et al., 2010). The heat treatment significantly increases both the time to growth and the minimal aw and pH necessary for germination and subsequent growth (Daelman et al., 2013). The decrease of spore heat resistance is also favoured by acidic sporulation conditions, acidic treatment or acidic recovery conditions (Baril et al., 2012). The emetic spores are referred as more heat resistant compared to the diarrheal type (Ankolekar and Labbé, 2009). Thermal resistance of both vegetative cells and spores of B. cereus (cocktail of three strains: one strain involved in food poisoning incident, diarrhoeal syndrome, one strain extensively used in sporulation and germination studies and one reference culture strain) reported in pork luncheon roll, vegetative cells: D55.0ºC = 6.4,

D65.0ºC = 1.0 min, z= 6.6 ºC; spores : D85.0ºC = 29.5, D90.0ºC = 10.1 D95.0ºC = 2.0, z= 8.6 ºC (Byrne, Dunne and Bolton, 2006), indicates that B. cereus spores seems to be more heat resistant than those of psychrotrophic strains of C. botulinum but less than Clostridium perfringens spores.

33 CHAPTER 2 ______

International guidelines (CAC, 1999) also report the importance of use challenge studies to confirm the effectiveness of the chosen hurdles against the pathogen(s) of concern, unless scientific evidence previously exists.

2.4 MICROBIOLOGICAL CHALLENGE TESTING

Microbiological challenge tests are useful instruments that are used by the industry to validate food safety processing, namely processes intended to inhibit target pathogen(s) growth or/and inactivation or/and to shelf life determination, based on the microbiological safety of the food, being very useful for foods that may support the growth of pathogen(s) and are chilled stored (Anonymous, 2003; NACMCF, 2010; ICMSF, 2011). Shelf life determination may also be based on the detection of microbial alteration, as well as physic-chemical and sensorial changes (Valero, Carrasco and García-Gimeno, 2012). However, for some foods, when a recontamination is probable to occur or when normal processing does not eliminate all the pathogens of concern, shelf life should be determined based on the period of time in which the pathogen(s) of concern (or toxin) will remain in harmless levels to the defined consumer (Zwietering et al., 2010). Additionally, microbiological hazard(s) may not be associated or precede sensorial (and/or physic-chemical) deterioration of foods and microbiological testing for confirmation of safety of finished foods, besides is often not a practical option because of the high number of samples needed (van Schothorst et al., 2009), has statistical and technical limitations, especially for pathogens that are normally present in low concentration and may not be detected, thus negative test results do not guarantee that a food is not contaminated with pathogens (Mossel, Moreno and Struijk, 2003). Instead, special attention on preventing cross- contamination and undercooking may have more impact on the public health than further reductions in the already small numbers of pathogens occasionally present in certain foods (IFT, 2013). Food safety objectives, which are based on risk assessments and on implemented GMP and HACCP systems, allows to integrate risk assessment and current hazard management practices, offering a practical tool to convert public health goals into values/targets that can be used by regulatory agencies and industry (IFT, 2013). In food safety management, the performance of a system implicates

34 CHAPTER 2 ______measures of the characteristics of its output and, besides microbiological testing, other approaches are necessary to effectively and efficiently reach public health food safety goals (IFT, 2013). Validation of food processes is based on the collection and evaluation of scientific, technical and observational information differing: from verification, which is used to determine that the control measures have been correctly implemented, and from monitoring, which is the continuous collection of information on a control measure at the time the control measure is applied. Validation provides a high degree of assurance that a particular process will consistently produce a food product meeting its pre-determined specifications and quality attributes, and includes: clear identification of hazards, available control measures, critical control points, critical limits and corrective actions, being fundamental for the successful implementation of HACCP system (CAC, 2008; Zwietering et al., 2010).

The principal of challenge testing is quite simple: if the goal is to eliminate/retard a target microorganism(s) (pathogenic or spoilage) growth through some technological process (hurdle) in a specific food, the solution is to challenge that target microorganism(s) in real processing conditions to check its behaviour. Challenge tests are reported to be very useful in REPFEDs, being part of the code of hygienic practice of international guidelines (CAC, 1999) for this type of foods. However, adequate challenge testing design is complex, being multifactor dependent, namely how the product is formulated, manufactured, packaged, distributed, prepared and consumed (Anonymous, 2003). The information obtained can be the basis for setting safety criteria at the critical control points in a food processing operation, but it remains necessary to verify the test results under field conditions (Notermans and in't Veld, 1994).

Food safety challenge studies are classified as: i) pathogen(s) growth inhibition studies, used to evaluate the ability of a particular food product formulation, with specific processing and packaging technology, to inhibit the growth of certain bacterial pathogen(s) under specific storage conditions (time/temperature), to demonstrate safety of a current formulation during extended shelf life under normal storage conditions during extended shelf life, or to determine whether formulation/processing changes are required if the product is subjected to temperature abuse, or to determine the effect of a modified formulation, process, or packaging technology; ii) pathogen inactivation studies, used to evaluate the ability of a particular food product formulation, a specific

35 CHAPTER 2 ______food manufacturing practice, or their combination, to cause inactivation of certain bacterial pathogen(s), held under specific storage conditions (time/temperature) and packaging. It also may be used to determine whether non-thermal technologies or combinations of pH, aw, preservatives, and holding for specified times at specific temperatures prior to release the product will provide sufficient lethality to achieve a safe food product and; iii) combined growth and inactivation studies, used to evaluate the ability of a particular food/process to inactivate certain bacterial pathogens and to inhibit the growth of certain other pathogenic bacteria or for achieving a level of inactivation followed by inhibition of the growth of survivors or contaminants introduced after processing (NACMCF, 2010). Vestergaard (2001) indicates the following factors that must be considered in a microbiological challenge study: i) selection of appropriate pathogens/surrogates, ii) level of challenge inoculum, iii) inoculum preparation and method of inoculation, iv) the duration of the study, v) formulation factors and storage conditions and, vi) sample analyses. Considering the selection of appropriate pathogens or surrogates, ideally, the microorganisms to be used are those that have been previously isolated from similar formulations and pathogens from known foodborne outbreaks to ensure if the formulation is capable to inhibit those organisms. The selection of the challenge target pathogen(s) depends on food type, food formulation, storage conditions (temperature and packaging technology), and history of the food (association with known illness outbreaks and/or evidence of potential growth). Potential pathogens of concern for growth studies in a food product with a pH > 5.4 and aw > 0.96 includes B. cereus, C. botulinum, C. perfringens, L. monocytogenes, pathogenic E. coli, Salmonella, S. aureus, V. parahaemolyticus and V. vulnificus (NACMCF, 2010). In MAP products, including meats, C. botulinum (proteolytic and non proteolytic strains) and other pathogens like Salmonella, enterohemorrhagic E. coli and L. monocytogenes are of concern (Vestergaard, 2001). However, if the product is not a seafood should not be challenged for Vibrio and, if is cooked, vegetative cells of pathogens have been eliminated and might be challenged with pathogenic spore formers. L. monocytogenes might be used if the study is designed to determine growth/inactivation due to recontamination with this organism in a RTE product. Present European recommendations (European Regulation No 2073/2005 replaced by European Regulation No 1441/2007) includes the study of the growth and/or survival of this pathogen of concern during the shelf life period to verify that the maximum concentration is < 100 ufc.g-1 through Listeria sp. challenge

36 CHAPTER 2 ______testing (Uyttendaele et al., 2004). In growth studies, the fastest growing pathogen(s) likely to be present in the food formulation should be used. Predictive models can be useful for determining which pathogen may grow faster under the conditions of the study (NACMCF, 2010). Once the appropriate target pathogen(s) have been selected should be included in the challenge study a mixture of not antagonistic multiple strains (frequently five or more) of the target pathogen(s) or surrogates, to account for potential strain variation, a single strain may not be the most resistant to each of the multiple stress factors involved in the product/process combination and strains with the shortest generation time may not have the shortest lag time under the test conditions (Vestergaard, 2001).

2.5 QUANTITATIVE DESCRIPTIVE ANALYSIS (QDA)

The concept of quality of a food is determined by several parameters: sensory attributes, chemical composition, physical properties, level of microbiological and toxicological contaminants, shelf life, packaging and labelling according to consumers requirements and acceptance (Molnár, 1995). However, in the vast majority of the cases, when shelf life of a food is the subject, normally is related with its sensory shelf life (Hough and Garitta, 2012). Besides the demand for safe food, there is always the need to consume a food with good taste (Cayot, 2007). The quality aspect related to sensory characteristics of traditional food products represent a determinant factor for its typicality (Cayot, 2007) and is a condition for product certification (European Regulation No 510/2006 replaced by the European Regulation No 1151/2012) stressing the need for sensory evaluation and to develop of testing procedures for several important commodities. Most of these procedures belong to the category of descriptive analyses, which are used to point out differences among product variants, storage conditions, identify drivers of hedonic responses and to examine relationships between chemical and sensory characteristics (Muñoz and Civille, 1998; Tuorila and Monteleone, 2009). The sensory characterization of a food product is frequently achieved by descriptive sensory analysis. Although there are several new methodological approaches developed as alternatives to conventional profiling (Valentin et al., 2012; Etaio et al., 2013), Quantitative Descriptive Analysis (QDA) (Stone et al., 1974) is still a reference technique.

37 CHAPTER 2 ______

This technique (ISO 11035, 1994) consists of a sequential survey of sensory terms for a product generated by a sensory panel (ISO 8586-1, 2001) using nontechnical language and supervised by a leader who should not be an active participant in the process (Barthélémy, 1990). This method provides a predeﬁned terminology to describe and quantify sensory perceptions as objectively as possible (Moskowitz, 1983). However, the validity of the results obtained depends on the correct initial selection of descriptors and on the ability of the judges to identify and quantify each speciﬁc attribute. Literature reports lexicon development strategies for descriptive sensory analysis of several meat and meat products (Johnson and Civille, 1986; Lyon, 1987; Rajalakshmi et al., 1987; Flores et al., 1997; Byrne et al, 1999a, 1999b; Carlucci et al., 1999; Gorraiz et al., 2000; Rodríguez- Pérez et al., 2003; Nissen et al., 2004; Rødbotten et al., 2004; Pérez-Cacho et al., 2005; Campo et al., 2006; Rason et al., 2007; Molinero et al., 2008; Patarata et al., 2008; Adhikari et al., 2011; Maughan et al., 2012; Larrauri et al., 2013; Vásquez-Araújo et al., 2013; Weerasinghe et al., 2013) and other foods namely for plant foods and ingredients (Smyth et al., 2012), rice (Bett-Garber et al., 2012), potato puree (Jiménez, Canet and Alvarez, 2013), soy sauce (Cherdchu, Chambers IV and Suwonsichon, 2013), and prebiotic gluten-free bread (Morais et al., 2014) with the objective of quality control and/or product certification. However, final food product shelf life validation depends on consumer acceptance of the food product which, in turns, will determine purchase intention. Several methods can be used to determine the sensory shelf life of food products (Giménez, Ares and Ares, 2012), all with advantages/disadvantages. Recently survival analysis was considered, both methodological and statistically, the most sound (Hough and Garitta, 2012). Nevertheless, results obtained from descriptive analysis are necessary for understanding product changes during storage. Furthermore, allows link the product characteristics with instrumental measures and with the consumer perception (Muñoz, Chambers IV and Hummer, 1996; Moussaoui and Varela, 2010; Varela and Ares, 2012; Hough et al., 2013).

38 CHAPTER 2 ______

2.6 REFERENCES

Abee, T., Groot, M. N., Tempelaars, M., Zwietering, M., Moezelaar, R. and van der Voort, M. (2011). Germination and outgrowth of spores of Bacillus cereus group members: Diversity and role of germinant receptors. Review article. Food Microbiology, 28, 199-208.

ACMSF (Advisory Committee on the Microbiological Safety of Food) (1992). Report on vacuum packaging and associated processes. Advisory Committee on the Microbiological Safety of Food. HMSO, London, U.K.

Adhikari, K., Chambers IV, E., Miller, R., Vázquez-Araújo, L., Bhumiratana, N. and Philip, C. (2011). Development of a lexicon for beef flavour in intact muscle. Journal of Sensory Studies, 26, 413-420.

Afchain, A. L., Carlin, F., Nguyen-Thé, C. and Albert, I. (2008). Improving quantitative exposure assessment by considering genetic diversity of B. cereus in cooked, pasteurised and chilled foods. International Journal of Food Microbiology, 128, 165-173.

Agata, N., Ohta, M. and Yokoyama, K. (2002). Production of Bacillus cereus emetic toxin (cereulide) in various foods. International Journal of Food Microbiology, 73, 23-27.

Ahn, D. U., Jo, C., Du, M., Olson, D. G. and Nam, K. C. (2000). Quality characteristics of pork patties irradiated and stored in different packaging and storage conditions. Meat Science, 56, 203-209.

Al-Nehlawi, A., Saldo, J., Vega, L. F. and Guri, S. (2013). Effect of high carbon dioxide atmosphere packaging and soluble gas stabilization pre-treatment on the shelf-life and quality of chicken drumsticks. Meat Science, 94, 1-8.

Ankolekar, C. and Labbé, R. G. (2009). Survival during cooking and growth from spores of diarrheal and emetic types of Bacillus cereus in rice. Journal of Food Protection, 72, 2386-2389.

Anonymous (2003). Microbiological challenge testing. Comprehensive Reviews in Food Science and Food Safety, 2, 46-50.

Ansorena, D. and Astiasarán, I. (2004). Effect of storage and packaging on fatty acid composition and oxidation in dry fermented sausages made with added olive oil and antioxidants. Meat Science, 67, 237-244.

Arnesen, L. P. S., Fagerlund, A. and Granum, P. E. (2008). From soil to gut: Bacillus cereus and its food poisoning toxins. Review article. FEMS Microbiology Reviews, 32, 579-606.

39 CHAPTER 2 ______

Audenaert, K., D'Haene, K., Messens, K., Ruyssen, T., Vandamme, P. and Huys, G., (2010). Diversity of lactic acid bacteria from modiﬁed atmosphere packaged sliced cooked meat products at sell-by date assessed by PCR-denaturing gradient gel electrophoresis. Food Microbiology, 27, 12-18.

Awuah, G. B., Ramaswamy, H. S. and Economides, A. (2007). Thermal processing and quality: principles and overview. Chemical Engineering and Processing, 46, 584-602.

Baker, D. A., Genigeorgis, C. and Garcia, G. (1990). Prevalence of Clostridium botulinum in seafood and significance of multiple incubation temperatures for determination of its presence and type in fresh retail fish. Journal of Food Protection, 53, 668-673.

Baril, E., Coroller, L., Couvert, O., Leguérinel, I., Postollec, F., Boulais, C., Carlin, F. and Mafart, P. (2012). Modeling heat resistance of Bacillus weihenstephanensis and Bacillus licheniformis spores as function of sporulation temperature and pH. Food Microbiology, 30, 29-36.

Barthélémy, J. (1990). Evaluation d` une grandeur sensorielle complexe: Description quantifiée. In Evaluation sensorielle: manuel méthodologique. Technique & Documentation-Lavoisier and APRIA (Ed.), 144-162, Librairies Lavoisier, Paris, France.

Bell, R. G., Penney, N., Gilbert, K. V., Moorhead, S. M. and Scott, S. M. (1996). The chilled storage life and retail display performance of vacuum and carbon dioxide packed hot deboned beef strip loins. Meat Science, 42, 371-376.

Bett-Garber, K. L., Lea, J. M., Champagne, E. T. and Mcclung, A. M. (2012). Whole-grain rice flavor associated with assorted bran colors. Journal of Sensory Studies, 27, 78-86.

Blakey, L. J. and Priesti, F. G. (1980). The occurrence of Bacillus cereus in some dried foods including pulses and cereals. Journal of Applied Bacteriology, 48, 291-302.

Blickstad, E. and Molin, G. (1983). The microbial flora of smoked pork loin and frankfurter sausage stored in different gas atmospheres at 4 ºC. Journal of Applied Bacteriology, 54, 45-56.

Borch, E. and Arinder, P. (2002). Bacteriological safety issues in red meat and ready- to-eat meat products, as well as control measures. Meat Science, 62, 381-390.

Borch, E., Kant-Muermans, M.-L. and Blixt, Y. (1996). Bacterial spoilage of meat and cured meat products. International Journal of Microbiology, 33, 103-120.

Borge, G. I. A., Skeie, M., Sørhaug, T., Langsrud, T. and Granum, P. E. (2001). Growth and toxin profiles of Bacillus cereus isolated from different food sources. International Journal of Food Microbiology, 69, 237-246.

40 CHAPTER 2 ______

Bredholt, S., Nesbakken, T. and Holck, A. (2001). Industrial application of an antilisterial strain of Lactobacillus sakei as a protective culture and its effect on the sensory acceptability of cooked, sliced, vacuum-packaged meats. International Journal of Food Microbiology, 66, 191-196.

Broda, D. M., DeLacy, K. M., Bell, R. G., Braggins, T. J. and Cook, R. L. (1996). Psychrotrophic Clostridium spp. associated with ‘blown pack’ spoilage of chilled vacuum-packed red meats and dog rolls in gas-impermeable plastic casings. International Journal of Food Microbiology, 29, 335-352.

Brody, A. L., Bugusu, B., Han, J. H., Koelsch, C. and McHugh, T. H. (2008). Innovative food packaging solutions. Journal of Food Science, 73, 107-116.

Byrne, B., Dunne, G. and Bolton, D. J. (2006). Thermal inactivation of Bacillus cereus and Clostridium perfringens vegetative cells and spores in pork luncheon roll. Food Microbiology, 23, 803-808.

Byrne, D. V., Bak, L. S., Bredie, W. L. P., Bertelsen, C. and Martens, M. (1999a). Development of a sensory vocabulary for warmed-over flavour: part I. In porcine meat. Journal of Sensory Studies, 14, 47-65.

Byrne, D. V., Bredie, W. L .P. and Martens, M. (1999b). Development of a sensory vocabulary for warmed-over flavour: part II. In chicken meat. Journal of Sensory Studies, 14, 67-78.

CAC (Codex Alimentarius Commission) (1999). Code of hygienic practice for refrigerated packaged foods with extended shelf life. CAC/RCP 46. Retrieved from http://www.codexalimentarius.org/download/standards/347/CXP_046e.pdf Accessed on 03-04-2013.

CAC (Codex Alimentarius Commission) (2008). Guidelines for the validation of food safety control measures. Retrieved from CAC/GL 69. Retrieved from http://www.codexalimentarius.org/download/standards/11022/CXG_069e.pdf Accessed on 03-04-2013.

Cachaldora, A., García, G., Lorenzo, J. M. and García-Fontán, M. C. (2013). Effect of modiﬁed atmosphere and vacuum packaging on some quality characteristics and the shelf life of morcilla, a typical cooked blood sausage. Meat Science, 93, 220-225.

Campo, M. M., Nute, G. R., Hughes, S. I., Enser, M., Wood, J. D. and Richardson R. I. (2006). Flavour perception of oxidation in beef. Meat Science, 72, 303-311.

41 CHAPTER 2 ______

Carlin, F., Brillard, B., Broussolle, V., Clavel, T., Duport, C., Jobin, M., Guinebretière, M.-H., Auger, S., Sorokine, A. and Nguyen-Thé, C. (2010). Adaptation of Bacillus cereus, an ubiquitous worldwide-distributed foodborne pathogen, to a changing environment. Food Research International, 43, 1885-1894.

Carlin, F., Broussolle, V., Perelle, S., Litman, S. and Fach, P. (2004). Prevalence of Clostridium botulinum in food raw materials used in REPFEDs manufactured in France. International Journal of Food Microbiology, 91, 141-145.

Carlin, F., Fricker, M., Pielaat, A., Heisterkamp, S., Shaheen, R., Salonen, M. S., Svensson, B., Nguyen-Thé, C. and Ehling-Schulz, M. (2006). Emetic toxin- producing strains of Bacillus cereus show distinct characteristics within the Bacillus cereus group. International Journal of Food Microbiology, 109, 132-138.

Carlin, F., Girardin, H., Peck , M. W., Stringer, S. C., Barker, G. C., Martinez, A., Fernandez, A., Fernandez, P., Waites, W. M., Movahedi, S., van Leusden, F., Nauta, M., Moezelaar, R., Del Torre, M. and Litman, S. (2000). Research on factors allowing a risk assessment of spore-forming pathogenic bacteria in cooked chilled foods containing vegetables: a FAIR collaborative project. International Journal of Food Microbiology, 60, 117-135.

Carlucci, A., Napolitano, F., Girolami, A. and Monteleone, E. (1999). Methodological approach to evaluate the effects of age at slaughter and storage temperature and time on sensory proﬁle of lamb meat. Meat Science, 52, 391-395.

Carr, T. P. and Marchello, J. A. (1986). Microbial changes of precooked beef slices as affected by packaging procedure. Journal of Food Protection, 49, 534-536.

Cayot, N. (2007). Sensory quality of traditional foods. Food Chemistry, 102, 445-453.

Cayré, M. E., Garroa, O. and Vignolob, G. (2005). Effect of storage temperature and gas permeability of packaging ﬁlm on the growth of lactic acid bacteria and Brochothrix thermosphacta in cooked meat emulsions. Food Microbiology, 22, 505-512.

Ceuppens, S., Van de Wiele, T., Rajkovic, A., Ferrer-Cabaceran, T., Heyndrickx, M., Boon, N. and Uyttendaele, M. (2012). Impact of intestinal microbiota and gastrointestinal conditions on the in vitro survival and growth of Bacillus cereus. International Journal of Food Microbiology, 155, 241-246.

Chen, C. C., Pearson, A. M., Gray, J. I., Fooladi, M. H. and Ku, P. K. (1984). Some factors inﬂuencing the nonheme iron content of meat and its implications in oxidation. Journal of Food Science, 49, 581-584.

Chenoll, E., Macián, M. C., Elizaquível, P. and Aznar, R. (2007). Lactic acid bacteria associated with vacuum-packed cooked meat product spoilage: population analysis by rDNA-based methods. Journal of Applied Microbiology, 102, 498-508.

42 CHAPTER 2 ______

Cherdchu, P., Chambers IV, E. and Suwonsichon, T. (2013). Sensory lexicon development using trained panelists in Thailand and the U.S.A.: soy sauce. Journal of Sensory Studies, 28, 248-255.

Choma, C., Guinebretière, M. H., Carlin, F., Schmitt, P., Velge, P., Granum, P. E. and Nguyen-Thé, C. (2000). Prevalence, characterization and growth of Bacillus cereus in commercial cooked chilled foods containing vegetables. Journal of Applied Microbiology, 88, 617-625.

Chon, J.-W., Kim, J.-H., Lee, S.-J., Hyeon, J.-Y. and Seo, K.-H. (2012). Toxin proﬁle, antibiotic resistance, and phenotypic and molecular characterization of Bacillus cereus in Sunsik. Food Microbiology, 32, 217-222.

Cilla, I., Martínez, L., Beltrán, J. S. and Roncalés, P. (2006). Dry-cured ham quality and acceptability as affected by the preservation system used for retail sale. Meat Science, 73, 581-589.

Clavel, T., Carlin, F., Lairon, D., Nguyen-Thé, C. and Schmitt, P. (2004). Survival of Bacillus cereus spores and vegetative cells in acid media simulating human stomach. Journal of Applied Microbiology, 97, 214-219.

Coleman, W. H., Zhang, P., Li, Y.-q and Setlow, P. (2010). Mechanism of killing of spores of Bacillus cereus and Bacillus megaterium by wet heat. Letters in Applied Microbiology, 50, 507-514.

Cronin, U. P. and Wilkinson, M. G. (2009). The growth, physiology and toxigenic potential of Bacillus cereus in cooked rice during storage temperature abuse. Food Control, 20, 822-828.

Curiel, J. A., Ruiz-Capillas, C., Rivas, B., Carrascosa, A. V., Jiménez-Colmenero, F. and Muñoz, R. (2011). Production of biogenic amines by lactic acid bacteria and enterobacteria isolated from fresh pork sausages packaged in different atmospheres and kept under refrigeration. Meat Science, 88, 368-373.

Daelman, J., Vermeulen, A., Willemyns, T., Ongenaert, R., Jacxsens, L., Uyttendaele, M. and Devlieghere, F. (2013). Growth/no growth models for heat-treated psychrotrophic Bacillus cereus spores under cold storage. International Journal of Food Microbiology, 161, 7-15.

Dahlsten, E., Korkeala, H., Somervuo, P. and Lindström, M. (2008). PCR assay for differentiating between Group I (proteolytic) and Group II (nonproteolytic) strains of Clostridium botulinum. International Journal of Food Microbiology, 124, 108-111.

43 CHAPTER 2 ______

Daskalov, H. (2006). The importance of Aeromonas hydrophila in food safety. Food Control, 17, 474-483.

De Santos, F., Rojas, M., Lockhorn, G. and Brewer, M. S. (2007). Eﬀect of carbon monoxide in modiﬁed atmosphere packaging, storage time and endpoint cooking temperature on the internal color of enhanced pork. Meat Science, 77, 520-527.

Del Torre, M., Corte, D. M. and Stecchini, M. L. (2001). Prevalence and behaviour of Bacillus cereus in a REPFED of Italian origin. International Journal of Food Microbiology, 63, 199-207.

Del Torre, M., Stecchini, M. L., Braconnier, A. and Peck, M. W. (2004). Prevalence of Clostridium species and behaviour of Clostridium botulinum in gnocchi, a REPFED of Italian origin. International Journal of Food Microbiology, 96, 115-131.

Delbrassinne, L., Andjelkovic, M., Rajkovic, A., Bottledoorn, N., Mahillon, J. and Van Loco, J. (2011). Follow-up of the Bacillus cereus emetic toxin production in penne pasta under household conditions using liquid chromatography coupled with mass spectrometry. Food Microbiology, 28, 1105-1109.

Devlieghere, F., Debevere, J. and Van Impe, J. (1998). Concentration of carbon dioxide in the water-phase as a parameter to model the effect of a modified atmosphere on microorganisms. International Journal of Food Microbiology, 43, 105-113.

Devlieghere, F., Van Belle, B. and Debevere, J. (1998). Shelf life of modiﬁed atmosphere packed cooked meat products: a predictive model. International Journal of Food Microbiology, 46, 57-70.

Diez, A. M., Björkroth, J., Jaime, I. and Rovira, J. (2009a). Microbial, sensory and volatile changes during the anaerobic cold storage of morcilla de Burgos previously inoculated with Weissella viridescens and Leuconostoc mesenteroides. International Journal of Food Microbiology, 131, 168-177.

Diez, A. M., Jaime, I. and Rovira, J. (2009). The inﬂuence of different preservation methods on spoilage bacteria populations inoculated in morcilla de Burgos during anaerobic cold storage. International Journal of Food Microbiology, 132, 91-99.

Diez, A. M., Santos, E. M., Jaime, I. and Rovira, J. (2008a). Application of organic acid salts and high-pressure treatments to improve the preservation of blood sausage. Food Microbiology, 25, 154–161.

Diez, A. M., Santos, E. M., Jaime, I. and Rovira, J. (2009b). Effectiveness of combined preservation methods to extend the shelf life of Morcilla de Burgos. Meat Science, 81, 171-177.

44 CHAPTER 2 ______

Diez, A. M., Urso, R., Rantsiou, K., Jaime, I., Rovira, J. and Cocolin, L. (2008b). Spoilage of blood sausages “morcilla de Burgos” treated with high hydrostatic pressure. International Journal of Food Microbiology, 123, 246-253.

Dong, Q.-L. (2012). Exposure assessment of Bacillus cereus in Chineese-style cooked rice. Journal of Food Process Engineering, 36, 1-8.

Doulgeraki, A. I., Ercolin, D., Villani, F. and Nychas, G.-H. E. (2012). Spoilage microbiota associated to the storage of raw meat in different conditions. International Journal of Food Microbiology, 157, 130-141.

Doyle, M. E., Mazzota, A. S., Wang, T., Wiseman, D. W. and Scott, V. N. (2001). Heat resistance of Listeria monocytogenes. Journal of Food Protection, 64, 410-429.

Drobniewski, F. A. (1993) Bacillus cereus and related species. Clinical Microbiology Reviews, 6, 324-338.

Dufrenne, J., Bijwaard, M., Te Giffel, M., Beumer, R. and Notermans, S. (1995). Characteristics of some psychrotrophic Bacillus cereus isolates. International Journal of Food Microbiology, 27, 175-183.

Duizer, L. M., Robertson, T. and Han, J. (2009). Requirements for packaging from an ageing consumer’s perspective. Packaging Technology and Science, 22, 187-197.

Dykes, G. A., Cloete, T. E. and von Holy, A. (1991). Quantification of microbial populations associated with the manufacture of vacuum packaged, smoked Vienna sausages. International Journal of Food Microbiology, 13, 239-248.

ECFF (European Chilled Food Federation) (2006). Recommendations for the Production of Pre-packaged Chilled Food. Retrieved from http://www.ecff.net/images/ECFF_Recommendations_2nd_ed_18_12_06.pdf Accessed on 03-04-2013.

EFSA (European Food Safety Authority) (2005a). Opinion of the Scientific Panel on Biological Hazards on the Request from the Commission Related to Clostridium spp. in foodstuffs. The EFSA Journal, 199, 1-65.

EFSA (European Food Safety Authority) (2005b). Opinion of the Scientific Panel on Biological Hazards on Bacillus cereus and other Bacillus spp. in foodstuffs. The EFSA Journal, 175, 1-48.

Eilert, S. J. (2005). New packaging technologies for the 21st century. Review article. Meat Science, 71, 122-127.

45 CHAPTER 2 ______

Etaio, I., Gil, P. F., Ojeda, M., Albisu, M., Salmerón, J. and Elortondo F. J. P. (2013). Evaluation of sensory quality of calf chops: A new methodological approach. Meat Science, 94, 105-114.

European Regulation No 510 (2006). Council Regulation of 20th March 2006. Official Journal of the European Union.2006, L93, 12-25.

European Regulation No 1151 (2012). Regulation of the European Parliament and of the Council of 21st November 2012. Official Journal of the European Union.2012, L343, 1-29.

European Regulation No 1169 (2011). European Parliament and Council Regulation of 25th October 2011. Official Journal of the European Union. 2011, L304, 18-63.

European Regulation No 1441 (2007). Commission Regulation of 5th December 2007. Official Journal of the European Union. 2007, L322, 12-29.

European Regulation No 2073 (2005). Commission Regulation of 15th November 2005. Official Journal of the European Union. 2005, L338, 1-26.

Farber, J. M. (1991). Microbiological aspects of modiﬁed-atmosphere packaging technology - A review. Journal of Food Protection, 54, 58-70.

FDA (Food and Drug Administration) (2009). Food Code 2009: Chapter 3 - Food. US Department of Health & Human Services, Food and Drug Administration. Retrieved from http://www.fda.gov/Food/GuidanceRegulation/RetailFoodProtection/FoodCode/uc m186451.htm#part3-5 Accessed on 6-01-2014.

Fekete, T. (2009). Bacillus species (not anthracis). Clinical Microbiology Newsletter, 31, 87-92.

Fernández, A., Ocio, M. J., Fernández, P. S., Rodrigo, M. and Martinez, A. (1999). Application of nonlinear regression analysis to the estimation of kinetic parameters for two enterotoxigenic strains of Bacillus cereus spores. Food Microbiology, 16, 607-613.

Fernández-Fernández, E., Vázquez-Odériz, M. L. and Romero-Rodríguez, M. A. (2002). Sensory characteristics of Galician chorizo sausage packed under vacuum and under modiﬁed atmospheres. Meat Science, 62, 67-71.

Flores, F., Ingram, D. A., Bett, K. L., Toldrá, F. and Spanier, A. M. (1997). Sensory characteristics of Spanish “serrano” dry-cured ham. Journal of Sensory Studies, 12, 169-179.

46 CHAPTER 2 ______

FSA (Food Standards Agency) (2008). Food Standards Agency guidance on the safety and shelf-life of vacuum and modiﬁed atmosphere packed chilled foods with respect to non-proteolytic Clostridium botulinum. Retrieved from http://food.gov.uk/multimedia/pdfs/publication/vacpacguide.pdf Accessed on 6-01- 2014.

FSAI (Food Safety Authority of Ireland) (2006). Industrial processing of heat-chill foods. Guidance Note No. 20. Retrieved from https://www.google.com/url?q=http://www.fsai.ie/WorkArea/DownloadAsset.aspx %3Fid%3D760&sa=U&ei=xGlrU86nFrDMsQSaiYFA&ved=0CAUQFjAA&client =internal-uds-cse&usg=AFQjCNEz6XBOoKlqb61etqX0Fc4arv_bsg Accessed on 12-02-2013.

FSAI (Food Safety Authority of Ireland) (2011). Microbial factsheet series. Issue No.1. Retrieved from http://www.fsai.ie/WorkArea/DownloadAsset.aspx?id=10919 Accessed on 12-02-2013.

Gandhi, M. and Chikindas, M. L. (2007). Listeria: A foodborne pathogen that knows how to survive. International Journal of Food Microbiology, 113, 1-15.

Gariepy, C., Amiot, J., Simard, R. E., Boudreau, A. and Raymond, D. P. (1986). Effect of vacuum-packing and storage in nitrogen and carbon dioxide atmospheres on the quality of fresh rabbit meat. Journal of Food Quality, 9, 289-309.

Gašperlin, L., Žlender, B. and Abram, V. (2001). Colour of beef heated to different temperatures as related to meat ageing. Meat Science, 59, 21-30.

Gibson A. M., Ellis-Brownlee, R. C. L., Cahill , M. E., Szabo, E. A., Fletcher, G. C. and Bremer, P. J. (2000). The effect of 100% CO2 on the growth of non proteolytic Clostridium botulinum at chill temperatures. International Journal of Food Microbiology, 54, 39-48.

Gill, C. O. and Penney, N. (1988). The effect of the initial gas volume to meat weight ratio on the storage life of chilled beef packaged under carbon dioxide. Meat Science, 22, 53-63.

Gill, C. O. and Tan, K. H. (1980). Effect of carbon dioxide on growth of meat spoilage bacteria. Applied and Environmental Microbiology, 39, 317-319.

Giménez, A., Ares, F. and Ares, G. (2012). Sensory shelf-life estimation: A review of current methodological approaches. Food Research International, 49,311-325.

47 CHAPTER 2 ______

Gokoglu, N., Pinar, Y., Uran, H. and Topuz, O. S. (2010). The effect of modified atmosphere packaging on the quality and shelf life of frankfurter type-sausages. Journal of Food Quality, 33, 367-380.

Gómez, M. and Lorenzo, J. M. (2012). Effect of packaging conditions on shelf-life of fresh foal meat. Meat Science, 91, 513-520.

Gonzalez, I., Lopez, M., Martinez, S., Bernardo, A. and Gonzalez, J. (1999). Thermal inactivation of Bacillus cereus spores formed at different temperatures. International Journal of Food Microbiology, 51, 81-84.

Gorraiz, C., Beriain, M. J., Chasco, J. and Iraizoz, M. (2000). Descriptive Analysis of meat from young ruminants in Mediterranean systems. Journal of Sensory Studies, 15, 137-150.

Goul, G. W. (2006). History of science - spores. Journal of Applied Microbiology, 101, 507-513.

Gounadaki, A. S., Skandamis, P. N., Drosinos, E. H. and Nychas, G.-J. E. (2008). Microbial ecology of food contact surfaces and products of small-scale facilities producing traditional sausages. Food Microbiology, 25, 313-323.

Gounina-Allouane, R., Broussolle, V. and Carlin, F. (2008). Inﬂuence of the sporulation temperature on the impact of the nutrients inosine and L-alanine on Bacillus cereus spore germination. Food Microbiology, 25, 202-206.

Graham, A. F., Mason, D. R., Maxwell, F. J. and Peck, M. W. (1997) Effect of pH and NaCl on growth from spores of non-proteolytic Clostridium botulinum at chill temperatures. Letters in Applied Microbiology, 24, 95-100.

Granum, P. E. and Lund, T. (1997). Bacillus cereus and its food poisoning toxins. Mini review. FEMS Microbiology Letters, 157, 223-228.

Grini, J. A., Sørheim, O. and Nissen, H. (1992). The effect of packaging materials and oxygen on the colour stability of sliced bologna. Packaging Technology and Science, 5, 313-320.

48 CHAPTER 2 ______

Guinebretière, M.-H., Thompson, F. L., Sorokin, A., Normand, P., Dawyndt, P., Ehling-Schtultz, M., Svensson, B., Sanchis, V., Nguyen-Thé, C., Heyndrickx, M. and De Vos, P. (2008). Ecological diversiﬁcation in the Bacillus cereus group. Environmental Microbiology, 10, 851-865.

Han, Y., Xu, X., Jiang, Y., Zhou, G., Sun, X. and Xu, B. (2010). Inactivation of food spoilage bacteria by high pressure processing: Evaluation with conventional media and PCR–DGGE analysis. Food Research International, 43, 1719-1724.

Hintlian, C. B. and Hotchkiss, J. H. (1987). Comparative growth of spoilage and pathogenic organisms on modified atmosphere packaged cooked beef. Journal of Food Protection, 50, 218-223.

Holdsworth, S. D. (2004). Optimizing the safety and quality of thermally processed packaged foods. In Improving the thermal processing of foods. Richardson, P. (Ed.), 1-27, CRC Press, Boca Raton, Florida, U.S.A.

Hotchkiss, J. H. and Langston, S. W. (1995). MAP of cooked meat and poultry products. In Principles of modified-atmosphere and sous vide product packaging. Farber, J. M. and Dodds, K. L. (Eds.), 137-152, Technomic Pub. Co. Inc., Lancester, U.S.A.

Hough, G. and Garitta, L. (2012). Methodology for sensory shelf-life estimation: a review. Journal of Sensory Studies, 27, 137-147.

Hough, G., Subramaniam, P., Narain, C. and Beeren, C. (2013). Collecting samples from the shelf: does this contribute to shelf-life knowledge? Journal of Sensory Studies, 28, 47-56.

Huang, L. (2007). Computer simulation of heat transfer during in-package pasteurization of beef frankfurters by hot water immersion. Journal of Food Engineering, 80, 839-849.

Hudson, J. A., Mott, S., J. and Penny, N. (1994). Growth of Listeria monocytogenes, Aeromonas hydrophila and Yersinia enterocolitica on vacuum and saturated carbon dioxide controlled atmosphere packaged sliced roast beef. Journal of Food Protection, 57, 204-208.

ICMSF (International Commission for the Microbiological Specifications of Foods) (2002). Microorganisms in foods 7: microbiological testing in food safety management. Kluwer Academic/Plenum Publishers, New York, U.S.A.

ICMSF (International Commission on Microbiological Specifications for Foods) (2011). Microorganisms in foods 8: use of data for assessing process control and product acceptance. Springer Science+Business Media, LLC, New York, U.S.A.

49 CHAPTER 2 ______

IFT (Institute of Food Technologists) (2013). FAQ: Application of science to food safety management. Retrieved from http://www.ift.org/knowledge-center/read-ift- publications/science-reports/expert-reports/emerging-microbiological-food-safety- issues/faq-application-of-science.aspx. Accessed on 18-02-2013.

Igene, J. O., King, J. A., Pearson, A. M. and Gray, J. I. (1979). Influence of heme pigments, nitrite, and non-heme iron on development of warmed-over flavour (WOF) in cooked meat. Journal of Agricultural and Food Chemistry, 27, 838-841.

Incze, K., Körmendy, L., Körmendy, I. and Zsarnóczay, G. (1999). Considerations of critical microorganisms and indicator enzymes in connection with the pasteurization of meat products. Meat Science, 51, 115-121.

International Organization for Standardization (ISO) 8586-1 (2001). Sensory analysis. General guidance for the selection, training and monitoring of assessors. Part 1: Selected assessors, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 11035 (1994). Sensory analysis. Identification and selection of descriptors for establishing a sensory profile by a multidimensional approach, International Organization for Standardization, Geneva, Switzerland.

Jackson, T. C., Acuff, G. R., Vanderzant, C., Sharp, T. R. and Savell, J. W. (1992). Identification and evaluation of volatile compounds of vacuum and modified atmosphere packaged beef strip loins. Meat Science, 31, 175-190.

Jakobsen, M. and Bertelsen, G. (2002). The use of CO2 in packaging of fresh red meats and its effect on chemical quality changes in the meat: A review. Journal of Muscle Foods, 13, 143-168.

Jiménez, M. J., Canet, W. and Alvarez, M. D. (2013). Sensory description of potato puree enriched with individual functional ingredients and their blends. Journal of Texture Studies, 44, 301-316.

Johnson, K. M. (1984). Bacillus cereus foodborne illness-an update. Journal of Food Protection, 47, 145-153.

Johnson, P. B. and Civille, G. V. (1986). A standardized lexicon of meat wof descriptors. Journal of Sensory Studies, 1, 99-104.

Juneja, V. K. (2000). Thermal inactivation of Salmonella serotypes in red meat as affected by fat content. Quantitative Microbiology, 2, 189-225.

50 CHAPTER 2 ______

Juneja, V. K. (2003). A comparative heat inactivation study of indigenous microflora in beef with that of Listeria monocytogenes, Salmonella serotypes and Escherichia coli O157:H7. Letters in Applied Microbiology, 37, 292-298.

Juneja, V. K. and Eblen, B. S. (1995). Influence of sodium chloride on thermal inactivation and recovery of non proteolytic Clostridium botulinum type B strain KAP B5 spores. Journal of Food Protection, 58, 813-816.

Juneja, V. K., Eblen, B. S., Marmer, B. S., Williams, A. C., Palumbo, S. A. and Miller, A. J. (1995). Thermal resistance of non proteolytic type B and E Clostridium botulinum spores in phosphate buffer and turkey slurry. Journal of Food Protection, 58, 758-763.

Juneja, V. K., Eblen, B. S. and Ransom, G. M. (2001). Thermal Inactivation of Salmonella spp. in chicken broth, beef, pork, turkey, and chicken: determination of D- and z-values. Journal of Food Science, 66, 146-152.

Juneja, V. K., Klein, P. G. and Marmer, B. S. (1998). Heat shock and thermotolerance of Escherichia coli O157:H7 in a model beef gravy system and ground beef. Journal Applied Microbiology, 84, 677-684.

Juneja, V. K. and Marks, H. M. (2003). Characterizing asymptotic D-values of Salmonella spp. subjected to different heating rates in sous-vide cooked beef. Innovative Food Science and Emerging Technologies, 4, 395-402.

Juneja, V. K., Marks, H. M. and Huang, L. (2003). Growth and heat resistance kinetic variation among various isolates of Salmonella and its application to risk assessment. Risk Analysis, 23, 199-213.

Juneja, V. K. and Marmer, B. S. (1996). Growth of Clostridium perfringens from spore inocula in sous-vide turkey products. International Journal of Food Microbiology, 32, 115-123.

Juneja, V. K. and Marmer, B. S. (1999). Lethality of heat to Escherichia coli O157:H7: D- and z-value determinations in turkey, lamb and pork. Food Research International, 32, 23-28.

Juneja, V. K. and Snyder, O. P. (2007). Sous vide and cook-chill processing of foods: concept development and microbiological safety. In Advances in thermal and non- thermal food preservation. Tewari, G. and Juneja, V. K. (Eds.), 145-163, Blackwell Publishing, Iowa, U.S.A.

Juneja, V. K., Snyder, O. P. and Cygnarowicz-Provost, M. (1994). Influence of cooling rate on outgrowth of Clostridium perfringens spores in cooked ground beef. Journal of Food Protection, 57, 1063-1067.

51 CHAPTER 2 ______

Juneja, V. K., Whiting, R. C., Marks, H. M. and Snyder, O. P. (1999). Predictive model for growth of Clostridium perfringens at temperatures applicable to cooling of cooked meat. Food Microbiology, 16, 335-349.

Kalpana, S. and Elsa, M. (1996). Effect of heat shock on the thermotolerance and protein composition of Yersinia enterocolitica in brain heart infusion broth and ground pork. Journal of Food Protection, 59, 360-364.

Karabagias, I., Badeka, A. and Kontominas, M. G. (2011). Shelf life extension of lamb meat using thyme or oregano essential oils and modiﬁed atmosphere packaging. Meat Science, 88, 109-116.

Kemp, J. D., Langlois, B. E., Akers, K., Means, W. J. and Aaron, D. K. (1988). Eﬀect of storage temperature and time on the quality of vacuum packaged dry-cured ham slices. Journal of Food Science, 53, 402-406.

Kerry, J. P., O’Grady, M. N. and Hogan, S. A. (2006). Past, current and potential utilisation of active and intelligent packaging systems for meat and muscle-based products: A review. Meat Science, 74, 113-130.

Killinger, K. M., Hunt, M. C., Campbell, R. E. and Kropf, D. H. (2000). Factors affecting premature browning during cooking of store-purchased ground beef. Journal of Food Science, 65, 585-587.

Kim, H., Kim, H., Bang, J., Kim, Y., Beuchat, L. R. and Ryu, J.-H. (2012). Reduction of Bacillus cereus spores in sikhye, a traditional Korean rice beverage, by modiﬁed tyndallization processes with and without carbon dioxide injection. Letters in Applied Microbiology, 55, 218-223.

Koort, J., Coenye, T., Santos, E. M., Molinero, C., Jaime, I., Rovira, J., Vandamme, P. and Björkroth, J. (2006). Diversity of Weissella viridescens strains associated with “Morcilla de Burgos”. International Journal of Food Microbiology, 109, 164-168.

Korkeala, H. J. and Björkroth, K. J. (1997). Microbiological spoilage and contamination of vacuum-packed cooked sausages. Journal of Food Protection, 60, 724-731.

Korkeala, H., Lindroth, S., Ahvenainen, R. and Alanko, T. (1987). Interrelationship between microbial numbers and other parameters in the spoilage of vacuum-packed cooked ring sausages. International Journal of Food Microbiology, 5, 311-341.

Korkeala, H., Lindroth, S., Suihko, M., Kuhmonen, A. and Penttilä, P.-L. (1985). Microbiological and sensory quality changes in blood pancakes and cooked ring sausage during storage. International Journal of Food Microbiology, 2, 279-292.

52 CHAPTER 2 ______

Korkeala, H. and Mäkelä, P. (1989). Characterization of lactic acid bacteria isolated from vacuum-packed cooked ring sausages. International Journal of Food Microbiology, 9, 33-43.

Korkeala, H., Suortti, T. and Mäkelä, P. (1988). Ropy slime formation in vacuum- packed cooked meat products caused by homofermentative lactobacilli and Leuconostoc species. International Journal of Food Microbiology, 7, 339-347.

Kotzekidou, P. and Bloukas, J. G. (1996). Effect of protective cultures and packaging ﬁlm permeability on shelf-life of sliced vacuum-packed cooked ham. Meat Science, 42, 333-345.

Kramer, J. M. and Gilbert, R. J. (1989). Bacillus cereus and other Bacillus species. In Food-borne Bacterial Pathogens. Doyle, J. P. (Ed.), 21-77, Marcel Dekker, New York, U.S.A.

Lamjed, S., Michèle, M., Fadia, N., Blaise, O., Claude, L. and Linda, S. (2006). Heat Inactivation of Listeria monocytogenes and Salmonella enterica serovar typhi in a typical bologna matrix during an industrial cooking-cooling cycle. Journal of Food Protection, 69, 3025-3030.

Lange, J. and Wyser, Y. (2003). Recent innovations in barrier technologies for plastic packaging - a review. Packaging Technology Science, 16, 149-158.

Larrauri, M., Barrionuevo, M. G., Riveros, C., Mestrallet, M. G., Zunino, M. P., Zygadlo, J. A., Grosso, N. R. and Nepote, V. (2013). Effect of peanut skin extract on chemical stability and sensory properties of salami during storage. Journal of the Science of Food and Agriculture, 93, 1751-1757.

Laursen, B. G., Byrne, D. V., Kirkegaard, J. B. and Leisner, J. J. (2009). Lactic acid bacteria associated with a heat-processed pork product and sources of variation affecting chemical indices of spoilage and sensory characteristics. Journal of Applied Microbiology, 106, 543-553.

Lawlor, K. A., Pierson, M. D., Hackney, C. R., Claus, J. R. and Marcy, J. E. (2000). Non proteolytic Clostridium botulinum toxigenesis in cooked turkey stored under modified atmospheres. Journal of Food Protection, 63, 1511-1516.

Lebert, I., Leroy, S., Giammarinaro, P., Lebert, A., Chacornac, J. P., Bover-Cid, S., Vidal-Carou, M. C. and Talon, R. (2007). Diversity of microorganisms in the environment and dry fermented sausages of small traditional French processing units. Meat Science, 76, 112-122.

Lee, K. T. (2010). Quality and safety aspects of meat products as affected by various physical manipulations of packaging materials. Review article. Meat Science, 86, 138-150.

53 CHAPTER 2 ______

Lee, D. S., Yam, K. L. and Piergiovanni, L. (2008). Food packaging science and technology. CRC Press, Boca Raton, Florida, U.S.A.

Lien, R., Hunt, M. C., Anderson, S., Kropf, D. H., Loughin, T. M., Dikeman, M. E. and Velazco, J. (2002a). Effects of endpoint temperature on the internal colour of pork loin chops of different quality. Journal of Food Science, 67, 1007-1010.

Lien, R., Hunt, M. C., Anderson, S., Kropf, D. H., Loughin, T. M., Dikeman, M. E. and Velazco, J. (2002b). Effects of endpoint temperature on the internal colour of pork patties of different myoglobin form, initial cooking state, and quality. Journal of Food Science, 67, 1011-1015.

Lin, S., Schraft, H., Odumeru, J. A. and Grifﬁths, M. W. (1998). Identiﬁcation of contamination sources of Bacillus cereus in pasteurized milk. International Journal of Food Microbiology, 43, 159-171.

Lindström, M., Kiviniemi, K. and Korkeala, H. (2006). Hazard and control of group II (non-proteolytic) Clostridium botulinum in modern food processing. International Journal of Food Microbiology, 108, 92-104.

Liu, F., Guo, Y. and Li, Y. (2006). Interactions of microorganisms during natural spoilage of pork at 5 ºC. Journal of Food Engineering, 72, 24-29.

Logan, N. A. (2011). Bacillus and relatives in foodborne illness. Review article. Journal of Applied Microbiology, 112, 417-429.

Lund, T. and Granum, P. E. (1996). Characterization of a non-haemolytic enterotoxin complex from Bacillus cereus isolated after a foodborne outbreak. FEMS Microbiology Letters, 141, 151-156.

Lund, B. M. and Peck, M. W. (1994). Heat resistance and recovery of non-proteolytic Clostridium botulinum in relation to refrigerated processed foods with an extended shelf-life. Journal of Applied Bacteriology, 76, 115-128.

Lynt, R. K., Kautter, D. A. and Solomon, H. M. (1982). Differences and similarities among proteolytic and non proteolytic strains of Clostridium botulinum types A, B, E, and F: a review. Journal of Food Protection, 45, 466-474.

Lyon, B. G. (1987). Development of chicken flavour descriptive attribute terms aided by multivariate statistical procedures. Journal of Sensory Studies, 2, 55-67.

Mäkelä, P. and Korkeala, H. (1987). Lactobacillus contamination of cooked ring sausages at sausage processing plants. International Journal of Food Microbiology, 5, 323-330.

54 CHAPTER 2 ______

Mäkelä, P., Schillinger, U., Korkeala, H. and Holzapfel, W. (1992). Classiﬁcation of ropy slime-producing lactic acid bacteria based on DNA–DNA homology and identiﬁcation of Lactobacilllus sake and Leuconostoc amelibiosum as dominant spoilage organisms in meat products. International Journal of Food Microbiology, 16, 167-172.

Mancini, A., Ramanathan, R., Suman, S. P., Konda, M. K. R., Joseph, P., Dady, G. A., Naveena, B. M. and López-López, I. (2010). Effects of lactate and modified atmosphere packaging on premature browning in cooked ground beef patties. Meat Science, 85, 339-346.

Márquez-González, M., Cabrera-Díaz, E, Hardin, M. D., Harris, K. B., Lucia, L. M. and Castillo, A. (2012). Survival and germination of Clostridium perfringens spores during heating and cooling of ground pork. Journal of Food Protection, 75, 682-689.

Marth, E. H. (1998). Extended shelf life refrigerated foods: microbiological quality and safety. Food Technology, 52, 57-62.

Matamoros, S., André, S., Hue, I., Prévost, H. and Pilet, M. F. (2010). Identiﬁcation of lactic acid bacteria involved in the spoilage of pasteurized ‘‘foie gras” products. Meat Science, 85, 467-471.

Maughan, C., Tansawat, R., Cornforth, D., Ward, R. and Martini, S. (2012). Development of a beef flavour lexicon and its application to compare the flavour profile and consumer acceptance of rib steaks from grass- or grain-fed cattle. Meat Science, 90, 116-121.

Mckenna, B. M., Lyng, J., Brunton, N. and Shirsat, N. (2006). Advances in radio frequency and ohmic heating of meats. Journal of Food Engineering, 77, 215-229.

McMillin, K. W. (2008). Where is MAP going? A review and future potential of modified atmosphere packaging for meat. Meat Science, 80, 43-65.

Meer, R. R., Baker, J. W., Bodyfelt, F. W. and Griffiths, M. W. (1991). Psychrotrophic Bacillus spp. in fluid milk products. A review. Journal of Food Protection, 54, 969-979.

Membré, J.-M., Yu, K.-K. D. and Blackburn, C. W. (2008). Use of sensitivity analysis to aid interpretation of a probabilistic Bacillus cereus spore lag time model applied to heat-treated chilled foods (REPFEDs). International Journal of Food Microbiology, 128, 28-33.

55 CHAPTER 2 ______

Molinero, C., Martínez, B., Rubio, B., Rovira, J. and Jaime, I. (2008). The eﬀects of extended curing on the microbiological, physicochemical and sensorial characteristics of Cecina de León. Meat Science, 80, 370-379.

Molnár, P. J. (1995). A model for overall description of food quality. Food Quality and Preference, 6, 185-190.

Morais, E. C., Cruz, A. G., Faria, J. A. F. and Bolini, H. M. A. (2014). Prebiotic gluten-free bread: sensory profiling and drivers of liking. LWT-Food Science and Technology, 55, 248-254.

Moskowitz, H. R. (1983). Product testing and sensory evaluation of foods: marketing and R&D approaches. Food and Nutrition Press, Westport, Connecticut, U.S.A.

Moussaoui, K. A. and Varela, P. (2010). Exploring consumer product proﬁling techniques and their linkage to a quantitative descriptive analysis. Food Quality and Preference, 21, 1088-1099.

Mullan, M. and McDowell, D. (2003). Modified atmosphere packaging. In Food packaging technology. Coles, R., McDowell, D. and Kirwan, M. J. (Eds.), 303-338, CRC Press, Boca Raton, Florida, U.S.A.

Muñoz, A.M., Chambers IV, E. and Hummer, S. (1996). A multifaceted category research study: how to understand a product category and its consumer responses. Journal of Sensory Studies, 11, 261-294.

Muñoz, A. M. and Civille, G. V. (1998). Universal, product and attribute speciﬁc scaling and the development of common lexicons in descriptive analysis. Journal of Sensory Studies, 13, 57-67.

Murphy, R. Y., Beard, B. L., Martin, E. M., Duncan, L. K. and Marcy, J. A. (2004). Comparative study of thermal inactivation of Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes in ground pork. Journal of Food Science, 69, 97-101.

NACMCF (National Advisory Committee on Microbiological Criteria for Foods), (2006). Requisite scientiﬁc parameters for establishing the equivalence of alternative methods of pasteurization. Journal of Food Protection, 69, 1190-1216.

56 CHAPTER 2 ______

NACMCF (National Advisory Committee on Microbiological Criteria for Foods), (2010). Parameters for determining inoculated pack/challenge study protocols. Journal of Food Protection, 73, 140-202.

Nauta, M. J., Litman, S., Barker, G. C. and Carlin, F. (2003). A retail and consumer phase model for exposure assessment of Bacillus cereus. International Journal of Food Microbiology, 83, 205-218.

Newell, D. G., Koopmans, M., Verhoef, L., Duizer, E., Aidara-Kane, A., Sprong, H., Opsteegh, M., Langelaar, M., Threfall, J., Scheutz, F., van der Giessen, J. and Kruse, H. (2010). Food-borne diseases - The challenges of 20 years ago still persist while new ones continue to emerge. International Journal of Food Microbiology, 139, S3-S15.

Nissen, L. R., Derek V., Byrne, D. V., Bertelsen, G. and Skibsted, L. H. (2004). The antioxidative activity of plant extracts in cooked pork patties as evaluated by descriptive sensory proﬁling and chemical analysis. Meat Science, 68, 485-495.

Nørrung, B. and Buncic, S. (2008). Microbial safety of meat in the European Union. Meat Science, 78, 14-24.

Notermans, S. and Batt, C. A. (1998). A risk assessment approach for food-borne Bacillus cereus and its toxins. Journal of Applied Microbiology, 84, 51S-61S.

Notermans S. and in't Veld, P. (1994). Microbiological challenge testing for ensuring safety of food products. International Journal of Food Microbiology, 24, 33-39.

Nychas, G-J. E., Skandamis, P. N., Tassou, C. C. and Koutsoumanis, K. P. (2008). Meat spoilage during distribution. Meat Science, 78, 77-89.

Oh, M. H., Ham, J. S. and Cox, J. M. (2012). Diversity and toxigenicity among members of the Bacillus cereus group. International Journal of Food Microbiology, 152, 1-8.

Osaili, T. M., Griffis, C. L., Martin, E. M., Beard, B. L., Keener, A. E. and Marcy, J. A. (2007). Thermal inactivation of Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes in breaded pork patties. Journal of Food Science, 72, 56-61.

Oteiza, J. M., Giannuzzi, L. and Califano, A. N. (2003). Thermal inactivation of Escherichia coli O157:H7 and Escherichia coli isolated from morcilla as affected by composition of the product. Food Research International, 36, 703-712.

57 CHAPTER 2 ______

Ouoba, L. I. I., Thorsen, L. and Varnam A. H. (2008). Enterotoxins and emetic toxins production by Bacillus cereus and other species of Bacillus isolated from Soumbala and Bikalga, African alkaline fermented food condiments. International Journal of Food Microbiology, 124, 224-230.

Patarata, L., Judas, I., Silva, J. A., Esteves, A. and Martins, C. (2008). A comparison of the physicochemical and sensory characteristics of alheira samples from diﬀerent- sized producers. Meat Science, 79, 131-138.

Peck, M. W. (2006). Clostridium botulinum and the safety of minimally heated, chilled foods: an emerging issue? Journal of Applied Microbiology, 101, 556-570.

Peck, M. W., Goodburn, K. E., Betts, R. P. and Stringer, S. C. (2008). Assessment of the potential for growth and neurotoxin formation by non-proteolytic Clostridium botulinum in short shelf-life commercial foods designed to be stored chilled. Trends in Food Science & Technology, 19, 207-216.

Peck, M. W. and Stringer, S. C. (2005). The safety of pasteurised in-pack chilled meat products with respect to the foodborne botulism hazard. Meat Science, 70, 461-475.

Penney, N., Hagyard, C. J. and Bell, R. G. (1993). Extension of shelf life of chilled sliced roast beef by carbon dioxide packaging. International Journal of Food Science and Technology, 28, 181-191.

Pérez-Cacho, M. P. R., Galán-Soldevilla, H., Crespo, F. L. and Recio, G. M. (2005). Determination of the sensory attributes of a Spanish dry-cured sausage. Meat Science, 71, 620-633.

Peterson, M. E., Pelroy, G. A., Poysky, F. T., Paranjpye, R. N., Dong, F. M., Pigott, G. M. and Eklund, M. E. (1997). Heat-pasteurization process for inactivation of nonproteolytic types of Clostridium botulinum in picked Dungeness crabmeat. Journal of Food Protection, 60, 928-934.

Planchon, S., Dargaignaratz, C., Levy, C., Ginies, C., Broussolle, V. and Carlin, F. (2011). Spores of Bacillus cereus strain KBAB4 produced at 10 ºC and 30 ºC display variations in their properties. Food Microbiology, 28, 291-297.

58 CHAPTER 2 ______

Portuguese Law nº 207 (2008). De 23 de outubro. Diário da República n.º 206. I Série, 7502-7512.

Rajalakshmi, D., Dhanaraj, S., Chand, N. and Govindarajan, V. S. (1987). Descriptive quality analysis of mutton. Journal of Sensory Studies, 2, 93-118.

Rajkovic, A., Uyttendaele, M., Ombregt, S.-A., Jaaskelainen, E., Salkinoja-Salonen, M. and Debevere, J. (2006). Influence of type of food on the kinetics and overall production of Bacillus cereus emetic toxin. Journal of Food Protection, 69, 847-852.

Rajkovic, A., Uyttendaele, M., Vermeulen, A., Andjelkovic, M., Fitz-James, I., in‘t Veld, P., Denon, Q., Vérhe, R. and Debevere, J. (2008). Heat resistance of Bacillus cereus emetic toxin, cereulide. Letters in Applied Microbiology, 46, 536-541.

Ramaswamy, H. and Marcotte, M. (2006). Food processing: principles and applications. CRC Press, Boca Raton, Florida, U.S.A.

Rason, J., Martin, J.-F., Dufour, E. and Lebecque, A. (2007). Diversity of the sensory characteristics of traditional dry sausages from the centre of France: Relation with regional manufacturing practice. Food Quality and Preference, 18, 517-530.

Reddy, N. R., Roman, M. G., Villanueva, M., Solomon, H. M., Kautter, D. A. and Rhodehamel, E. J. (1997). Shelf life and Clostridium botulinum toxin development during storage of modified atmosphere-packaged fresh catfish fillets. Journal of Food Science, 62, 878-884.

Robertson, G. L. (1993). Food packaging - principles and practice. Marcel Dekker, Inc., New York, U.S.A.

Rødbotten, M., Kubberød, E., Lea, P. and Ueland, Ø. (2004). A sensory map of the meat universe. Sensory proﬁle of meat from 15 species. Meat Science, 68, 137-144.

Rodríguez-Pérez, M. R., Zurera-Cosano, G., García-Gimeno, R. M., Barco-Alcala, E. and Castillejo-Rodríguez, A. M. (2003). Sensory and microbiological quality evaluation of vacuum-packaged sliced cooked chicken breast. Shelf-life estimation. Journal of Food Quality, 26, 105-122.

Rubio, B., Martínez, B., García-Cachán, M. D., Rovira, J. and Jaime, I. (2008). Effect of the packaging method and the storage time on lipid oxidation and colour stability on dry fermented sausage salchichón manufactured with raw material with a high level of mono and polyunsaturated fatty acids. Meat Science, 80, 1182-1187.

59 CHAPTER 2 ______

Rubio, B., Martínez, B., Sánchez, M. F., García- Cachán, M. D., Rovira, J. and Jaime, I. (2007b). Study of the shelf life of a dry fermented sausage “salchichon” made from raw material enriched in monounsaturated and polyunsaturated fatty acids and stored under modified atmospheres. Meat Science, 76, 128-137.

Ruiz-Capillas, C. and Jiménez-Colmenero, F. (2010). The effect of an argon- containing packaging atmosphere on the quality of fresh pork sausages kept at 1 °C. Food Control, 21, 1331-1337.

Samapundo, S., Everaert, H., Wandutu, J. N., Rajkovic, A., Uyttendaele, M. and Devlieghere, F. (2011). The inﬂuence of headspace and dissolved oxygen level on growth and haemolytic BL enterotoxin production of a psychrotolerant Bacillus weihenstephanensis isolate on potato based ready-to-eat food products. Food Microbiology, 28, 298-304.

Samelis, J. (2006). Managing microbial spoilage in the meat industry. In Food spoilage organisms. Blackburn, C. D. W. (Ed.), 213-286, Woodhead Publishing Limited, Cambridge, U.K.

Samelis, J., Metaxopoulos, J., Vlassi, M. and Pappa, A. (1998). Stability and safety of traditional Greek salami – A microbiological ecology study. International Journal of Food Microbiology, 44, 69-82.

Santos, E. M., Diez, A. M., González-Fernández, C., Jaime, I. and Rovira, J. (2005a). Microbiological and sensory changes in “Morcilla de Burgos” preserved in air, vacuum and modified atmosphere packaging. Meat Science, 71, 249-255.

60 CHAPTER 2 ______

Santos, E. M., Jaime, I., Rovira, J., Lyhs, U., Korkeala, H. and Björkroth, J. (2005b). Characterization and identification of lactic acid bacteria in ‘‘morcilla de Burgos’’. International Journal of Food Microbiology, 97, 285-296.

Scott, V. N. and Bernard, D. T. (1982). Heat resistance of spores from non-proteolytic type B Clostridium botulinum. Journal of Food Protection, 45, 909-912.

Scott, V. N. and Bernard, D. T. (1985). The effect of lysozyme on the apparent heat resistance of non-proteolytic type B Clostridium botulinum. Journal of Food Safety, 7, 145-154.

Selby, T. L., Berzins, A., Gerrard, D. E., Corvalan, C. M., Grant, A. L. and Linton, R. H. (2006). Microbial heat resistance of Listeria monocytogenes and the impact on ready- to-eat meat quality after post-packaging pasteurization. Meat Science, 74, 425-434.

Sen, A. R., Naveena, B. M., Muthukumar, M. and Vaithiyanathan, S. (2011). Colour, myoglobin denaturation and storage stability of raw and cooked mutton chops at different end point cooking temperature. Journal of Food Science Technology, 48, 1-6.

Senesi, S. Salvetti, S., Celandroni, F. and Ghelardi, E. (2010). Features of Bacillus cereus swarm cells. Research in Microbiology, 161, 743-749.

Shaw, G. H. (2004). The use of data loggers to validate thermal processes. In Improving the thermal processing of foods. Richardson, P. (Ed.), 353-363, CRC Press, Boca Raton, Florida, U.S.A.

Silliker, J. H. (1981). The influence of atmospheres containing elevated levels of CO2 (carbon dioxide) on growth of psychrotrophic organisms in meat and poultry (Pseudomonas and closely related bacteria). In Psychrotrophic microorganisms in spoilage and pathogenicity. Roberts, T. A. (Ed.), 369-375, Academic Press, London, U.K.

Silva, F. V. M. and Gibbs, P. A. (2010). Non-proteolytic Clostridium botulinum spores in low-acid cold distributed foods and design of pasteurization processes. Review. Trends in Food Science and Technology 21, 95-105.

Singh, R. K. (2007). Aseptic processing. In Advances in thermal and non-thermal food preservation. Tewari, G. and Juneja, V. K. (Eds.), 43-61, Blackwell Publishing, Iowa, U.S.A.

61 CHAPTER 2 ______

Sivertsvik, M. and Jensen, J. S. (2005). Solubility and absorption rate of carbon dioxide into non-respiring foods. Part 3: Cooked meat products. Journal of Food Engineering, 70, 499-505.

Smith, J. P., Ramaswamy, H. S. and Simpson, B. K. (1990). Developments in food packaging technology. Part II. Storage aspects. Trends in Food Science & Technology, 1, 111–118.

Smyth, H. E., Sanderson, J. E. and Sultanbawa, Y. (2012). Lexicon for the sensory description of Australian native food plants and ingredients. Journal of Sensory Studies, 27, 471-481.

Sørheim, O., Nissen, H. and Nesbakken, T. (1999). The storage life of beef and pork packaged in an atmosphere with low carbon monoxide and high carbon dioxide. Meat Science, 52, 157-164.

Stolzenbach, S., Leisner, J. J. and Byrne, D. V. (2009). Sensory shelf life determination of a processed meat product ‘rullepølse’ and microbial metabolites as potential indicators. Meat Science, 83, 285-292

Stone, H., Sidel, J., Oliver, S., Woolsey, A. and Singleton, R. C. (1974). Sensory evaluation by quantitative descriptive analysis. Food Technology, 28, 24-34.

Stringer, S. C., Webb, M. D. and Peck, M. W. (2009). Contrasting effects of heat treatment and incubation temperature on germination and outgrowth of individual spores of non proteolytic Clostridium botulinum bacteria. Applied and Environmental Microbiology, 75, 2712-2719.

Summo, C., Caponio, F. and Pasqualone, A. (2006). Effect of vacuum-packaging storage on the quality level of ripened sausages. Meat Science, 74, 249-254.

Te Giffel, M. C., Beumer, R. R., Granum, P. E. and Rombouts, F. M. (1997). Isolation and characterisation of Bacillus cereus from pasteurised milk in household refrigerators in the Netherlands. International Journal of Food Microbiology, 34, 307-318.

62 CHAPTER 2 ______

Te Giffel, M. C., Beumer, R. R., Leijendekkers, S. and Rombouts, F. M. (1996). Incidence of Bacillus cereus and Bacillus subtilis in foods in the Netherlands. Food Microbiology, 13, 53-58.

Tsigarida, E., Hugas, M. and Robinson, T. (2009). The EFSA scientific panel on biological hazards first mandate: May 2003-May 2006. Insight into scientific advice on food hygiene and microbiology. Trends in Food Science & Technology, 20, 587-594.

Tuorila, H. and Monteleone, E. (2009). Sensory food science in the changing society: opportunities, needs, and challenges. Trends in Food Science & Technology, 20, 54-62.

Uyttendaele, M., Rajkovic, A., Benos, G., François, K., Devlieghere, F. and Debevere, J. (2004). Evaluation of a challenge testing protocol to assess the stability of ready- to-eat cooked meat products against growth of Listeria monocytogenes. International Journal of Food Microbiology, 90, 219-236.

Valentin, D., Chollet, S., Lelièvre, M. and Abdi, H. (2012). Quick and dirty but still pretty good: a review of new descriptive methods in food science. International Journal of Food Science and Technology, 47, 1563-1578.

Valero, A., Carrasco, E. and García-Gimeno, R. M. (2012). Principles and methodologies for the determination of shelf-life in foods. In Trends in vital food and control engineering. Eissa, A. H. A. (Ed.), 3-42, InTech, DOI: 10.5772/35353. Available from http://www.intechopen.com/books/trends-in-vital-food-and-control- engineering/principles-and-methodologies-for-the-determination-of-shelf-life-in- foods. Accessed on 18-02-2013.

Valero, M., Hernández-Herrero, L. A., Fernández, P. S. and Salmerón, M. C. (2002). Characterization of Bacillus cereus isolates from fresh vegetables and refrigerated minimally processed foods by biochemical and physiological tests. Food Microbiology, 19, 491-499.

Vandendriessche, F. (2008). Meat products in the past, today and in the future. Meat Science, 78, 104-113. van der Voort, M. and Abee, T. (2012). Sporulation environment of emetic toxin- producing Bacillus cereus strains determines spore size, heat resistance and germination capacity. Journal of Applied Microbiology, 114, 1201-1210.

63 CHAPTER 2 ______van Schothorst, M., Zwietering, M. H., Ross, T., Buchanan, R. L., Cole, M. B. and International Commission on Microbiological Speciﬁcations for Foods (ICMSF) (2009). Relating microbiological criteria to food safety objectives and performance objectives. Food Control, 20, 967-979.

Varela, P. and Ares, G. (2012). Sensory profiling, the blurred line between sensory and consumer science. A review of novel methods for product characterization. Food Research International, 48, 893-908.

Vasilopoulos, C., De Maere, H., De Mey, E., Paelinck, H., De Vuyst, L. and Leroy, F. (2010). Technology-induced selection towards the spoilage microbiota of artisan-type cooked ham packed under modiﬁed atmosphere. Food Microbiology, 27, 77-84.

Vasilopoulos, C., Ravyts, F., De Maere, H., De Mey, E., Paelinck, H., De Vuyst, L. and Leroy, F. (2008). Evaluation of the spoilage lactic acid bacteria in modiﬁed- atmosphere-packaged artisan-type cooked ham using culture-dependent and culture- independent approaches. Journal of Applied Microbiology, 104, 1341-1353.

Vázquez-Araújo, L., Chambers IV, E., Adhikari, K., Hough, G. and Carbonell- Barrachina, A. A. (2013). Influence of various traditional seasonings on beef flavor: United States, Spanish, and Argentinian practices. Meat Science, 93, 61-66.

Vermeiren, L., Devlieghere, F., Vandekinderen, I., Rajtak, U. and Debevere, J. (2006). The sensory acceptability of cooked meat products treated with a protective culture depends on glucose content and buﬀering capacity: A case study with Lactobacillus sakei 10A. Meat Science, 74, 532-545.

Vestergaard, E. M. (2001). Building product confidence with challenge studies. Dairy Food and Environmental Sanitation, 21, 206-209.

Viana, E. S., Gomide, L. A. M. and Vanetti, M. C. D. (2005). Eﬀect of modiﬁed atmospheres on microbiological, color and sensory properties of refrigerated pork. Meat Science, 71, 696-705. von Holy, A., Cloete, T. E. and Holzapfel, W. H. (1991). Quantification and characterization of microbial populations associated with spoiled, vacuum-packed Vienna sausages. Food Microbiology, 8, 95-104.

Walls, I. (2006). Role of quantitative risk assessment and food safety objectives in managing Listeria monocytogenes on ready-to-eat meats. Meat Science, 74, 66-75.

64 CHAPTER 2 ______

Wang, G., Sabja, D. P., Sarker, M. R., Green, C., Setlow, P. and Li, Y.-q. (2012). Effects of wet heat treatment on the germination of individual spores of Clostridium perfringens. Journal of Applied Microbiology, 113, 824-836.

Weerasinghe, S., Williams, J. B., Mukherjee, D., Tidwell, D. K., Chang, S. and Haque, Z. U. (2013). Quality and sensory characteristics of cubed beef steak dipped in edible protective solutions of thermized cheddar whey. Journal of Food Quality, 36, 77-90.

Willardsen, R. R., Busta, F. F. and Allen, C. E. (1979). Growth of Clostridium perfringens in three different beef media and fluid thioglycollate medium at static and constantly rising temperatures. Journal of Food Protection, 42, 144-148.

Zhang, L., Lyng, J. G. and Brunton, N. P. (2004). Eﬀect of radio frequency cooking on the texture, colour and sensory properties of a large diameter comminute meat product. Meat Science, 68, 257-268.

Zhou, G. H., Xu, X. L. and Liu, Y. (2010). Preservation technologies for fresh meat - A review. Meat Science, 86, 119-128.

Zwietering, M. H., Stewart, C. M., Whiting, R. C. and International Commission on Microbiological Speciﬁcations for Foods (ICMSF) (2010). Validation of control measures in a food chain using the FSO concept. Food Control, 21, 1716-1722.

CHAPTER 3

Pereira, J. A., Ferreira-Dias, S., Dionísio, L., Patarata, L. and Matos, T. J. S. Application of unsteady-state heat transfer equations to thermal process of Morcela de Arroz from Monchique region, a Portuguese traditional blood sausage. Submitted for publication in Journal of Food Quality

CHAPTER 3 ______

3. APPLICATION OF UNSTEADY-STATE HEAT TRANSFER EQUATIONS TO THERMAL

PROCESS OF MORCELA DE ARROZ FROM MONCHIQUE REGION, A PORTUGUESE

TRADITIONAL BLOOD SAUSAGE

Pereira, J. A.a,b, Ferreira-Dias, S.b,c, Dionísio, L.d,e, Patarata, L.f and Matos, T. J. S.b,c

aDepartment of Food Engineering, Instituto Superior de Engenharia, University of Algarve, Campus da Penha, Estrada da Penha, 8005-139 Faro, Portugal bCEER, Biosystems Engineering, Tapada da Ajuda, 1349-017 Lisbon, Portugal cInstituto Superior de Agronomia, University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal dDepartment of Biological Sciences and Bioengineering, Faculdade de Ciências e Tecnologia, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal eICAAM - Institute of Mediterranean Agricultural and Environmental Sciences, P.O. box 94, University of Évora, 702-554 Évora, Portugal fCECAV, Centro de Ciência Animal e Veterinária, Universidade de Trás-os-Montes e Alto Douro, Quinta de Prados, 5000-801 Vila Real, Portugal

3.1 ABSTRACT

Morcela de Arroz (MA) is a popular cooked blood and rice sausage produced following empirical procedures in the Monchique region, south of Portugal. Presenting high pH (about 6.5) and water activity (around 0.97), this product is commercialized without preservatives, unpacked and with restricted commercialization period (one week at 0-5ºC). The application of unsteady-state heat transfer equations (USHTE) to predict temperature profile inside the product, during thermal processing, is of great importance for thermal process optimization to obtain safe products. The aim of this study was (i) to assess the adequacy of USHTE using Heisler charts to predict temperature in the center of the product and the time needed to attain the temperature needed and (ii) to determine if the thermal treatment traditionally applied to this cooked blood sausage complies with the recommendations for refrigerated processed food of extended durability. A very good fit of USHTE to the experimental time/temperature data was observed. Thus, the use of Heisler charts showed to be a useful and accurate technique to define time/temperature values in thermal processing

CHAPTER 3 ______

of MA, without the need for in situ monitoring time and temperature. MA cooking process ensures a 6D reduction of spores of psychrotrophic non proteolytic C. botulinum strains. Generally, MA complies with the recommendations for the safe production of vacuum and modified atmosphere packed chilled foods with a shelf life longer than 10 days, with respect to the growth and toxin production by non proteolytic C. botulinum. B. cereus and C. perfringens were not detected in any of the tested samples. The thermal process involving temperatures above 95ºC for about 38 min in the core of MA is capable to destroy these more heat resistant spores.

3.2 INTRODUCTION

Morcela de Arroz (MA) is a popular cooked blood and rice sausage produced in the Monchique region, south of Portugal. It is made with ground pork meats and blood, rice, ground onion, spearmint, ground spices (cumin, allspice, clove and pepper) and salt, all stuffed in natural casings (pork stomach, pork or beef intestine), tied with thread and then boiled in molho de moura, a broth prepared with water, whole onions, whole garlic, salt and laurel leaves. This ready-to-eat (RTE) meat product supports the economy of several industries, due to its uniqueness among the traditional Portuguese meat products. MA is a cooked low acidic food (pH about 6.5) with high water activity (around 0.97). It is traditionally produced without preservatives and commercialized unpacked with a restricted period of commercialization (one week at 0-5 ºC) in a restricted area (Pereira, Ferreira-Dias and Matos, 2008). The use of an appropriate packaging technology, e.g., modified atmosphere packaging (MAP) or vacuum packaging (VP) (Santos et al., 2005) might extend the shelf life of this meat product converting it in a refrigerated processed food of extended durability (REPFED). In the RTE food industry, several problems may occur during heating or cooling periods where heat is transferred under unsteady-state conditions. Perishable RTE meat products have to be subjected to adequate thermal treatment in order to prevent the survival of pathogenic microorganisms and to reduce the risk of both microbial and enzymatic degradation during storage. Effective thermal treatment of low acid (pH higher than 4.6) cooked RTE meat products is a critical control point to ensure microbiological safety (Juneja, 2003). The

CHAPTER 3 ______

knowledge of temperature profile on real thermal process conditions of cooked blood sausages is crucial to guarantee product safety and to optimize thermal process. To ensure a microbiologically safe product, the specific thermal process must be validated, which comprises the determination of the coldest spot in the product and data to demonstrate that the required critical binomial time/temperature parameters are met at this point, according to the thermal inactivation kinetics of the specific food most heat resistant pathogen of concern (NACMCF, 2006). At present, two heat treatment recommendations, considered as safe harbours and based on thermobacteriology data and evidence over the last 2 decades, are used for the safe production of low acid REPFEDs. The first one is based on a minimal thermal treatment of 2 min/70 ºC allowing a 6D reduction of Listeria monocytogenes (ICMSF, 2002), which is considered the most heat resistant facultative psychrotrophic vegetative pathogen of concern in RTE meat products. D-value is the decimal reduction time required to reduce the microbial population by 1 log cycle at a constant temperature. The second recommendation comprises a minimal thermal treatment of 10 min/90 ºC or equivalent lethality time/temperature combinations (ACMSF, 1992) achieved at the slowest heating point in the product to ensure a 6D reduction of the number of spores of psychrotrophic non proteolytic Clostridium botulinum strains (Group II). These strains, producers of toxins of types B, E, or F, are considered the most dangerous in low acid REPFEDs. Thus, they are the target organisms for the control of spore forming pathogens (Carlin et al., 2000), as a consequence of their capacity to survive to heat process usually followed for these products, and of their posterior germination, outgrow and toxin production at temperatures as low as 3 ºC (Graham et al., 1997). Botulism outbreaks associated with non proteolytic C. botulinum have occurred most frequently with processed ﬁsh, namely salted, dried, hot smoked-vacuum packed and fermented marine products (Lindström, Kiviniemi and Korkeala, 2006), some of them involving temperature abuse, being the majority of European outbreaks associated to type B neurotoxin (Peck, 2006). Spores of psychrotrophic strains of C. botulinum are moderately heat resistant, since heating at 80-85 ºC for 5-10 min inactivates the spore germination system (Lund and Peck, 1994). D-values reported in literature for meat products are scarce and often refers to fish and other seafood because of their associated high risk of botulism: 0.43≤ D90.0ºC ≤8.2 min and 8.3≤ z ≤9.8 ºC, for types B and E respectively (Lindström, Kiviniemi and Korkeala, 2006; Silva and Gibbs, 2010). From

CHAPTER 3 ______

seafood, crab meat presents the highest heat resistance in strains type B, D94.4ºC =2.9 min and z-value = 8.6 ºC (Peterson et al., 1997) and, values reported for pork, are of

1.5≤ D82.2ºC ≤ 32.3 min, 6.5 ≤ z ≤16.5 ºC (type B) (Scott and Bernard, 1982). The z- value is defined as the number of Celsius degrees required to reduce the D-value by a factor of 10. For turkey slurry with added lysozyme, the values are of D90.0ºC = 0.8 min, z = 9.4ºC (type B); D85.0ºC =1.2 min, z = 9.9 ºC (type E) (Juneja et al., 1995) and, for turkey slurry with 1% salt (wt/vol) and added lysozyme, are D90.0ºC = 1.1 min, D85.0ºC = 7.8 min, z = 10.1 ºC (type B) (Juneja and Eblen, 1995). Other spore former bacteria of particular concern in REPFEDs are Clostridium perfringens and Bacillus cereus. D-values reported in literature indicated that B. cereus spores are more heat resistant than those of psychrotrophic strains of C. botulinum but less than C. perfringens spores (Byrne, Dunne and Bolton, 2006). The aim of this study was (i) to assess the adequacy of USHTE using Heisler charts to predict temperature in the center of the product and the time needed to attain the temperature needed and (ii) to determine if the thermal treatment traditionally applied to this cooked blood sausage complies with the recommendations for REPFEDs.

3.3 MATERIALS AND METHODS

3.3.1 Processing of the MA

Formulation included Alentejano pig breed meat (54%), rice (13%), pork blood (6%), water (17%), lard (4%), ground onion (3%), salt (1%) and ground spices mixture (cumin, clove, allspice, pepper and mint) (2%). Meats were minced (screed with diameter apertures of 2.0 cm) and mixed with raw rice, pork blood, water, ground onion and mint, lard, salt and ground spices. The mixture was mechanically stuffed into 60-65 mm natural beef casings and hand tied with cotton thread. MA was then transferred into a stainless steel cooking vessel batch system (0.715 m in diameter and 0.565 m in height) heated by a gas burner and cooked in the molho de moura (a broth prepared with water, whole onions, whole garlic, salt and laurel leaves). Sufficient amount of heating fluid (molho de moura) for covering the products was put into the cooking vessel. The heating fluid was heated until reach the process temperature, before

CHAPTER 3 ______

immersion of the products for cooking. After cooking, products were air cooled at room temperature and then transferred into a refrigerated chamber.

3.3.2 Sampling

Raw MA sausages presented an average length of 16.9 ± 1.9 cm, average diameter of 5.1 ± 0.2 cm and average weight of 315.6 ± 39.5 g. Thermocouples (Type SSS-20120-G700-TS 12, Ellab A/S, Hilleroed, Denmark) were placed in the center of each MA and in the heating fluid (Type SSR-60020-G700-TS, Ellab A/S, Hilleroed, Denmark) in order to register heat penetration curves. The thermocouples were connected to a data acquisition system (CTF9008, Ellab A/S, Hilleroed, Denmark), with temperatures recorded every minute. Products were weighed and measured (diameter and length) before and after cooking. During cooling and chilling, the core temperature of the products and of the air was also recorded and the data acquisition interval was set as 1 min. From each of the four batches studied, six MA were collected (three for microbiological analysis and three for physicochemical analysis) and transported to the laboratory in sterile sample bags under refrigeration and analyzed immediately on arrival to the laboratory. A total of 24 heat penetration curves from four product batches were obtained and six sausages from each batch were randomly used for this study.

3.3.3 Cooking loss

The cooking loss (w/w, %) was calculated from the weight ratios of the whole batch, before and after cooking and expressed as a percentage.

CHAPTER 3 ______

3.3.4 Microbiological analysis

Twenty-five gram slices of MA were taken aseptically and homogenized with 225 mL of sterile Ringer´s solution (Merck, Darmstadt, Germany) for 90 s in sterile plastic bag in a lab blender (Masticator, IUL Instruments, Barcelona, Spain). Serial decimal dilutions in sterile Ringer´s solution were prepared and 0.1 mL samples of appropriate dilutions were spread onto selective agar plates. Enumeration of Bacillus cereus was performed according to ISO 7932 (2004) and Clostridium perfringens as described in ISO 7937 (2004) standard.

3.3.5 pH and water activity assays

The pH potentiometric measurements were carried out by inserting the electrode of the pH meter (GLP 21, Crison, Barcelona, Spain) directly into the sample (20.4 ± 0.3ºC).

Prior to water activity (aw) measurements, each sample was ground and homogenized in a domestic food processor. Rotronic Hygroskop DT humidity sensor (Rotronic, HygroLab,

Bassersdorf, Switzerland) was employed to measure aw at 25.0 ± 0.01 ºC.

3.3.6 Proximate composition of MA

Chemical composition was determined according to ISO recommended procedures: moisture (ISO 1442, 1997), total fat (ISO 1443, 1973), protein (Kjeldahl; N x 6.25) (ISO 937, 1978), NaCl (ISO 1841-1, 1996) and ash (ISO 936, 1998) contents. Carbohydrates content was estimated by subtracting to 100% the moisture, protein, fat and ash percentages.

CHAPTER 3 ______

3.3.7 Unsteady state heat transfer equations (USHTE) using Heisler charts

A cylindrical geometry was considered, thermal and physical properties were collected from literature (Table 3.1) (Earle, 1983; Geankoplis, 1993; McCabe, Smith and Harriott, 1993) and the adimensional numbers were calculated using these values.

The convection-heat transfer coefficient (hc) was calculated using the following dimensionless equation for natural convection about vertical cylinders equation (Earle, 1983):

(Nu) = 0.53(Pr.Gr)0.25 for 104 < (Pr.Gr) < 109 (eq. 1) where, Nu is the Nusselt number, which consists of the ratio of convective to conductive heat transfer across the boundary; Gr is the Grashof number, which approximates the ratio of the buoyancy to viscous forces acting on a fluid, and Pr is the Prandtl number, defined as the ratio of kinematic viscosity to thermal diffusivity (Earle, 1983; Geankoplis, 1993; McCabe , Smith and Harriott, 1993). Also,

Nu = hcL/k (eq.2)

-2 -1 -1 where, hc is the convection-heat transfer coefficient (Jm s K ), L is the characteristic length (m) and k is the thermal conductivity of the fluid (Jm-1s-1K-1).

Pr = cpµ/k (eq. 3) where, cp is the specific heat at constant pressure (Jkg-1K-1), µ is the viscosity (Nsm-2) and k is the thermal conductivity of the fluid (Jm-1s-1K-1).

Gr = L3ρ2gβΔT/µ2 (eq. 4) where, L is the characteristic length (m), ρ is the density (Kgm-3), g is the gravitational acceleration (ms-2), β is the thermal expansion coefficient (K-1), ΔT is the difference temperature term and µ is the viscosity of the fluid (Nsm-2).

CHAPTER 3 ______

Table 3.1 Thermal and physical properties considered for MA (Earle, 1983; Geankoplis, 1993; McCabe , Smith and Harriott, 1993). MA  (Kgm-³) a 1070.0 Cp (JKg -1K-1) b 3.350x103

k (Jm-1s-1K-1) c 0.480

α (m2s-1) d 1.339x10-7

β (K-1) e 6.986x10-4

Gr f 5.055x107

Prg 1.799

Nu h 51.753

-2 -1 -1 i hc (Jm s K ) 100.548 a- density; b- specific heat; c - thermal conductivity; d - thermal diffusivity; e - thermal expansion coefficient; f - Grashof number; g - Prandtl number; h - Nusselt number; i - convection-heat transfer coefficient.

3.4 RESULTS AND DISCUSSION

3.4.1 MA characterization and processing

MA with a pH around 6.5 and aw of 0.97, is a product of difficult preservation (Table 3.2). These results are similar to those found in other cooked blood sausages from Argentina, Spain and Slovenia (Oteiza, Giannuzzi and Califano, 2003; Santos et al., 2003; Gašperlin et al., 2012) but different from morcilla, a cured-cooked blood sausage from Spain (Cachaldora et al., 2013). MA proximate composition shows that water is the major component (53.9%), but lower than the amounts found by Santos et al. (2003) in morcilla de Burgos, a Spanish cooked blood sausage (62.2%) or in Krvavica (61.5%), a Slovenian cooked blood sausage (Gašperlin et al., 2012). This difference can be explained by the use in both products of pre-cooked rice and other starch components, with a high water content, before mixing with the rest of the ingredients, while in MA formulation, the added rice (about 13%) is not pre-cooked.

CHAPTER 3 ______

Table 3.2 Mean and standard deviation (sd) of physicochemical parameters and proximate composition of MA. Mean sd pH 6.4 0.06 aw 0.97 0.01 Water (g/100 g) 53.9 1.15 Protein (g/100 g) 15.8 0.64 Fat (g/100 g) 14.1 1.82 Carbohydrate (g/100 g) 14.3 0.92 Ash (g/100 g) 1.9 0.11 NaCl (g/100 g) 1.9 0.09

The proximate composition of MA shows that it is a nutritious meat product (Table 3.2). The protein content of MA is higher than that found in other cooked blood sausages, namely in morcilla de Burgos (13.1%) and Krvavica (10.4%) which may be due to the high content of meat (54%) in MA formulation. Morcilla de Burgos includes a high content of onion and rice and it does not include meat in its formulation but only blood, up to 20% (Santos et al., 2003) while Krvavica includes blood (up to 20%) and other offal, pork skin, entrails, meat from the pig head, trimmings and cracklings (10%) and a major level of starch components (Gašperlin et al., 2012). MA presents approximately half of the fat content of morcilla de Burgos (28.7) and Argentinean morcilla (32%) (Oteiza et al., 2003) but similar fat and higher carbohydrate contents comparatively to Krvavica (Gašperlin et al., 2012). The average cooking loss of 4 batches was 20.3%. Cooking of meat is a multiphase and multicomponent transport with phase change, in which water evaporates, fat melts (Dhall, Halder and Datta, 2012) and protein denaturation occurs (Promeyrat, Daudin and Gatellier, 2013) with increasing temperature. Cooking of meat products in water is reported to present the lowest cooking loss comparatively to other thermal processes, namely dry air and steam cooking, because the product is surrounded by water and little evaporation occurs (Cheng, Sun and Scannell, 2005). MA, in which no phosphates or other chemical additives are used, loses about 20% of its initial weight. Wang et al. (2010) in blood curd, a Chinese blood food, made from a mixture of fresh porcine or duck blood and water and coagulated by heat, reported a cooking loss of 26.5%. The specificities of MA formulation, namely the use

CHAPTER 3 ______

of a high proportion of meat in which approximately half is lean, therefore with more myofibrillar protein and higher water binding capacity may partially explain the obtained results of cooking loss. Myofibrillar proteins are responsible for holding most of the water retained in muscle and during cooking, in the temperature range from 40ºC to 90ºC, they denature and shrink with substantial water loss (Tornberg, 2005). Additionally, MA formulation includes raw rice at level of about 13%, this cereal, depending on cultivar, is capable of considerable water uptake during cooking in water as a consequence of starch gelatinization (Shinde et al., 2014). Therefore, the inclusion of rice at a level of about 13% in MA formulation could also explain the observed cooking loss of the product. MA has a long thermal process (Figure 3.1). The average total time of thermal process (including product cooling and chilling) was between 335.2 ± 5.9 and 396.0 ± 8.1 min, for the four studied batches (Table 3.3).

Temperature (ºC)

100.0 T core MA batch 1 95.0 T media batch 1 90.0 T core MA batch 2 T media batch 2 85.0 T core MA batch 3 80.0 T media batch 3 75.0 T core MA batch 4 T media batch 4 70.0 65.0 60.0 55.0 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 0 15 30 45 60 75 90 105120 135150165180 195210225 240255270285 300315330345 360375390405 time (min) Figure 3.1 Thermal processing of MA.

CHAPTER 3 ______

Heating medium average processing temperature varied from 94.4 ± 5.1 to 95.2 ± 4.6 ºC, among the four batches. Average cooking time (until the beginning of cooling) was between 88.0 ± 2.6 and 103.3 ± 0.5 min while product core maximum temperature was around 98ºC in the four batches (Table 3.3). This temperature is slightly higher than the cooking temperatures reported in several studies, namely in Spanish (Santos et al., 2003, 2005; Cachaldora et al., 2013), Argentinian (Oteiza et al., 2003, 2006) and Slovenian (Gašperlin et al., 2012) blood sausages that are comprised between 85 and 96ºC during a period of time varying, according to the product, from 30 min to 4 hours.

Table 3.3 Thermal process parameters of MA (mean ± sd in four batches).

Batch 1 Batch 2 Batch 3 Batch 4

Heating medium initial T (ºC) 86.2 ± 2.6 85.8 ± 5.5 83.9 ± 0.4 87.7 ± 5.2 Heating Heating medium average processing T (ºC) 92.6 ± 5.9 95.2 ± 4.6 94.4 ± 5.1 92.7 ± 6.6 Heating medium T máx. (ºC) 98.5 ± 0.1 98.7 ± 0.2 98.5 ± 0.1 98.6 ± 0.5

Product average processing T (ºC) 72.3 ± 0.7 81.9 ± 1.5 76.1 ± 2.2 78.6 ± 1.8 Product T máx. (ºC) 97.8 ± 0.5 98.2 ± 0.3 97.5 ± 0.8 98.0 ± 0.3 Cooking Time to reach product T máx. (min) 84.8 ± 4.2 94.8 ± 6.6 87.0 ± 7.8 82.2 ± 7.7 Time at T product > 90 ºC (min) 37.8 ± 2.5 65.7 ± 1.0 51.7 ± 7.7 46.0 ± 5.2 Time at T product > 95 ºC (min) 24.7 ± 4.6 53.5 ± 4.5 39.3 ± 8.4 32.5 ± 10.9 Time of cooking (min) 88.0 ± 2.6 103.3 ± 0.5 100.3 ± 1.4 94.0 ± 0.6

Air room medium average T (ºC) 20.2 ± 1.7 22.1 ± 1.2 24.2 ± 1.1 22.5 ± 1.3 Time of product cooling to 35.0 ºC (min) 73.2 ± 2.9 64.3 ± 4.8 82.7 ± 3.7 68.7 ± 5.0 Cooling Total time of product cooling 108.0 ± 2.4 78.7 ± 0.5 78.7 ± 1.2 77.8 ± 0.7 Cooled product T (ºC) 27.2 ± 1.1 30.8 ± 1.5 36.1 ± 1.3 32.3 ± 1.4

Air chilling medium average T (ºC) 4.2 ± 1.2 5.5 ± 1.3 4.4 ± 1.0 4.2 ± 1.3 Chilling Time of product chilling to 5.0 ºC (min) 288.2 ± 11.3 231.8 ± 6.1 295.7 ± 8.0 250.2 ± 6.9 Total duration of thermal process (min) 376.2 ± 12.7 335.2 ± 5.9 396.0 ± 8.1 344.2 ± 6.9

Thermal inactivation of both vegetative and bacteria spores is a critical control point in the safe preparation of many RTE foods. According to the Food Standards Agency (FSA, 2008), if the heat treatment is the single control factor in the production of chilled VP and MAP foods, the minimum heat treatment that should be used to manufacture is 90°C for 10 min or equivalent, achieved at the slowest heating point in the product. FSA (2008) recommends for the safe production of VP and MAP chilled

CHAPTER 3 ______

foods with a shelf life longer than 10 days, with respect to growth and toxin production by non proteolytic C. botulinum strains includes, single or in combination, a minimal thermal treatment of 10 min/90 ºC, achieved at the slowest heating point in the product, and a refrigeration temperature of between 3 and 8 ºC, which should be maintained in the entire cold chain; a pH value lower than 5 throughout the food and throughout all components of complex foods; an aw lower than 0.97, throughout the food and throughout all components of complex foods; a minimum salt level of 3.5% in the aqueous phase, throughout the food and throughout all components of complex foods; or a combination of heat and preservative factors which can be shown consistently to prevent growth and toxin production by non proteolytic C. botulinum. MA complies with almost of all of these recommendations, except for the pH value which is about 6.5 (Table 3.2). MA thermal process comprises in average, a core product temperature higher than 90.0 ºC for 50.3 min (Table 3.3), which largely exceeds the most conservative safe harbour of 90.0 ºC/10min. Product refrigeration temperature is within the range of 3 to 8 ºC (Table 3.3), the aw of the product is lower than 0.97 (Table 3.2) and complies with the minimum salt level of 3.5% in the aqueous phase (Table 3.2). Besides the adequate binomial time/temperature during thermal processing, the cooling phase must be rapidly achieved to prevent germination of bacterial spores which are not destroyed by cooking and multiplication of their vegetative cells (Juneja and Marks, 1999). Considering the cooling phase, C. perfringens is the spore former bacteria of concern by presenting the fastest growth between 25 ºC and 45 ºC (Poumeyrol et al., 2014). Almost all outbreaks resulted from slow cooling of foods or holding without refrigeration. The effective prevention of food borne diseases caused by this spore former is achieved by appropriate cooking, fast cooling through the temperature range of 55-15 °C, holding foods at temperatures lower than 10-12 °C, and re-heating the product to an internal temperature of 72 °C before consumption (EFSA, 2005a). According to Daelman et al. (2013), reheating by the consumer cannot be considered an effective method to assure REPFEDs safety with respect to heat resistant pathogenic microorganisms. MA cooling and chilling complies with FDA (2009) recommendations, who indicates that cooked potential hazardous foods should be cooled from 57 ºC to 21 ºC

CHAPTER 3 ______

within 2 hours and within a total of 6 hours from 57 ºC to ≤ 5 ºC. MA cooling took, on average, 85 min in the range of 57 ºC to 21 ºC and 237 min (3.95 hours) in the range of 57 ºC to 5 ºC.

3.4.2 Temperature/time prediction using Heisler charts

The adequacy of unsteady-state heat transfer equations (USHTE) using Heisler charts to predict time/temperature values in the center of the product, were tested for MA, using equations 1 to 4 and data presented in Tables 3.1 and 3.3. The predicted results, together with the experimentally measured values, are presented in Table 3.4. For the 4 batches, the differences between experimental and predicted temperatures were only of 0.6, 0.7, 0.0 and 0.5 ºC, respectively. These results show very small differences between the experimental and predicted temperatures in the four batches studied, indicating a very good fit of USHTE to the experimental data as well as the adequacy of the thermal and physical properties collected from the literature (Earle, 1983; Geankoplis, 1993; McCabe, Smith and Harriott, 1993) and used in these calculations. For the 4 batches, differences between the experimental and predicted times were 10, 10, 15 and 10 min, respectively (Table 3.4).

Table 3.4 Experimental temperature and time and predicted temperature and time of MA in four batches. Batch 1 Batch 2 Batch 3 Batch 4

Experimental T (ºC) 97.8 98.2 97.5 98.0 Experimental time (min) 85 95 87 82 Predicted T (ºC) 98.4 98.2 98.2 98.5 Predicted time (min) 95 105 102 92

Considering the duration of the thermal heating process (88-103 min), the results indicate a good prediction accuracy of the model used. These findings demonstrate the feasibility of using the USHTE model to predict, control and optimize thermal processing of MA, in terms of food safety. In addition, the USHTE model allows for 81

CHAPTER 3 ______

adjusting external conditions, and product dimensions, without the need for the continuous monitoring of MA internal temperature with thermocouples.

3.4.3 Microbiological analysis of MA

Bacillus cereus and Clostridium perfringens were not detected in any of the tested samples of MA. The thermal processing of MA, involving temperatures in the core of the product above 90 ºC, on average, for 50.3 min, and above 95 ºC, on average, for 37.5 min (Figure 3.1 and Table 3.3), allows for the elimination of both vegetative forms and spores. The D-values of C. perfringens spores, in pork luncheon roll, reported by Byrne, Dunne and Bolton (2006), are: D90.0ºC = 30.6 min, D95.0ºC = 9.7 min, D100.0ºC = 1.9 min, z= 8.3 ºC. These values indicate a higher heat resistance of C. perfringens spores than those exhibited by psychrotrophic strains of C. botulinum. B. cereus is reported to have a generation time of 11 to 27 min, capable of growth in the pH range from 4.9 to 9.3 and an optimum growth temperature of 28-35 ºC with a minimum of 4-5 ºC and maximum of 55 ºC (Afchain et al., 2008). However, strains capable to multiply below 7 °C and strains able to multiply above 45 °C, are considered not very common (EFSA, 2005b). Psychrotrophic strains of B. cereus, although less heat resistant (Guinebretière et al., 2008), involved in food poisoning, have been isolated in cooked chilled foods containing vegetables (Carlin et al., 2000) and meat products, namely meat and rice dishes and meat stew (Borge et al., 2001) being of special interest in the modern RTE refrigerated foods industry (Afchain et al., 2008). Thermal resistance values of spores of B. cereus, reported in pork luncheon roll, are: D85.0ºC = 29.5 min, D90.0ºC = 10.1min, D95.0ºC = 2.0 min, z= 8.6 ºC (Byrne, Dunne and Bolton, 2006). These values indicate that B. cereus spores are more heat resistant than those of psychrotrophic strains of C. botulinum but less resistant than C. perfringens spores. The maximum average temperature of MA (product T máx.) in the four batches was 97.9 ºC (Table 3.3) and the thermal process of MA (Figure 3.1 and Table 3.3) shows that the product is, on average, 37.5 min at a temperature above 95 ºC. This time/temperature combination allows the thermal destruction of both B. cereus and C.

CHAPTER 3 ______

perfringens spores according to the D-values reported in the literature. Furthermore, MA cooling and chilling complies with FDA (2009) recommendations. These operations are short enough to limit the growth of C. perfringens and B. cereus and therefore should be considered to be used as a critical control point (Poumeyrol et al., 2010). Additionally, the Good Hygienic and Good Manufacturing Practices implemented by the producer may determine low initial contamination levels for both spore formers and therefore to increase the effectiveness of the thermal process and help to explain why the presence of these spore formers were not detected in the present work.

3.5 CONCLUSIONS

A very good fit of USHTE to the experimental time/temperature data was observed. The USHTE model used allows for adjusting external conditions and product dimensions, without the need for the continuous monitoring of MA internal temperature. The model can easily be used to predict the adequacy of MA thermal process in terms of food safety. The cooking process applied to MA is long and allows to safely extending the refrigerated shelf life of this product by ensuring a reduction of the most heat resistant facultative psychrotrophic vegetative pathogen of concern in RTE meat products, Listeria monocytogenes, largely above the 6D recommended for this type of meat product. The thermal process also ensures a 6D reduction of the number of spores of psychrotrophic non proteolytic C. botulinum strains. Generally, MA complies with the recommendations for the safe production of VP and MAP chilled foods with a shelf life longer than 10 days, with respect to the growth and toxin production by non proteolytic C. botulinum. The traditional thermal process, although following empirical procedures, involving temperatures above 95 ºC for about 38 min in the core of MA, is capable of destroying other more heat resistant spores namely of B. cereus and C. perfringens, according to D-values reported in the literature. However, this thermal process cannot

CHAPTER 3 ______

theoretically destroy the most heat resistant proteolytic strains of C. botulinum. That might be a problem only if a temperature abuse occurs. Further research is needed to investigate the safety and microbiological stability of this product during shelf life, namely with the application of packaging technologies. Good Hygienic and Manufacturing Practices must continue to be included in the control of the process, with particular attention to the thermal process and storage temperature.

3.6 ACKNOWLEDGEMENTS

This work received financial support from FCT, Fundação para a Ciência e a Tecnologia, Portugal, through the PhD grant SFRH/BD/63341/2009. This study was also supported by the Strategic Project PEst-OE/AGR/UI0245/2011 of CEER, a research unit supported by the FCT.

3.7 REFERENCES

Afchain, A. L., Carlin, F., Nguyen-Thé, C. and Albert, I. (2008). Improving quantitative exposure assessment by considering genetic diversity of B. cereus in cooked, pasteurized and chilled foods. International Journal of Food Microbiology, 128, 165-173.

Byrne, B., Dunne, G. and Bolton, D. J. (2006). Thermal inactivation of Bacillus cereus and Clostridium perfringens vegetative cells and spores in pork luncheon roll. Food Microbiology, 23, 803-808.

CHAPTER 3 ______

Carlin, F., Girardin, H., Peck, M. W., Stringer, S. C., Barker, G. C., Martinez, A., Fernandez, A., Fernandez, P., Waites, W. M., Movahedi, S., van Leusden, F., Nauta, M., Moezelaar, R., Del Torre, M. and Litman, S. (2000). Research on factors allowing a risk assessment of spore-forming pathogenic bacteria in cooked chilled foods containing vegetables: a FAIR collaborative project. International Journal of Food Microbiology, 60, 117-135.

Cheng, Q., Sun, D.-W. and Scannell, A. G. M. (2005). Feasibility of water cooking for pork ham processing as compared with traditional dry and wet air cooking methods. Journal of Food Engineering, 67, 427-433.

Daelman, J., Jacxsens, L. Devlieghere, F. and Uyttendaele, M. (2013). Microbial safety and quality of various types of cooked chilled foods. Food Control, 30, 510-517.

Dhall, A., Halder, A. and Datta, A. K. (2012). Multiphase and multicomponent transport with phase change during meat cooking. Journal of Food Engineering, 113, 299-309.

Earle, R. L. (1983). Unit operations in food processing. 2nd ed., Pergamon Press, Oxford, England.

EFSA (European Food Safety Authority) (2005a). Opinion of the scientific panel on biological hazards on the request from the commission related to Clostridium spp. in foodstuffs. EFSA Journal, 199, 1-65.

EFSA (European Food Safety Authority) (2005b). Opinion of the scientific panel on biological hazards on Bacillus cereus and other Bacillus spp. in foodstuffs. EFSA Journal, 175, 1-48.

CHAPTER 3 ______

Gašperlin, L., Skvarča, M., Žlender, B., Lušnic, M. and Polak, T. (2012). Quality assessment of Slovenian krvavica, a traditional blood sausage: sensory evaluation. Journal of Food Processing and Preservation, 38, 97-105.

Geankoplis, C. J. (1993). Transport processes and unit operations. 3rd ed., Prentice- Hall International Inc., New Jersey, U.S.A.

International Organization for Standardization (ISO) 936 (1998). Meat and meat products. Determination of ash content, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 937 (1978). Meat and meat products. Determination of nitrogen content, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 1442 (1997). Meat and meat products. Determination of moisture content, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 1443 (1973). Meat and meat products. Determination of total fat content, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 1841-1 (1996). Meat and meat products. Determination of chloride content. Part 1: Volhard method, International Organization for Standardization, Geneva, Switzerland.

CHAPTER 3 ______

International Organization for Standardization (ISO) 7932 (2004). Microbiology of food and animal feeding stuffs. Horizontal method for the enumeration of presumptive Bacillus cereus. Colony-count technique at 30 ºC, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 7937 (2004). Microbiology of food and animal feeding stuffs. Horizontal method for the enumeration of Clostridium perfringens. Colony-count technique, International Organization for Standardization, Geneva, Switzerland.

Juneja, V. K. and Marks, H. M. (1999). Proteolytic Clostridium botulinum growth at 12-48 ºC simulating the cooling of cooked meat: development of a predictive model. Food Microbiology, 16, 583-592.

Lund, B. M. and Peck, M. W. (1994) Heat resistance and recovery of non-proteolytic Clostridium botulinum in relation to refrigerated processed foods with an extended shelf-life. Journal of Applied Bacteriology, 76, 115-128.

McCabe, W. L., Smith, J. C. and Harriott, P. (1993). Unit operations of chemical engineering. 5th ed., McGraw-Hill, Inc., New York, U.S.A.

CHAPTER 3 ______

Peck, M. W. (2006). Clostridium botulinum and the safety of minimally heated, chilled foods: an emerging issue? Journal of Applied Microbiology, 101, 556-570.

Peterson, M. E., Pelroy, G. A., Poysky, F. T., Paranjpye, R. N., Dong, F. M., Pigott, G. M. and Eklund, M. E. (1997). Heat-pasteurization process for inactivation of non proteolytic types of Clostridium botulinum in picked Dungeness crabmeat. Journal of Food Protection, 60, 928-934.

Poumeyrol, G., Morelli, E., Rosset, P. and Noel, V. (2014). Probabilistic evaluation of Clostridium perfringens potential growth in order to validate a cooling process of cooked dishes in catering. Food Control, 35, 293-299.

Poumeyrol, G., Rosset, P., Noel, V. and Morelli, E. (2010). HACCP methodology implementation of meat pâté hazard analysis in pork butchery. Food Control, 21, 1500-1506.

Promeyrat, A., Daudin, J. D. and Gatellier, P. (2013). Kinetics of protein physicochemical changes induced by heating in meat using mimetic models: (1) Relative effects of heat and oxidants. Food Chemistry, 138, 581-589.

Scott, V. N. and Bernard, D. T. (1982). Heat resistance of spores from non-proteolytic type B Clostridium botulinum. Journal of Food Protection, 45, 909-912.

CHAPTER 3 ______

Shinde, Y. H., Vijayadwhaja, A., Pandit, A. B. and Joshi, J. B. (2014). Kinetics of cooking of rice: A review. Journal of Food Engineering, 123, 113-129.

Tornberg, E. (2005). Effects of heat on meat proteins - implications on structure and quality of meat products. Meat Science, 70, 493-508.

Wang, P., Han, M.-Y., Dong, Q.-L., Xu, X.-L., Zhou, G.-H. and Lu, S.-L. (2010). Principal component analysis for textural properties of selected blood curd. Journal of Texture Studies, 41, 757-773.

CHAPTER 4

Pereira, J. A., Dionísio, L., Matos, T. J. S. and Patarata, L. Sensory lexicon development for a Portuguese cooked blood sausage - Morcela de Arroz de Monchique. Submitted for publication in Journal of Sensory Studies

CHAPTER 4 ______

4. SENSORY LEXICON DEVELOPMENT FOR A PORTUGUESE COOKED BLOOD SAUSAGE -

MORCELA DE ARROZ DE MONCHIQUE

Pereira, J. A.a,b*, Dionísio, L.c,d, Matos, T. J. S. b,e and Patarata, L.f

aDepartment of Food Engineering, Instituto Superior de Engenharia, University of Algarve, Campus da Penha, Estrada da Penha, 8005-139 Faro, Portugal bCEER - Biosystems Engineering, Tapada da Ajuda, 1349-017 Lisbon, Portugal cDepartment of Biological Sciences and Bioengineering, Faculdade de Ciências e Tecnologia, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal dICAAM - Institute of Mediterranean Agricultural and Environmental Sciences, P.O. box 94, University of Évora, 702-554 Évora, Portugal eDepartment of Science and Biosystems Engineering, Instituto Superior de Agronomia, Technical University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal fCECAV, Centro de Ciência Animal e Veterinária, Universidade de Trás-os-Montes e Alto Douro, Quinta de Prados, 5000-801 Vila Real, Portugal

4.1 ABSTRACT

Morcela de Arroz (MA) is a popular cooked blood and rice sausage produced in the Monchique region, south of Portugal, whose characteristics are not well known and no data is available about its sensory characteristics. The aim of this study was to build up terminology to develop a quantitative descriptive analysis for Morcela de Arroz from Monchique region in comparison to other blood and rice sausages, in order to improve its sensory description.

4.2 PRACTICAL APPLICATIONS

This study provides the necessary lexicon to be used in quantitative descriptive analysis of Morcela de Arroz, particularly to discriminate the origin of the product. That lexicon will allow the characterization of traditional MA viewing to apply for one of the EC special labels certification: Protected Designation of Origin (PDO) or

93 CHAPTER 4 ______

Protected Geographical Indication (PGI). The list of selected descriptors and references should help researchers to elucidate about the factors which influence typical appearance, odour, flavour and texture, and also to improve sensory quality control in this type of meat products.

4.3 INTRODUCTION

Morcela de Arroz (MA) is a popular cooked blood sausage produced in the Monchique region, south of Portugal. This ready-to-eat meat product supports the economy of several industries, due to its uniqueness among the traditional Portuguese meat products. It is made with grinded pork meat and blood, rice, grinded onion, spearmint, spices (cumin, allspice, clove and black pepper) and salt, stuffed in natural casings (pork stomach, pork or cattle intestine), tied with thread and then boiled in a broth prepared with water, whole onions, whole garlic, salt and laurel leaves (Pereira, Ferreira-Dias and Matos, 2008). Cooked blood sausages are traditional meat products popular in several countries, namely morcela in Portugal, morcilla in Spain (Santos et al., 2003) and Argentina (Oteiza et al., 2006), sanguinaccio in Italy, black pudding in UK, blutwurst in Germany, boudin noir in France, kaszanka in Poland and krvavica in Slovenia (Gašperlin et al., 2012). Their consumption is popular in many European countries and the ingredients used in their preparation differ according to the producing region. The EC is involved in the protection of the quality of traditional foods from specific regions to promote rural areas and support local population. Special labels, namely Protected Designation of Origin (PDO) and Protected Geographical Indication (PGI), have been created in the EC (European Regulation No 510/2006). Monchique MA has a high potential to apply for one of these special labels certification. That process implies that the product must be well characterized, namely its sensorial traits. However, with the best of our knowledge, there is no report about MA in the literature. The sensory characterization of a food product is frequently achieved by descriptive sensory analysis. Although there are several new methodological approaches developed as alternatives to conventional profiling (Valentin et al., 2012), Quantitative Descriptive Analysis (QDA) (Stone et al., 1974) is still a reference technique. This technique consists of a sequential survey of sensory terms for a product generated by a sensory panel using nontechnical language and supervised by a leader who should not be an active participant

94 CHAPTER 4 ______in the process. This method provides a predeﬁned terminology to describe and quantify sensory perceptions as objectively as possible (Moskowitz, 1983). However, the validity of the results obtained depends on the correct initial selection of descriptors and on the ability of the judges to identify and quantify each speciﬁc attribute. Literature reports lexicon development strategies for descriptive sensory analysis of several meat and meat products (Pérez-Cacho et al., 2005; Patarata et al., 2008; Adhikari et al., 2011; Maughan et al. 2012; Vásquez-Araújo et al., 2013) and other foods namely for plant foods and ingredients (Smyth et al., 2012), rice (Bett-Garber et al. ,2012), potato puree (Jiménez, Canet and Alvarez, 2013), soy sauce (Cherdchu, Chambers IV and Suwonsichon, 2013), and prebiotic gluten-free bread (Morais et al., 2014) with the objective of quality control and/or product certification. The aim of this study was to build up terminology to develop a quantitative descriptive analysis for Morcela de Arroz from Monchique region in comparison to other blood and rice sausages, in order to improve its sensory description.

4.4 MATERIALS AND METHODS

4.4.1 Formation of the panel

A panel of 18 assessors (fourteen men and four women) was selected from an initial group of 63 recruited individuals. This procedure was conducted according the recommendations of ISO 8586-1 (1993) during a total of 17 sessions. In a first stage, with all the individuals, it was evaluated: i) their capacity to detect and identify tastes thresholds (sweet, salty, acid, bitter, metallic, umami); ii) their ability to memorize and identify aromas (cumin, clove, garlic, butter, rancid, spearmint, bitter almond, mushrooms, floral, lemon and banana), all sessions were preceded by familiarization sessions with the basic tastes and aromas solutions (chemicals were supplied by Proaromática-Aromas Alimentares, Lda, Lisbon, Portugal); iii) their discriminative capacity trough ranking of 4 different intensity levels of cumin aroma; ranking the hardness of 4 dry sausages with different levels of drying, and, a triangle test with similar blood sausages from two different origins and; iv) the texture description test (oranges, double cream, cooked leeks, Gouda cheese, chips, raw carrots, pears puree, granulated sugar, caramel topping and rice cake). From this assemble of 14 sessions,

95 CHAPTER 4 ______were eliminated the 40% of assessors with worst performance. In the second stage, from the remain 38 assessors with better performance in the first stage, the selection of the final panel was based on their discriminative capacity and their ability of assessments repeatability, calculated using the data obtained in repeated quantitative descriptive analyses of samples of MA from three different origins, analysed repeatedly in three sessions held at 1-week intervals.

4.4.2 Samples of MA

Morcela de Arroz de Monchique is produced on widely varying scales, from the smallest farm production units to industrial facilities that supply the majority of the market demand. It is a meat product with unique technological and sensory characteristics that supports the economy of several industries. This study was conducted with samples from 12 producers (5 from Monchique and 7 from other regions) in the first phase (Figure 4.1), and samples from 7 producers in the following stages of lexicon development (3 from Monchique and 4 from other regions). At the last stage, validation of the vocabulary adequacy, samples from two producers from each origin were used. One sample was comprised by three or four morcelas, depending on individual size. Samples used in the initial open free generation of vocabulary, in the testing with a 5-point+zero scale and in the final validation were obtained from different batches. According to Lawless and Civille (2013) the sample set should be enough to provide a fair representation of the entire product category. The appropriate size of the sample set is dependent on the category. In the present study, the samples used from Monchique region and from other origins are representative of the main producers of this type of blood sausages. Samples of Morcela de Arroz from Monchique region were processed in local producer’s facilities according to the traditional method. Typical formulation included Alentejano pig breed meat (50-55%), rice (10-15%), pork blood (20-25%), lard (2-6%), grinded onion (2.5-5%), salt (about 1%) and ground spices (ground cumin, ground allspice and sometimes ground clove and black pepper) and spearmint. Meat was minced and mixed with raw rice, pork blood, grinded onion, lard, salt, ground spices and spearmint. The mixture was stuffed into 55-75 mm natural casings and tied with cotton thread. MA was then cooked for about 2 h in a broth prepared with water, whole onions, whole garlic, salt and laurel leaves, which is locally referred as molho de moura.

96 CHAPTER 4 ______

Initial open free generation of vocabulary

12 samples – 4 sessions 5 MA1-Mo2; 7 MA-Other

89 raw terms

Screening for hedonistic and quantitative terms, terms referring to the product itself and terms indicated in the incorrect sensorial dimension by the panel leader

Grouping of synonymous and descriptors referring unequivocally to the same characteristic by consensus among panellists – 1 session

76 terms

Testing with a 5-point+zero scale

7 samples – 4 sessions 3 MA-Mo; 4 MA-Other

Reduction based on geometric mean – M < 0.30 were eliminated

52 terms

First reduction of terms based on PCA3 to eliminate similar terms used in odor and flavor dimension

42 terms

Second reduction of terms based on PCA

Final list with 14 terms

Definitions and references for the 14 terms

4 samples (2 MA-Mo; 2 MA-Other) 2 repetitions – 2 sessions

Figure 4.1 Overview of the procedure used for identification and selection of descriptors (adapted from ISO 11035, 1994). 1 MA-Morcela de Arroz, 2 Mo-Monchique, 3 PCA-Principal component analysis.

97 CHAPTER 4 ______

After cooking, products were cooled in a refrigerated chamber (4 ± 1 ºC) and were ready for consumption. Monchique MA samples were kept under refrigeration before analysis during 48h. Samples of other blood and rice sausage types were obtained from commercial suppliers. These morcelas are commercialized only scalded and must be cooked before consumption. Cooking was performed in boiling water for 30 minutes, products were than cooled and stored in refrigeration conditions until use (48h maximum).

4.4.3 Identification and selection of descriptors

The descriptor set was generated according ISO 11035 (1994). The overview of the procedure is illustrated in Figure 4.1. In four sessions involving free and open generation of vocabulary, assessors used the descriptors which they considered most accurate for describing samples of MA from five different producers of Monchique region and samples of commercial MA from seven different origins. The list of terms was first screened for inadequately used terms (hedonistic terms, quantitative terms, terms referring to the product itself, such as “MA taste”, and other incorrect terms of sensorial dimension indicated by the panel leader). Additionally a reduction was conducted based on the grouping of synonymous and descriptors which unequivocally referred the same characteristic. The choice between these terms was based on the opinion of the panellists about its clarity and on the frequency of occurrence of the term in the first sessions of identification of descriptors. This preliminary sorting of descriptors were performed by consensus among panellists, and resulted in a list of 76 terms. The terms were discussed with the panellists to make sure that they have well understood the meaning of each descriptor. The doubts were clarified using available samples for that effect and discussed by the group. The panellists were then presented with seven MA samples (3 from Monchique and 4 from other origins). Panellists were asked, for each descriptor, to judge the perceived intensity in a 6 point scale, where zero corresponded to the absence of the perception and five to the highest intensity, in order to reduce the number of terms based on its geometric mean. Descriptors presenting geometric mean values lower than 0.30 were discarded. Principal component analysis was used to consecutively reduce the list to the 14 descriptors retained.

98 CHAPTER 4 ______

For the final list of 14 descriptors, definitions and references were established, proposed by the panel leader, and discussed with the panellists (Table 4.4). In the present study it was used, whenever possible, as a reference Morcela de Arroz appropriately modified to achieve the desired characteristic intensity. According to Braghieri et al. (2012) the use of the product under study rather than other matrixes as a reference might be helpful to generalize sensory concepts. The scale to be used was defined as a 10 numeric unipolar discrete scale, with zero representing the absence of the characteristic, and 1 to 9 progressively more intensity in the characteristic (ISO 4121, 2003). A final group of analyses were made to validate the adequacy of the descriptors set to distinguish MA from Monchique from those from other provenances, as suggested by Lawless and Civille (2013). Two samples from each provenance were tested in duplicate (one for each session).

4.4.4 Evaluation of the samples

Sessions took place in a sensory analysis laboratory equipped with individual assessment booths and uniform lighting, equivalent to daylight, and temperature conditions (ISO 8589, 2007). In the first four sessions involving free and open generation of vocabulary samples were presented simultaneously. In the following sessions, samples were monadically presented to the panellists. Samples were served in slices (two slices from different morcelas of the same batch) of approximately 1 cm thick in glass Petri dishes with cover and assigned with a random three-digit number, at room temperature (20 ± 2 ºC, tempered for at least 1 h before sensory analysis). The serving order of the samples was random. The assessors were asked to individually evaluate 2 to 3 samples over 40-min sessions. Sessions took place once or twice a week. Commercial spring water and unsalted crackers were provided to clean their palates and reduce build-up of any flavours from one sample to the next.

4.4.5 Data analysis

Data from sensory analysis was evaluated by principal component analysis (PCA). Logistic regression was used to evaluate the ability of the PCA solution to

99 CHAPTER 4 ______predict the samples origin. The eﬀect of MA’s origin (Monchique or other origin) was tested by comparing the means for each variable by one-way ANOVA. Statistical signiﬁcance was set at p values less than 0.05. Data analysis was carried out using XLSTAT 2012.6.09 (Addinsoft, New York).

4.5 RESULTS AND DISCUSSION

In the first group of four sessions, panellists freely generated 89 terms that could describe the sensory characteristics of the samples of 12 origins supplied (5 from Monchique and 7 from other origin). The list was first screened for inadequately terms (hedonistic terms, quantitative terms, terms referring to the product itself, such as “MA taste”, and terms indicated as incorrect in sensorial dimension, such as references to texture in the flavour dimension). Additionally, a reduction of terms was conducted based on the grouping of synonymous and descriptors that unequivocally refers the same characteristic. The choice between these terms was based on the opinion of the panellists about its clarity and on the frequency of occurrence of the term in the first sessions of descriptors identification. A total of 13 terms were removed. Terms used simultaneously in the odour and flavour dimension were kept to study at a subsequent stage. Once the flavour is a complex sense that involves, besides the taste perceived in tongue and the trigeminal stimuli, the odour characteristics perceived by retronasal via, it is expected that some identical descriptors are identified simultaneously in both odour and flavour (Holley, 2006). Several authors (Santos et al., 2003; N’Kouka et al., 2004; Hongsoongnern and Chambers IV, 2008; Zeppa, Bertolino and Rolle, 2012) reported a similar simultaneous occurrence in odour and flavour vocabulary identification for descriptive analysis of several foods.

The reduced list resulted in 76 terms (Table 4.1) and was discussed with the panellists to make sure that they had well understood the meaning of each descriptor. The doubts were clarified using samples available for that effect and discussed by the group. That list included 12 terms for the aspect, 18 for the odour, 33 for flavour and 13 for the texture. The panellists were then presented with the seven variants of MA (3 from Monchique and 4 from other origin) and were asked, for each descriptor, to judge the

100 CHAPTER 4 ______perceived intensity in a 6 point scale, where zero corresponded to the absence of the perception and five to the highest intensity. In the first session of this phase it was included the revision of the understood terminology and one sample was analysed. In the following three sessions two samples were analysed by session.

Table 4.1 List of terms identified by the panellists in the first open sessions and respective geometric mean obtained after analysis with a 6-point scale. Aspect Odour Flavour Texture Terms M Terms M Terms M Terms M R.1 amount 0.80 O.5 Intensity 0.80 Persistence 0.80 Juiciness 0.76 Colour 0.80 Meat 0.70 O. Intensity 0.79 Cohesiveness 0.75 Cohesion 0.80 Blood 0.69 Spices 0.73 Firmness 0.74 Moisten 0.78 Cumin 0.68 Saltiness 0.70 Hardness 0.72 R. dimension 0.77 Fat 0.65 Blood 0.69 Fibrousness 0.72 F.2 dimension 0.74 Pepper 0.57 Rice 0.68 Friability 0.72 F. amount 0.73 Smoke 0.54 Meat 0.64 Granularity 0.71 R. cooking 0.73 Spearmint 0.52 Piquant 0.58 Pasty 0.70 Uniformity 0.73 Wine 0.52 Cumin 0.59 Residue 0.70 P.4 ingredients 0.72 Garlic 0.48 Fat 0.59 Roughness 0.69 M.3 amount 0.70 Clove 0.47 Pepper 0.51 Elasticity 0.66 M. dimension 0.70 Lauryl < 0.3 Smoke 0.48 Adhesiveness 0.64 Rancidity < 0.3 Wine 0.48 Floury < 0.3 Meat broth < 0.3 Sweet 0.47

Coriander < 0.3 Clove 0.42

Herbs < 0.3 Spearmint 0.42

Synthetic < 0.3 Garlic 0.41

Fermented < 0.3 Acid < 0.3

Rancidity < 0.3

Bread < 0.3

Coriander < 0.3

Butter like < 0.3

Pork (boar-taint) < 0.3

Fermented < 0.3

Toasted < 0.3

Bitter < 0.3

Lauryl < 0.3

Umami < 0.3

Meat broth < 0.3

Burned < 0.3

Herbs < 0.3

Synthetic < 0.3

Olive oil < 0.3

1 R.- Rice, 2 F.- Fat, 3 M.- Meat, 4 P.- Proportion between ingredients, 5 O.- Overall.

101 CHAPTER 4 ______

The geometric mean (M) was calculated from the frequency of use of each descriptor and from its relative intensity (Table 4.1). It was arbitrarily considered the cut- point of M = 0.30 to keep a term for further analysis. All terms describing aspect attributes presented high geometric mean (M ≥ 0.70). A similar pattern was observed for texture. Only the term Floury presented an M < 0.30. On the contrary, in the odour and flavour dimensions, almost half of the terms were abandoned (39% and 48%, respectively), indicating that they were referring to attributes rarely detected and/or detected with very low intensity. The higher occurrence of terms with reduced geometric mean for odour and flavour than of other sensory traits was also observed by Pérez-Cacho et al. (2005) in dry- cured sausages. The odour descriptors with low geometric mean where also identified in flavour. Descriptors related specifically to taste – salty, sweet, acid, bitter and to trigeminal stimuli – piquant, and the most part of flavour descriptors with low geometric mean where retronasal odour stimuli, and some of them made reference to complex odours, like meat broth, bread, olive oil or toasted, which might have different interpretation by the panellists (Lawless and Civille, 2013), resulting in the observed reduced geometric mean. After that sorting based on geometric mean, 52 terms were kept. In order to determine the relationship between the attributes retained, viewing the detection of redundant terms and those more discriminant between Monchique MA and MA from other origins, a principal component analysis was performed. From the group of 52 variables, a first reduction was performed to eliminate 10 terms that were simultaneously used for odour and flavour: Blood, Meat, Fat, Smoke, Cumin, Pepper, Wine, Spearmint, Garlic and Clove. It was observed that the most part of these attributes were better represented in the PCA solution with the odour terms than with the correspondent flavour ones, as they presented higher communalities for the first 3 principal components (data not presented), thus being more meaningful sensory descriptors. Although there is no consensus about the relationship between orthonasal (‘forward’ way through the nose) and retronasal (‘backward’ way through the nasopharynx) perception of odour (Holley, 2006), there are evidences that the retronasal ability to identify odorants is inferior than the orthonasal for a range of odorants (Negoias et al., 2008; Bojanowsky and Hummel, 2012), indicating that, for the same odour, the perception is better by smelling the sample. Additionally, we had feedback from the panellists that the evaluation while the samples are masticated was more difficult, because they had a high number of flavour and texture characteristics to analyse. After removing the ten terms from flavour also indicated for odour, 42 variables were used to calculate the first three principal components (PC) (Table 4.2).

102 CHAPTER 4 ______

Table 4.2 Factor loadings and communalities of the first three components after varimax normalized rotation for the initial solution (with 42 variables) and for the final solution (with the selected 14 variables). Initial solution Final solution

PC1 1 PC 2 PC 3 CM2 PC 1 PC 2 PC 3 CM

Aspect

Colour -0.28 0.01 0.08 0.08

Moisten -0.52 0.13 0.47 0.51 -0.59 0.19 -0.42 0.56 Uniformity -0.35 0.01 -0.07 0.13

Rice amount 0.63 0.01 -0.33 0.50

Rice dimension 0.58 0.20 -0.45 0.58 0.69 0.00 0.14 0.50 Rice cooking -0.05 0.21 -0.43 0.23

Meat amount 0.31 0.17 -0.42 0.30

Meat dimension 0.58 0.13 -0.33 0.46 0.63 0.15 0.20 0.46 Fat amount 0.22 -0.08 0.67 0.50

Fat dimension 0.02 0.01 0.72 0.52 0.14 -0.01 -0.77 0.61 P.3 ingredients 0.21 0.17 -0.43 0.26

Cohesion 0.46 -0.17 0.18 0.27

Odour

Overall Intensity 0.05 0.57 0.09 0.34

Blood -0.06 0.21 0.46 0.26

Meat 0.19 0.10 0.12 0.06

Fat -0.06 0.02 0.65 0.43

Smoke -0.64 0.29 0.13 0.51 -0.66 0.21 0.00 0.48 Cumin 0.23 0.14 -0.66 0.50 0.35 0.13 0.69 0.62 Pepper -0.28 0.32 -0.08 0.19

Wine 0.03 0.36 0.37 0.27

Spearmint 0.28 0.23 -0.59 0.48 0.35 0.11 0.63 0.53 Garlic -0.03 0.52 -0.07 0.28

Clove -0.08 0.42 -0.18 0.22

Flavour

Overall Intensity 0.00 0.73 -0.02 0.53 -0.01 0.86 0.10 0.76 Saltiness -0.09 0.12 0.42 0.20

Rice 0.44 0.12 0.12 0.22

Piquant 0.06 0.56 0.23 0.37

Sweet -0.07 0.32 -0.10 0.12

Spices -0.02 0.62 -0.24 0.45 0.00 0.67 0.38 0.60 Persistence -0.07 0.66 0.20 0.48 -0.14 0.81 -0.15 0.70 Texture

Hardness 0.80 0.01 -0.23 0.70 0.86 0.08 0.09 0.76 Roughness 0.34 0.32 -0.07 0.22

Firmness 0.66 -0.16 0.15 0.49 0.71 0.04 -0.30 0.60 Fibrousness 0.38 0.24 0.13 0.22

Granularity 0.56 0.25 -0.12 0.39

Friability 0.33 0.38 -0.15 0.27

Cohesiveness 0.43 -0.30 0.10 0.29

Juiciness -0.41 0.02 0.33 0.28

Pasty -0.62 0.15 0.44 0.60 -0.70 0.18 -0.28 0.61 Elasticity 0.17 0.04 0.47 0.25

Adhesiveness -0.54 0.37 0.35 0.55 -0.64 0.32 -0.25 0.58 Residue 0.19 0.59 0.08 0.39

Cumulative explained 13.54 23.58 35.40 29.37 44.64 59.69 variance (%) 1 PC - Factor loadings for each principal component, 2 CM - Communality, 3 P.- Proportion between ingredients.

103 CHAPTER 4 ______

Loadings of each variable to each PC after varimax normalized rotation and the communalities from three principal components are presented in Table 4.2. Using this set of variables, only 35.4% of the total variance was explained by the three PCs. In order to determine which aspects were of greatest importance in MA sensory characterization, a reduction in the number of variables was carried out on the basis of higher factor loadings (FL > 0.5) and communalities (CM > 0.4) and on the basis of redundancy between descriptors. Three descriptors with FL > 0.5 and CM > 0.4 were abandoned because they were redundant with other better represented in the PCA analysis: rice amount with rice dimension, fat amount and fat odour with fat dimension. Fourteen descriptors were retained: Aspect: Moisten, Rice dimension, Meat dimension, Fat dimension; Odour: Smoke, Cumin, Spearmint; Flavour: Overall intensity, Spices, Persistence; Texture: Hardness, Firmness, Pasty, Adhesiveness. The number of components retained in the final solution was based on the Kaiser criterion for the analysis of eigenvalues (> 1); the proportion of variance retained was 59.69% (Table 4.2). Figure 4.2 shows the projection of the 14 variables on the three-dimensional space defined by the three PCs. The first and second principal components (Figure 4.2, a) accounted for about 45% of the data variance. The first component is defined by the visual evaluation of the meat and rice dimension, by the hard and firm texture in the right side of the plan, with FL ≥ 0.63. On the opposite side of the plan, the visual evaluation of moisten, the smoke odour and the adhesive and pasty texture are related between them, with absolute FL ≥ 0.59. All the flavour characteristics are related and are represented in the PC2 with FL ≥ 0.67. The PC3, represented with PC1 in Figure 4.2 (b), is defined only by three characteristics, the cumin and spearmint odour in the upper part of the plan, with FL ≥ 0.63, and fat dimension in the lower part (FL = -0.77). These relationships between the sensory attributes are highly related to the specificity of the MA from Monchique, when compared with MA from other origins.

104 CHAPTER 4 ______

1 F-Overall Intensity F -Persistence (a) 0.75 F 3-Spices

0.5

T 4-Adhesiveness 0.25 O 2-Smoke A-Meat Dimension A1-Moisten O-Spearmint T-Pasty T-Hardness O-Cumin 0 T-Firmness A-Fat Dimension A-Rice Dimension

PC2(15.26 %) -0.25

-0.5

-0.75

-1 -1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1 PC1 (29.37 %)

1 (b) 0.75 O -Cumin O-Spearrmint 0.5 F 3-Spices 0.25 A-Meat Dimension F-Overall Intensity A-Rice Dimension O 2-Smoke T-Hardness 0 F 3-Persistence PC3(15.05 %) -0.25 T 4-Adhesiveness T-Firmness T-Pasty A1-Moisten -0.5

-0.75 A-Fat Dimension

-1 -1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1 PC1 (29.37 %)

Figure 4.2 Loadings for the PC1-PC2 (a) and PC1-PC3 (b) dimensions, after varimax normalized rotation, of the 14 variables selected. 1 A- Aspect, 2 O- Odour, 3 F- Flavour, 4 T- Texture.

As it can be seen in Figure 4.3, which represent the scores of the samples in the space defined by the three principal components considered, a clear discrimination between Monchique MA (M) and those from other origins (O), particularly over the

105 CHAPTER 4 ______first and the third PC, can be observed. The Monchique MA samples are located in the right side of the PC1 (Figure 4.3, a), indicating that these samples have a less moisten aspect, the meat and rice dimensions are bigger, and products are harder and firmer than those from the other origins.

3 (a) O 2 O O O O M M M O OO M M O MO O M M M M 1 M M M O MO MO O O O OM M O M MOM O O O O O M O O MM O M O O 0 O O O O M O M O O OO OM O M O O M O M M OM O O OM O MM M M O O OO PC2(15.26 %) O M M O M O M MM O M M M -1 O M O O M O O O O O O O OM O M M M O M -2 O M

-3 -3 -2 -1 0 1 2 3 PC1 (29.37 %)

3 M M (b) M 2 M M M O M M M M M MM M M MMM 1 O O M M O MMM M M O O M M M O O MM M M O O M M O M O O M MM M OOO M M M 0 O O MO O O O M MM M M O OO O O M O OO O O O O O O M PC3(15.05 %) O O O O OOO OMO O O O O -1 O O OM O O O O MOO O OO O O O O O O -2 O O

-3 -3 -2 -1 0 1 2 3 PC1 (29.37 %)

Figure 4.3 Scores of Monchique Morcela de Arroz (M) and Other origin Morcela de arroz (O) on PC1- PC2 (a) and PC1-PC3 (b) dimensions obtained from the 14 variables selected.

106 CHAPTER 4 ______

Considering the variables related to the left side of the PC1, Monchique MA was also less adhesive and pasty, and presented a lower smoke odour. The PC2, which was defined by the flavour characteristics, does not discriminate samples from both origins. On the plan defined by the PC1 and PC3 (Fig. 4.3, b) it is possible to observe that samples from Monchique (M) are located mainly in the upper part of the plan, indicating that these samples have higher cumin and spearmint odour and a fat of reduced dimension. These traits of Monchique MA are related to the specificities of its formulation, namely the use of a high proportion of meat in which approximately half is lean, and the use of only one type of rice Carolino (Japonica subspecies of the Oryza sativa L.), which is a long size grain (above 6 mm), that justifies the visual and texture differences observed. Although Morcela de Arroz formulation includes considerable rice amount (up to 15%) and this grain presents specific flavour traits (Bett-Garber et al., 2012), in the present study the rice flavour showed less relevancy in the data reduction by PCA than other traits. It was probably due to the strong aromatization. Monchique MA does not undergoes a smoking process, it is stored without package in the same storage rooms where smoked meat products are kept, resulting in the acquisition of a very discreet smoke odour. Additionally, the discriminative tendency observed on the PC3 is due to the aromatization of the Monchique MA that is seasoned mainly with cumin and spearmint. All remain spices are used in a much lower proportion. The aromatization with spices is often considered a major distinctive characteristic of meat products, namely in dry-cured sausages (Roncalés, 1995) and in Spanish and Slovenian cooked blood sausages (Santos et al., 2003; Gašperlin et al., 2012). In order to confirm the adequacy of the PCA reduction of descriptors in assigning the origin of MA samples, binary logistic regression was performed, using the origin as the predicted variable and the scores of the observations on the three PCs retained as predictors (Table 4.3). The second principal component was not retained in the logistic regression model because it was not significant, that is coherent with the observation of the projection of the variables on the space defined by PC1 and PC2, as no tendency of discrimination was graphically observed. It was confirmed that PC1 is an excellent predictor of MA origin, as revealed by the high odds ratio, indicating that if 1 unit is increased on the score of samples in PC1, samples are 103.401 times more likely to belong to the Monchique origin. The third principal component is also a good predictor for the origin of the MA, as revealed by

107 CHAPTER 4 ______the high odds ratio (OR = 67.843). This logistic model exhibited a high performance of well-classified events (93%), as indicated by the high area under the curve (AUC = 0.98) determined in the ROC (Receiver Operating Characteristic) plot. As AUC approaches 1.0 it represent a perfect separation of the test values of the two groups, while at 0.5 there are no apparent distributional difference between them (Zweig and Campbell, 1993).

Table 4.3 Results of the logistic regression model for the origin of Morcela de Arroz (Monchique in relation to other origins) considering the scores PC1 and PC3 as predictors. Variables Beta SE1 p Odds ratio 95% CI2 PC1 4.639 1.226 < 0.001 103.401 9.349-1143.614 PC3 4.217 1.054 <.0.001 67.843 8.592-535.708

Log likelihood = 33.093, p < 0.001 1 SE - Standard Error, 2 CI - Confidence Interval.

The use of multivariate analysis to reduce data and segregate samples has been also successfully used, namely to differentiate the type of manufacturing process used to produce chorizo de cebolla (homemade and industrial) (Salgado et al., 2006) or to produce alheira (Patarata et al., 2008), and to identify different varieties of morcilla de Burgos (Santos et al., 2003) and of krvavica (Gašperlin et al., 2012). After defining the descriptors set for characterization of Monchique MA, definitions of descriptors and references to help panellists to anchor their evaluation in the scale were defined (Table 4.4). In order to validate the adequacy of the selected vocabulary, a new set of analysis was performed. Panellists were asked to evaluate the fourteen descriptors, in a 9 point scale (plus zero for the absence of the characteristic) using 2 MA from different origins (Monchique and other) evaluated in duplicate in separate sessions. The sensory characteristics of Monchique MA in comparison to MA from other origins are presented in Table 4.5 and the correspondent sensory profile is illustrated in Figure 4.4. With the selected vocabulary, it is possible to observe that all characteristics of MA aspect, odour and texture are highly discriminative among MA from different origins. The overall intensity of flavour and the spice flavour was also highly discriminative. No differences were observed in the persistence of flavour.

108 CHAPTER 4 ______

Table 4.4 Descriptors selected to characterize Morcela de Arroz (MA), definitions and references used. Intensity scale Descriptor 1 9

Aspect Moisten: Moistness Surface property relating Dry Moisten to the perception of the amount of water Slices of MA dried 3 Slices of freshly cooked days at 80% RH, 10 ºC MA Rice dimension: Visual evaluation of rice size Small Big Cooked broken grain rice Cooked long grain rice Meat dimension: Evaluation of the size of Small Big perceptible meat pieces < 0.3 cm > 2 cm Fat dimension: Evaluation of the size of Small Big perceptible fat pieces < 0.3 cm > 2 cm Odour Smoke: The aromatics associated with wood Low High smoke Slices of not smoked MA Slices of MA smoked for 3h with smoke generated from beech wood scraps Cumin: Characteristic odour associated with Low High this spice Commercial cumin Commercial cumin aroma1 (10-7 g/L) aroma (10-4 g/L) Spearmint: Characteristic odour associated Low High with this spice Commercial spearmint Commercial spearmint aroma2 (10-5 g/L) aroma (10-2 g/L) Flavour Global intensity Low High Intensity of the sum of all ﬂavours Slices of 3 different MA Slices of 3 different MA (combination of the smell, taste and served at 4 ºC served at 37 ºC sensations perceived during tasting) Spices Low High MA spice mix (ground cumin, spearmint, MA spice mixture3 MA spice mix allspice, pepper and clove) suspension in suspension in water suspension in water water (0.002%) (0.2%) Persistence Low High Duration of the sample flavour after < 10 s > 60 s swallowing Texture Hardness Soft Hard The force required to bite through the sample Slices of freshly cooked Slices of MA dried 3 MA days at 80% RH, 10 ºC Firmness Elastic Firm Resistance to deformation when force is Slices of freshly cooked Slices of MA dried 3 applied MA days at 80% RH, 10 ºC Pasty Not pasty Very pasty ‘Mouth coating’ sensation produced by ﬂour- Slices of MA dried 3 Slices of overcooked water paste during the mastication process. days at 80% RH, 10 ºC MA (3 h, boiling) Adhesiveness Not adhesive (sticky) Very adhesive (sticky) The degree to which the surface of the Cooked egg white Cooked rice puree sample adheres to the palate when compressing it with the tongue 1, 2 Cumin aroma (10DR183A), Spearmint aroma (10PRO122), respectively, Proaromática-Aromas Alimentares, Lda (Lisbon, Portugal); 3 Ground cumin, clove, allspice and pepper, mixture supplied by one MA producer.

109 CHAPTER 4 ______

Table 4.5 Sensory attributes of Morcela de Arroz from Monchique and other origins (Mean  Standard deviation). Sensory attribute Monchique Other origins p Aspect Moisten ± ± 4.75 1.38 6.34 1.24 < 0.001 Rice dimension ± ± 6.41 1.35 3.09 1.83 < 0.001 Meat dimension ± ± 6.07 1.11 3.02 1.27 < 0.001 Fat dimension ± ± 3.27 1.23 4.98 1.82 < 0.001 Odour Smoke ± ± 2.64 1.85 5.16 1.39 < 0.001 Cumin ± ± 5.73 1.60 3.25 1.45 < 0.001 Spearmint ± ± 4.43 1.74 1.88 1.15 < 0.001 Flavour Overall Intensity ± ± 5.66 1.46 4.23 1.58 < 0.001 Spices ± ± 4.45 1.50 3.14 1.53 < 0.001 Persistence ± ± 5.45 1.56 5.86 1.09 0.157 Texture Hardness ± ± 6.04 1.15 3.86 1.37 < 0.001 Firmness ± ± 6.14 1.09 5.32 1.69 0.008 Pasty ± ± 2.81 1.31 4.38 1.80 < 0.001 Adhesiveness ± ± 2.41 1.42 3.54 1.69 0.001

A-Moisten 10.0 A-Rice T-Adhesivennes Dimension A-Meat T-Pasty Dimension 5.0

A-Fat T-Firmness Dimension 0.0

T-Hardness O-Smoke

F-Persistense O-Cumin

F-Spices O-Spearmint F-Overall Intensity Monchique Other origins

Figure 4.4 Sensory profile of Morcela de Arroz from Monchique (full line) and other origins (dashed line). A- Aspect, O- Odour, F- Flavour, T- Texture.

110 CHAPTER 4 ______

These results confirm the adequacy of the selected vocabulary to discriminate between Monchique MA from those from other provenances. The discrimination observed in this validation was better than that conducing by the fourteen descriptors set. The introduction of references for the selected descriptors and the use of a wider scale increased the performance of the panellists.

4.6 CONCLUSIONS

The present study enabled us to establish the lexicon for describing the sensory characteristics of Monchique Morcela de Arroz. With the selected descriptors, it is possible to observe that Monchique MA has a less moisten surface, the pieces of meat and the rice are bigger than those from other origins. On the contrary, the size of fat pieces is smaller. Monchique MA has a weaker smoke odour, but presents a clear higher odour note to cumin and spearmint, which can be assumed as a distinctive trait. The overall flavour and the spices flavours are slightly higher. Due to the amount of meat, and to the size of its pieces, Monchique MA is harder and firmer, less pasty and adhesive.

4.7 ACKNOWLEDGEMENTS

This work received financial support from FCT-Portuguese Foundation for Science and Technology through PhD grant SFRH/BD/63341/2009. Authors gratefully acknowledge all members of the sensory panel for their participation.

4.8 REFERENCES

111 CHAPTER 4 ______

Bett-Garber, K. L., Lea, J. M., Champagne, E. T. and Mcclung, A. M. (2012). Whole- grain rice flavour associated with assorted bran colours. Journal of Sensory Studies, 27, 78-86.

Bojanowski, V. and Hummel, T. (2012). Retronasal perception of odours. Physiology & Behavior, 107, 484-487.

Braghieri, A., Piazzolla, N., Carlucci, A., Monteleone, E., Girolami, A. and Napolitano, F. (2012). Development and validation of a quantitative frame of reference for meat sensory evaluation. Food Quality and Preference, 25, 63-68.

Cherdchu, P., Chambers IV, E. and Suwonsichon, T. (2013). Sensory lexicon development using trained panelists in Thailand and the U.S.A.: soy sauce. Journal of Sensory Studies, 28, 248-255.

European Regulation No 510 (2006). Council Regulation of 20th March 2006. Official Journal of the European Union.2006, L93, 12-25.

Holley, A. C. (2006). Processing information about flavour. In Flavour in food. Voilley, A. and Etievant, P. (Eds.), 36-61, Woodhead Publishing Limited, Cambridge, U.K.

Hongsoongnern, P. and Chambers IV, E. (2008). A lexicon for texture and flavour characteristics of fresh and processed tomatoes. Journal of Sensory Studies, 23, 583-599.

International Organization for Standardization (ISO) 4121 (2003). Sensory analysis. Guidelines for the use of quantitative response scales, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 8589 (2007). Sensory analysis. General guidance for the design of test rooms, International Organization for Standardization, Geneva, Switzerland.

112 CHAPTER 4 ______

Jiménez, M. J., Canet, W. and Alvarez, M. D. (2013). Sensory description of potato puree enriched with individual functional ingredients and their blends. Journal of Texture Studies, 44, 301-316.

Lawless, L. J. R. and Civille, G. V. (2013). Developing lexicons: a review. Journal of Sensory Studies, 28, 270-281.

Morais, E. C., Cruz, A. G., Faria, J. A. F. and Bolini, H. M. A. (2014). Prebiotic gluten-free bread: sensory profiling and drivers of liking. LWT-Food Science and Technology, 55, 248-254.

Moskowitz, H. R. (1983). Product testing and sensory evaluation of foods: marketing and R&D approaches. Food and Nutrition Press, Westport, Connecticut, U.S.A.

Negoias, S., Visschers, R., Boelrijk, A. and Hummel, T. (2008). New ways to understand aroma perception. Food Chemistry, 108, 1247-1254.

N’Kouka, K. D., Klein, B. P. and Lee, S.-Y. (2004). Developing a lexicon for descriptive analysis of soymilks. Journal of Food Science, 69, 259-263.

113 CHAPTER 4 ______

Pérez-Cacho, M. P. R., Galán-Soldevilla, H., Crespo, F. L. and Recio, G. M. (2005). Determination of the sensory attributes of a Spanish dry-cured sausage. Meat Science, 71, 620-633.

Roncalés, P. (1995). Tecnología, cambios bioquímicos y calidad sensorial de los embutidos curados. Alimentación, Equipos y Tecnología, (Enero/Febrero), 73-82.

Salgado, A., Fontán, M., Franco, I., López, M. and Carballo, J. (2006). Effect of the type of manufacture (homemade or industrial) on the biochemical characteristics of Chorizo de cebolla (a Spanish traditional sausage). Food Control, 7, 213-221.

Smyth, H. E., Sanderson, J. E. and Sultanbawa, Y. (2012). Lexicon for the sensory description of Australian native food plants and ingredients. Journal of Sensory Studies, 27, 471-481.

Stone, H., Sidel, J., Oliver, S., Woolsey, A. and Singleton, R. C. (1974). Sensory evaluation by quantitative descriptive analysis. Food Technology, 28, 24-34.

Vázquez-Araújo, L., Chambers IV, E. Adhikari, K., Hough, G. and Carbonell- Barrachina, A. A. (2013). Influence of various traditional seasonings on beef flavour: United States, Spanish, and Argentinian practices. Meat Science, 93, 61-66.

Zeppa, G. Bertolino, M. and Rolle, L. (2012). Quantitative descriptive analysis of Italian polenta produced with different corn cultivars. Journal of Science and Food Agriculture, 92,412-417.

Zweig, M. H. and Campbell, G. (1993). Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry, 39, 561-577.

114

CHAPTER 5

Pereira, J. A., Dionísio, L., Patarata, L. and Matos, T. J. S. Effect of packaging technology on microbiological and sensory quality of a cooked blood sausage, Morcela de Arroz, from Monchique region of Portugal. Submitted for publication in Meat Science Journal

115

CHAPTER 5 ______

5. EFFECT OF PACKAGING TECHNOLOGY ON MICROBIOLOGICAL AND SENSORY

QUALITY OF A COOKED BLOOD SAUSAGE, MORCELA DE ARROZ, FROM MONCHIQUE

REGION OF PORTUGAL

Pereira, J. A.a,b, Dionísio, L.c,d, Patarata, L.e and Matos, T. J. S.b,f

aDepartment of Food Engineering, Instituto Superior de Engenharia, University of Algarve, Campus da Penha, Estrada da Penha, 8005-139 Faro, Portugal bCEER - Biosystems Engineering, Tapada da Ajuda, 1349-017 Lisbon, Portugal cDepartment of Biological Sciences and Bioengineering, Faculdade de Ciências e Tecnologia, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal dICAAM - Institute of Mediterranean Agricultural and Environmental Sciences, P.O. box 94, University of Évora, 702-554 Évora, Portugal eCECAV - Centro de Ciência Animal e Veterinária, Universidade de Trás-os-Montes e Alto Douro, Quinta de Prados, 5000-801 Vila Real, Portugal fDepartment of Science and Biosystems Engineering, Instituto Superior de Agronomia, Technical University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal

5.1 ABSTRACT

Morcela de Arroz (MA), a popular Portuguese blood sausage, with high pH and water activity (aw), is traditionally commercialized without preservatives and unpacked. This study evaluated the best packaging solution to extend MA shelf life stored at 4 ± 1 ºC for 44 days: without packaging (WP), vacuum (VP) and modified atmosphere packaging (MAP) (80% CO2; 20% N2). Mesophilic (MTVC), psychrotrophic (PTVC), lactic acid bacteria (LAB), pseudomonads, moulds and yeasts, Enterobacteriaceae, Listeria monocytogenes, Salmonella spp., Bacillus cereus,

Clostridium perfringens, sensory properties, pH, moisture and aw were studied.

Moisture and aw decreased (p < 0.05) in WP. pH decreased in WP and MAP during storage. MTVC and PTVC counts increased to values around 7 log CFU/g at 44 days of storage. LAB and Enterobacteriaceae counts were higher (p < 0.05) in VP. Pseudomonads were inhibited (p < 0.05) by MAP after 8 days of storage. Sensory parameters were affected (p < 0.05) by packaging and storage time. Globally, MAP performed better.

117

CHAPTER 5 ______

5.2 INTRODUCTION

Morcela de arroz (MA) is a popular cooked rice blood sausage, a ready-to-eat (RTE) meat product that is traditionally produced in Serra de Monchique (south of Portugal). This product supports the economy of several industries, due to its uniqueness among the traditional Portuguese meat products. MA is produced without addition of chemical preservatives and without the use of any packaging technology. Distribution and storage are performed in a very short period (one week at 0-5 ºC). The short shelf life has been an important hurdle to the increase of the production through the diversification of the markets, namely more distant ones. To our knowledge, there is no information about MA, namely its microbiological, sensory or physic-chemical characteristics, that hinders any attempt to modify the commercialization conditions. Despite the information available about similar blood sausages (Roseiro et al., 1998; Santos et al., 2003; Cachaldora et al., 2013), the specificity of MA makes its shelf life difficult to evaluate by comparison or extrapolation. Blood sausages are products of difficult preservation due to the rich nutrient composition, high pH (about 6.0) and high aw reaching values higher than 0.97. Although, the high temperature and cooking time (95-96 ºC for around 1 h) during processing eliminates the vegetative forms of microorganisms and only more heat resistant bacteria spores survive the treatment (Silva and Gibbs, 2010), post-cooking contamination on the surface of the product is caused by handling of the product during cooling, refrigeration and post processing operations (Mäkelä and Korkeala, 1987; Oteiza et al., 2006; Santos et al., 2005a). The use of packaging technology can extend shelf life of meat products by improving safety and stability along shelf life. Vacuum packaging (VP), being cost- effective and easy to apply (McMillin, 2008), has been so far the most widely used packaging technique for different type of meat products (Borch, Kant-Muermans and Blixt, 1996). However, it presents some disadvantages, such as product deformation and exudate forming. To overcome this problem, modified atmosphere packaging (MAP) has become popular among meat products (Hintlian and Hotchkiss, 1987; Pexara, Metaxopoulos and Drosinos, 2002; Matos, Barreto and Bernardo, 2005).

CO2/N2 atmospheres are accepted as particularly adequate for the preservation of meat 118

CHAPTER 5 ______

products, mainly because of the strong growth inhibition of the majority of microorganisms. It is accepted that CO2 increases the lag phase and generation time of microorganisms, even those that can growth in the absence of oxygen. This inhibitory effect is due to the reduction of pH, inhibition of certain enzymes involved in energy production and damage of the cell membrane. N2 is used as a filler gas and it has no recognized inhibitory effect. The proportion of CO2/N2 in MAP is variable with the product and the desired shelf life (Mullan and McDowell, 2003). Among the several studies made with blood sausages shelf life, Santos et al. (2005a) concluded that the combination of 80% CO2 and 20% N2 resulted in an improved control of spoilage microbiota and shelf life extension. The aim of the present study was to evaluate the microbiological, physicochemical and sensory characteristics of Morcela de Arroz packaged in VP,

MAP (80% CO2 and 20% N2) and without packaging (WP) stored at 4 ± 1 ºC in order to investigate the best packaging solution to extend the shelf life period.

5.3 MATERIALS AND METHODS

5.3.1 Formulation and processing of the MA

MA was processed in a local industry according to the traditional method. Formulation included Alentejano pig breed meat (loin, shoulder, neck and bacon) (54%); rice (13%), pork blood (6%), water (17%), lard (4%), grinded onion (3%), salt (1%) and spices (ground cumin, ground clove, ground pepper and mint) (2%). Meats were minced in plates with holes of 2 cm diameter and mixed with raw rice, pork blood, water, grinded onion and mint, lard, salt and ground spices (Pereira, Ferreira- Dias and Matos, 2008). The mixture was stuffed into 60-65 mm diameter natural beef casings. MA were then transferred into a stainless steel cooking vessel and cooked in molho de moura, a broth prepared with water, whole onions, whole garlic, salt and laurel leaves at 91 ± 5 ºC for 1.5 h. After cooking (individual weight of each MA around 250 g), products were cooled at room temperature for about 2 hours and then transferred into a refrigerated chamber (4 ± 1 ºC) until the next day (day 0). The experiment was conducted with three production batches separated in time by 2 months. 119

CHAPTER 5 ______

5.3.2 Packaging of the MA

On day 0, 45 samples for each packaging technology studied were individual packed in the producer facility: vacuum packed (VP) in a Belca V-40S packaging machine (Belca, Villabona, Spain) and MAP (80% CO2/20% N2) in a Haratek 200A packaging machine (Haratek, Astigarraga, Spain) both in oxygen and water impermeable bags Combitherm XX (dimensions of 230 x 286 mm), 115 µm thickness (Wipak Walsrode GmbH & Co.KG, Walsrode, Germany), with an oxygen transmission rate of 3.0 cm3/m2.24 h.atm at 23 ºC and 75% RH, carbon dioxide transmission rate of 13.5 cm3/m2.24 h.atm at 23 ºC and 75% RH and a water vapor transmission rate of 1.0 g/m2.24 h at 23 ºC and 85% RH. Samples to be studied without package (WP) were transported to the laboratory in sterile sample bags. All samples were transported under refrigeration to the laboratory and stored at 4 ± 1 ºC and 80% of RH (Fitoclima s 600 PLH, Aralab, Lisbon, Portugal). Selection of 80% of RH was based on the results of RH measurement in two industries and two retailer’s refrigerators. WP samples were taken out from the sterile bags and stored without any package, to simulate the usual practice of commercialization of MA. Nine extra sausages were collected from each batch and transported in the same conditions (under refrigeration in sterile sample bags) to the analysis of the first sampling time, and analyzed immediately on arrival to the laboratory.

5.3.3 Sampling procedure

One experimental unit, thereafter designated as repetition, was composed by two MA, one to perform microbiological and physicochemical analysis and the other for sensory analysis, once only one MA was not enough for all analysis due to the high consumption of sample by the eighteen member of sensory analysis panel. The experiment was conducted with 9 repetitions for each experimental condition (packaging x storage time). These 9 repetitions were obtained from 3 producing batches, separated in time by 2 months, collecting 3 samples from each of the 3 batches. Thus, for each storage time, 27 samples were analyzed [3 packaging (WP/VP/MAP) x 3 production batches x 3 repetitions from each batch]. In the first 120

CHAPTER 5 ______

storage time (day 0) only 9 samples were analyzed, once there was no packaging yet (3 production batches x 3 repetitions in each batch). During storage period samples were collected at days 8, 15, 22, 30 and 44 for microbiological, physicochemical and sensory analysis. At first sampling time (day 0) samples not packaged were analyzed immediately after arrival to the laboratory. Extra MA packed in each condition were maintained for all the experimental conditions to be used in case of package damage and vacuum or MAP loss. Samples used for sensory analysis panel selection, for training and for lexicon development are indicated in the respective sections (5.3.6.1.1 and 5.3.6.1.2).

5.3.4 Physic-chemical analysis

5.3.4.1 Analysis of gas composition

Oxygen and carbon dioxide concentrations in the head space of the MAP packages were measured with a digital O2 and CO2 analyzer (OXYBABY V O2/CO2 WITT-Gasetechnik, Witten, Germany). 6 ml gas samples were drawn by the needle of the analyzer through a rubber seal glued on the surface of the pack. Detection limits for O2 and

CO2 were ± 0.1% and ± 2%, respectively.

5.3.4.2 pH, moisture and aw parameters

The potentiometric measurements of pH were carried out by inserting the pin electrode of the pH meter (GLP 21, Crison, Barcelona, Spain) directly into the sample

(20.5 ± 0.3 ºC). Prior to the moisture and aw measurements, each sample was ground and homogenized in a domestic food processor. Moisture was determined using a Mettler LP16-M infrared scale (Mettler Instruments AG, Zurich, Switzerland). Rotronic Hygroskop DT was employed to measure the water activity at 25.0 ± 0.01 ºC (Rotronic, HygroLab, Bassersdorf, Switzerland). Determinations were made in triplicate.

121

CHAPTER 5 ______

5.3.5 Microbiological analysis

Twenty-five grams slices of MA were taken aseptically and homogenized with 225 ml of sterile Ringer´s solution (Merck, Darmstadt, Germany) for 90 sec in sterile plastic bag in a lab blender (Masticator, IUL Instruments, Barcelona, Spain). Serial decimal dilutions in sterile Ringer´s solution were prepared and 1 ml or 0.1 ml samples of appropriate dilutions were poured or spread onto nutrient agar and selective agar plates. Mesophilic total viable count (MTVC) was determined on Plate Count Agar (PCA, Merck), incubated at 30 ºC for 72 h; psychrotrophics (PTVC) were determined on PCA, incubated at 6.5 ºC for 10 days; lactic acid bacteria (LAB) on Man Rogosa and Sharpe (Merck) and incubated on anaerobic conditions at 30 ºC for 2-3 days; Enterobacteriaceae on Violet Red Bile Glucose Agar (Merck), incubated at 37 ºC for 24 h; Pseudomonas spp. on Pseudomonads Agar Base (Oxoid, Basingstoke, UK) with Cetrimide, Fucidin and Cephalosporin supplement, incubated at 25 ºC for 48 h, molds and yeasts on Chloramphenicol Rose Bengal (Merck) and incubated at 25 ºC for 5 days; detection of Salmonella spp. was performed according to EN 12824 (1997) and detection of Listeria monocytogenes according to ISO 11290-1 (1996); Bacillus cereus according to ISO 7932 (2004) and Clostridium perfringens as described in ISO 7937 (2004). The results are presented as log CFU/g. For statistical purposes, when the microorganism count was below the detection limit, it was considered to be zero and when countable colonies were present, but below the countable range, the estimated count was considered for data analysis.

5.3.6 Sensory analysis

5.3.6.1 Quantitative descriptive Analysis (QDA)

5.3.6.1.1 Formation of the panel

A panel of 18 assessors was selected from an initial group of 63 recruited individuals (internal recruitment) and subsequently trained how to carry out the necessary analysis. Selection of a shortlist of 38 individuals was based on their 122

CHAPTER 5 ______

capacity to detect and identify different tastes and aromas, ability to memorize sensory attributes, discriminative capacity and accuracy of their vocabulary in describing sensory parameters. Selection of the final 18 assessors was based on their discriminative capacity and repeatability calculated using the data obtained in repeated quantitative descriptive analyses of nine samples of MA in three sessions held at 1- week intervals (ISO 8586-1, 1993). Before performing the evaluation of the samples of the present work, the panel was trained during 10 sessions of 45-60 min. The panelists performed sensory analysis on a total of 27 samples of MA from 12 different origins (2 or 3 samples in each session).

5.3.6.1.2 Generation of vocabulary

Descriptors for the QDA were generated in five stages according to ISO 11035 (1994): 1- sessions of free and open generation of vocabulary, where assessors used the descriptors they considered more correct for describing the fresh and aged MA with different storage times, correspondent to different levels of spoilage; 2- synthesis of vocabulary, determination of frequency of use and elimination of incorrectly used descriptors; 3- test of an initial list of nineteen descriptors to detect redundant descriptors and selection of the most discriminative descriptors between fresh and not fresh product; 5- test of the final list. These stages were organized over nine sessions. For the final list of 16 descriptors, definitions and references were established, proposed by the panel leader, and discussed with the panelists. It was used as references the freshly cooked MA and unpackaged refrigerated MA stored for 7 to 44 days to achieve the desired sensory parameter intensity. The choice of using the product under study rather than other matrixes as a reference was considered to be more useful to generalize sensory concepts, as suggested by Braghieri et al. (2012). The descriptors retained in the final list to evaluate shelf life of the MA were: Aspect- moisture: moistness surface property relating to the perception of the amount of water; slime: presence of viscous substance in the surface; typical color: characteristic dark brown color; brightness: light reflection due to melting of fat, greasy surface; Aroma- typical odor: characteristic meaty and spicy odor; off-odors: 123

CHAPTER 5 ______

any atypical odor considered not expected in the product; rancid: associated with oxidized fat and oils, putrid: the pungent ammonic like odor in advanced spoiled meat or fish; sour/fermented: pungent odor associated to spoiled cooked rice; spoiled wine/vinegary: acetic odor associated to vinegar; moldy/mildew: characteristic odor of moulds; Flavour- typical flavour: characteristic meaty, salted and slightly sweet flavour; acid: taste like vinegar or lemon juice; bitter: acrid and harsh taste, as the peel of an orange or coffee dregs; global evaluation of freshness: multifactorial evaluation of freshness, by opposition to spoilage. Additionally, a freshness global evaluation of the entire packed product (external appearance) was introduced. Assessors were asked to score each descriptor in each sample on a scale from 1 (low intensity) to 5 (intense).

5.3.6.1.3 Evaluation of the samples

Sessions took place on a sensory analysis laboratory equipped with individual assessment booths and uniform lightening and temperature conditions. To guarantee test repeatability, samples were cut in 1 cm thick slices and placed in Petri dishes codified with a randomized three digit number and evaluated at room temperature. Spring water and unsalted crackers were provided to clean the mouth between samplings.

5.3.7 Statistical analysis

The experiment was composed by a total of 144 observations used for statistical analysis [9 samples for day 0 (3 batches x 3 repetitions) + 135 samples (5 storage times x 3 packaging x 3 production batches x 3 repetitions from each batch]. The analysis of variance was performed with the packaging or storage time as individual effects on the microbiological and sensory variables by one-way ANOVA procedures. When the effect of storage time was analyzed by ANOVA, data from sampling time 0 (day 0) was used for the three packaging conditions. The differences between the mean values were determined using the Tukey test. The relationship between variables was established through Pearson correlation.

124

CHAPTER 5 ______

The chi-squared test was applied to evaluate the relationship between the presence/absence of Bacillus cereus in MA and (1) the packaging technology in each sampling time and (2) the storage period in each package. Statistical significance was set at p values less than 0.05. Data analysis was performed using the IBM SPSS Statistics for Windows, Version 19.0 (IBM Corp., 2010).

5.4 RESULTS

5.4.1 Physic-chemical analysis

The CO2 concentration of the MAP samples decreased by 19% during the storage period. The initial moisture content was 52.49 ± 2.08% (Figure 5.1) and significantly

(p < 0.05) decreased in the WP samples during storage time (20% loss). The initial aw was 0.966 ± 0.005 (Figure 5.2) and followed a similar trend to that observed for moisture (decreased nearly 0.02 units until 44 days).

A 55 F A b b b b b

50 b b b b b 45 E WP a D 40 VP a Moisture (%) Moisture C a MAP 35 B A a a 30 0 8 15 22 30 44 Storage time (days)

Figure 5.1 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on moisture content (%). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-F, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

125

CHAPTER 5 ______

B 0.98 AB A AB b AB a b b D ABAB b AB AB B 0.97 A a b b b A b 0.96 CD a BC B w

a a BC 0.95 a a A WP a 0.94 VP MAP 0.93 0 8 15 22 30 44 Storage time (days) Figure 5.2 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on water activity (aw). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

At day 0, the pH of MA was 6.36 ± 0.06 (Figure 5.3). During most part of the storage time, the MAP MA presented a significantly (p < 0.05) lower pH comparatively to WP and VP samples, reaching the lowest values at day 22 of the study (pH = 6.19 ± 0.06).

6.50

C A C C A A b A A b a b b A A b BC AB A a 6.25 b

pH B b b B a AB A a A a a WP a VP MAP 6.00 0 8 15 22 30 44 Storage time (days)

Figure 5.3 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on pH. Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-C, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

126

CHAPTER 5 ______

5.4.2 Microbiological analysis

Clostridium perfringens, Listeria monocytogenes and Salmonella spp. we’re not detected in any of the samples tested during all the experiment in any of the package technologies used. Bacillus cereus (results not shown) was not detected in MA freshly prepared (day 0). At 8 days of storage and thereafter it was punctually detected in the three packaging conditions, in a total of 18 samples, without any significant (p > 0.05) association either to packaging or storage time, except at 22 days of storage when all the samples with B. cereus were stored WP (2 = 9.39, p = 0.09). The results of the counts of total viable count at 30 ºC (MTVC) and at 6.5 ºC (PTVC), LAB, Enterobacteriaceae and Pseudomonas spp. are presented in Figures 5.4, 5.5, 5.6, 5.7 and 5.8, respectively. During the storage of MA the MTVC and PTVC increased concomitantly from a count of 2.42 ± 0.69 log CFU/g and 1.69 ± 1.06 log CFU/g, respectively, to values around 7 log CFU/g at the end of the storage period (44 days). The increase in the TVC counts was about 1 log CFU/g in each storage interval studied. Generally both TVC were not affected by the packaging methodology. Only at 22 days of storage a higher (p < 0.05) count of mesophilic was observed in samples stored without package than those packaged in MAP.

9 CD D 8 C a a 7 b CD D B C a a a 6 AB ab C B C a a 5 a a AB 4 A a B B WP Log CFU/g Log a 3 a A VP 2 a MAP 1 0 0 8 15 22 30 44 Storage time (days)

Figure 5.4 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20 %N2) and of storage time (44 days) at 4 ± 1 ºC on mesophilic total viable counts (MTVC) (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

127

CHAPTER 5 ______

9 C C 8 C a a BC a 7 CD D a BCD AB a a 6 ABC a a 5 a AB D 4 a CD a BC Log CFU/g Log A a 3 BC WP a a 2 AB VP a 1 MAP 0 0 8 15 22 30 44 Storage time (days)

Figure 5.5 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on psychrotrophic total viable counts (PTVC) (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

For the entire storage time of the study the counts of LAB (Figure 5.6) were always higher (p < 0.05) in the VP samples than in WP and MAP, except in the day 15 of storage. The counts of LAB increased faster in the VP MA, as inferred by the significant differences (p < 0.05) between the counts at the beginning of the experiment and after 8 days of storage. The higher increase of LAB occurred between 8 and 15 days of storage, nearly 1.5 log CFU/g.

9 8 CD 7 C b D CD 6 a C B b b 5 b B a C B a C a 4 a ab B B 3 C B a Log CFU/g Log a a a WP 2 A A VP 1 a MAP 0 0 8 15 22 30 44 Stoarge time (days)

Figure 5.6 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on lactic acid bacteria (LAB) (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-b, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-D, means with different letter are different (same packaging technology along time of storage) (p < 0.05). 128

CHAPTER 5 ______

The count of Enterobacteriaceae (Figure 5.7) was generally lower in MAP samples. At the beginning of the experiment, any Enterobacteriaeceae colony was detected. During storage it presented an increase, that was always lower in samples packaged in MAP, with significant differences (p < 0.05) at 8, 22, 30 and 44 days of storage comparatively to VP samples and at days 8, 22 and 30 comparatively to the WP samples.

9 8 7 6 C C BC c 5 b AB a B C B 4 b b b AB b WP Log CFU/g Log 3 AB a b AB VP 2 a A A A MAP A A 1 A a a a a a 0 0 8 15 22 30 44 Storage time (days)

Figure 5.7 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on Enterobacteriaceae (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-c, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-C, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

Pseudomonas spp. (Figure 5.8) were not detected at day 0 of the study, although, in the WP samples their counts significantly increased (p < 0.05) along the storage period, achieving about 1.5 log CFU/g on day 8 of the study and reaching a plateau of 4.5 log CFU/g at 22 days of storage, after what their counts remained similar until the end of the storage period but were always significantly higher (p < 0.05) comparatively to the packaged MA. In the packaged MA, Pseudomonas spp. showed a growth rate that was similar both on VP and MAP only until 8 days of storage (p > 0.05). After 8 days of storage on, the MAP MA presented significantly lower (p < 0.05) counts of Pseudomonas spp. comparatively to the VP MA. The growth of Pseudomonads in

129

CHAPTER 5 ______

MAP was not significantly (p > 0.05) affected by the storage time. The count of moulds and yeasts was below the detection limit in all samples until 15 days of storage. At day 22 and thereafter these microorganisms were punctually detected, 2 samples in day 22, 4 in day 30 and 2 in day 44. No association was established between the occurrence of moulds and yeasts and the package technology used.

9 8 7 6 B b B C C C 5 b c c c 4 A a B

Log CFU/g Log 3 b B B WP A B b b VP 2 a b A MAP 1 A A A A a a a 0 a 0 8 15 22 30 44 Storage time (days)

Figure 5.8 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on Pseudomonas spp. (Log CFU/g). Results are presented as mean ± sd (n = 9 for each packaging/storage time). a-c, means with different letter are different (between packaging technologies at each day of storage) (p < 0.05). A-C, means with different letter are different (same packaging technology along time of storage) (p < 0.05).

5.4.3 Sensory analysis

The results of the sensory analysis of WP, VP and MAP MA during 44 days on refrigerated storage (4 ± 1 ºC) are presented in Table 5.1. Sensory parameters were significantly (p < 0.05) affected by the method of packaging and storage time. In general, the sensory quality decreased during storage time. MA stored in MAP presented higher mean scores than VP and this higher than WP samples. In the freshly prepared MA, no off-odours, rancid, putrid, sour/fermented, spoiled wine/vinegary, mouldy/mildew and putrid odours were detected. The WP presented lower (p < 0.05)

130

CHAPTER 5 ______

moisture aspect, typical colour and typical odour and flavour at 8 days of storage and thereafter than VP and MAP MA. Presence of off-odours and rancid odour began to be significantly higher (p < 0.05) in the WP samples after 8 days of storage. In the WP samples the rancid odour at 44 days of storage was threefold higher than in freshly prepared MA. At 8 days of storage, WP samples began to be considered less (p < 0.05) fresh than those packaged, regardless the evaluation made in slices or in packaged MA. The development of the depreciative characteristics: slime, sour/fermented, spoiled wine/vinegary and mouldy/mildew odour and acid and bitter tastes were slower than the rancid and off-odours, as they became significantly different in WP samples only after 22 days of storage. From the sensory point of view, VP and MAP had a similar performance for the most of the tested attributes. However, MAP samples were generally more bright (p < 0.05 after day 8), and, for longer storage periods (22 days and after), it presented a higher (p < 0.05) typical flavour and global freshness (slices or packaged). The evaluation of global freshness in slices or in entire packaged MAP MA during all the experiment reveals that these characteristics were stable during the tested period. The relationship between microbial counts and sensory characteristics was established by Pearson correlations for each packaging technology.

131

CHAPTER 5 ______

Table 5.1 Effect of packaging technology (without packaging -WP, vacuum packaging -VP and modified atmosphere packaging -MAP, 80% CO2 and 20% N2) and of storage time (44 days) at 4 ± 1 ºC on sensory parameters of Morcela de Arroz (MA). Results are presented as mean ± standard deviation (n = 9 for each packaging/storage time).

1-The first sampling time (day 0) was made before packaging. Mean values were repeated for the three packages only for comparison between storage times. 2- For global freshness of packaged product, in the day 0 of the study assessors evaluated only the unpackaged product. 3- Global freshness of the entire packed product (external appearance). a-c - means with different lowercase letter in the same column/parameter are significantly different (p < 0.05). A-D- means with different uppercase letter in the same row are significantly different (p < 0.05).

132

CHAPTER 5 ______

Considering only the global freshness (Table 5.2), it was observed high significant correlations between total viable counts (MTVC and PTVC) and the sensory evaluation of freshness. These correlations were higher in WP samples. When the correlations were established between the specific spoilage microorganisms and the global freshness, it was possible to observe that LAB were particularly important in the deterioration of VP and MAP, Enterobacteriaceae in VP, and Pseudomonas spp. in WP samples, as can be observed by the higher correlation coefficient.

Table 5.2 Selected Pearson correlations between microbial counts and sensorial evaluation of global freshness (r, p). Microbial group WP VP MAP

MTVC -0.86, < 0.001 -0.72, 0.004 -0.72, 0.004 PTVC -0.83, < 0.001 -0,73, 0.003 -0,77, 0.001 LAB -0.74, 0.002 -0.85, < 0.001 -0.88, < 0.001 Enterobacteriaceae -0.49, ns -0.76, 0.002 -0.33, ns Pseudomonas spp. -0.90, < 0.001 -0.74, < 0.001 -0.33, ns

5.5 DISCUSSION

The results for the mean initial pH, aw and moisture content were similar to those found by Santos et al. (2003) in a Spanish blood sausage which confirm that MA is a RTE meat product of difficult preservation. The significant decrease in pH values in the MAP samples comparatively to the

VP and WP was probably due to the solubilisation of CO2 in the product, once the conditions were favourable for that, namely the high CO2 content of the packages and the low temperature used in the experiment (Mullan and McDowell, 2003). Additionally, no correlation was observed between pH evolution and LAB or other microbial group counts for any of the packaged methods, indicating that the pH evolution was independent of the microbial growth. The traditional method of distribution and commercialization of MA without package (WP) was responsible for a considerable moisture loss. The drying effect

133

CHAPTER 5 ______

during cold storage has several consequences, beyond the direct economic ones due to weight loss. The reduction of aw resulting from the drying of MA was small, when related to the total moisture loss (around 20%). That unexpected relationship might be a consequence of the composition of MA, namely the high level of rice and blood. The binding of water to these components of MA is expected to be weaker than the binding in sausages made only with meat and fat, once carbohydrates have a lower water binding capacity than muscular proteins (Le Maguer, 1987), and the effect of blood on the water holding capacity is variable (Jarmoluk and Pietrasik, 2003). Despite the general faster deterioration detected by the sensory panel, WP samples did not have a clearly worst microbiological profile, as it might be expected. The reduction of aw due to the drying effect produced a delay in the growth of the microbiota. Despite the small absolute reduction in the aw during the experiment (0.02 units of aw), it occurred at an interval that began to have inhibitory effect on microbiota, particularly on Gram- negative bacteria (Adams and Moss, 2008). The presence of spore formers like B. cereus was probably associated to the contamination of the raw materials and/or to the insufficient heat treatment to inactivate the spores. Nevertheless it does not compromise the safety of MA, once the infective dose for diarrheal or emetic syndrome is clearly higher (Arnesen, Fagerlund and Granum, 2008) than the counts observed in MA. The post cooking contamination was probably the main responsible for the MTVC, PTVC and LAB counts observed immediately after processing. Even for thermoduric non-spore former gram positive bacteria, the heat treatment associated to the cooking of MA was very intense to allow their survival. Once MA was produced at one industry that produces dry-cured meat products, it is acceptable to consider that the contamination of MA was made from the house flora of the industry, composed mainly by Gram positive bacteria, particularly LAB (Mäkelä and Korkeala, 1987). The total viable counts might be also composed by spore formers that survived to the heat treatment (Korkeala et al., 1985). The behaviour of total viable counts (MTVC and PTVC) was similar along the experiment, without a clear pattern of differences between MA packaged with different technologies. These results might be justified by the qualitative composition of these total counts. Once the different spoilage microorganisms have particular behaviours in each packaging technology, and it is expected to have a substitution effect when one group is inhibited, the similar MTVC and PTVC were composed by the specific dominant spoilage group in each package. 134

CHAPTER 5 ______

LAB were the dominant spoilage microbiota in VP samples as they usually are in VP meat products (Korkeala and Björkroth, 1997). The microaerophilic needs justifies the ability to grow well in VP, once they can use trace amounts of oxygen dissolved in the VP product (Buchanan, 1986; Korkeala and Björkroth, 1997). MAP and WP samples had similar LAB counts. The reduction of growth in MAP MA can be attributed to the direct effect of carbon dioxide, once it has signiﬁcant effect on the extension of the lag phase and generation time of certain LAB (Devlieghere, Debevere and Van Impe, 1998). That was confirmed by the results of Santos et al. (2005a), who found an inhibitory effect of MAP (80 % CO2 / 20% N2) in the growth of LAB comparatively to VP. However, other authors (Samelis and Georgiadou, 2000; Pexara, Metaxopoulos and Drosinos, 2002; Cachaldora et al., 2013) reported that the use of MAP with different amounts of carbon dioxide in cooked meat products did not affect the growth rate of LAB when compared to the use of VP. That apparent controversy might be due to the LAB species present. According to the results obtained in Spanish blood sausages (Santos et al., 2005a, 2005b; Diez et al., 2008) the associated spoilage LAB were heterofermentative mainly members of Leuconostoc and Weissella genus, that were reported to be more inhibited by high concentration of carbon dioxide. In MA stored WP, the low level of carbon dioxide, combined with the high aw in the beginning of the storage and low temperature gave a competitive disadvantage to LAB against other psychrotrophic aerobes, namely Pseudomonas spp. (Huis in't Veld, 1996). The behaviour of Enterobacteriaceae was similar to that of Spanish Morcilla using the same MAP of the present work (Santos et al., 2005a). In VP and WP samples, Enterobacteriaceae were detected from the day 8 of storage and grew until the day 22. After that, the reduction observed in the Enterobacteriaceae in WP samples was probably due to the reduction of aw that was more pronounced in this interval. The continuous growth of Enterobacteriaceae until the end of storage period in VP samples was associated to their facultative anaerobic character. The inhibitory effect of carbon dioxide, which is particularly efficient in Gram-negative bacteria (Mullan and McDowell, 2003), justifies the low counts observed in MAP. As expected, the behaviour of Pseudomonas spp. was highly conditioned by the packaging technology.

In WP these aerobic microorganisms growth to higher levels, while the aw was compatible with their needs. VP retarded considerably the growth of Pseudomonas spp. The best inhibitory effect was obtained when MA was packaged in MAP, once this group of spoilage microorganism was present, during all the experiment, in very 135

CHAPTER 5 ______

low counts. In addition to oxygen depletion in the packaging by substitution by nitrogen and carbon dioxide, the direct inhibitory effect of CO2 was the responsible for the strong inhibition of these Gram-negative aerobic bacteria. The absence of package as it is usually done by the producers of MA revealed to be the worst choice for the maintenance of the sensory characteristics. These MA supported well the growth of all the spoilage microorganisms studied, resulting in the development of off-odours and sooner than when package is used. Additionally, the absence of package could have contributed to the loss of the characteristic aromas by volatilization and the oxidation of lipids, as inferred by the higher scores for rancid odour. The oxidation of lipids might also have been favoured by the high count of Pseudomonas spp. in these samples. These bacteria have an important lipolytic activity (Huis in't Veld, 1996), resulting at one increase of free fatty acids which favours the oxidation phenomena once free fatty acids are more easily oxidized than when esterified in triglycerides (Hamilton, 1994). The introduction of VP represents an improvement in the shelf life. The product maintained good sensorial characteristics during storage, however, mainly due to the growth of LAB and Enterobacteriaceae it losses the freshness. The use of MAP revealed to be the best solution to increase shelf life of MA, once almost all the sensory traits were maintained during the 44 days of storage. Though these samples presented high counts of LAB, the principal spoilage microbial group in these MA, the impact of these microorganisms in sensory characteristics was modest, once they have a weak lipolytic and catabolic activity over nitrogen fraction, when compared to Enterobacteriaceae and Pseudomonas spp. (Huis in't Veld, 1996). Depending on the initial quality of the product, storage temperature and package construction and design, shelf life of cooked meat products can be extend by two- to four-fold with MAP (Hotchkiss and Langston, 1995). CO2/N2 atmospheres are referred as particularly adequate for the preservation of cooked meat products, mainly because of the strong inhibition of growth of the great majority of microorganisms and of the oxidative phenomena (Hintlian and Hotchkiss, 1987; Matos, Barreto and Bernardo, 2005; Summo et al., 2010).

136

CHAPTER 5 ______

5.6 CONCLUSIONS

The present traditional storage and distribution procedure of MA from Monchique region, without packaging, compromises the shelf life of this difficult to preserve ready-to-eat blood sausage, which was evidenced by sensory and microbiological results, and contributes to considerable economic losses for the producers by compromising the commercialization and the conquest of other markets. Accordingly to the results of the behaviour of the microbial groups during the storage period, Enterobacteriacea is associated particularly to the loss of freshness in VP, and Pseudomonas spp. in WP samples.

MAP using 80% CO2 and 20% N2, by its inhibition effect on the growth of LAB, Enterobacteriaceae and Pseudomonas spp. and on rancidity revealed to be the most efficient on extending the shelf life of Morcela de Arroz.

5.7 ACKNOWLEDGEMENTS

This work received financial support from FCT-Portuguese Foundation for Science and Technology through PhD grant SFRH/BD/63341/2009.

5.8 REFERENCES

Adams, M. R. and Moss, M. O. (2008). Food microbiology. 3rd ed, The Royal Society of Chemistry, Cambridge, U.K.

Arnesen, L. P. S., Fagerlund, A. and Granum, P. E. (2008). From soil to gut: Bacillus cereus and its food poisoning toxins. Review article. FEMS Microbiology Reviews, 32, 579-606.

Borch, E., M. Kant-Muermans, L. and Blixt, Y. (1996). Bacterial spoilage of meat and cured meat products. International Journal of Food Microbiology, 33, 103-120.

137

CHAPTER 5 ______

Buchanan, R. L. (1986). Processed meats as a microbial environment. Food Technology, 40, 134-138.

Diez, A. M., Urso, R., Rantsiou, K., Jaime, I., Rovira, J. and Cocolin, L. (2008). Spoilage of blood sausages “morcilla de Burgos” treated with high hydrostatic pressure. International Journal of Food Microbiology, 123, 246-253.

EN 12824 (1997). Microbiology of food and animal feeding stuffs. Horizontal method for the detection of Salmonella spp. Part 1: Detection method, European Committee for Standardization, Belgium, Brussels.

Hamilton, R. J. (1994). The chemistry of rancidity in foods. In Rancidity in Foods. Allen, J. C. and Hamilton, R. J. (Eds.), 1-21. Blackie Academic and Professional, London, U.K.

Hintlian, C. B., and Hotchkiss, J. H. (1987). Comparative growth of spoilage and pathogenic organisms on modified atmosphere packaged cooked beef. Journal of Food Protection, 50, 218-223.

Hotchkiss, J. H. and Langston, S. W. (1995). MAP of cooked meat and poultry products. In Principles of modified atmosphere and sous vide product packaging. Farber, J. M. and Dodds, K. L. (Eds.), 137-152, Technomic Publishing Co, Lancaster, U.S.A.

Huis in't Veld, J. H. J. (1996). Microbial and biochemical spoilage of foods: an overview. International Journal of Food Microbiology, 33, 1-18.

IBM Corp. (2010). IBM SPSS Statistics for Windows, Version 19.0. IBM Corp., New York, U.S.A.

138

CHAPTER 5 ______

International Organization for Standardization (ISO) 8586-1 (1993). Sensory analyses. General guidance for the selection, training and monitoring of assessors. Part 1: Selected assessors, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 11290-1 (1996). Microbiology of food and animal feeding stuffs. Horizontal method for the detection and enumeration of Listeria monocytogenes. Part 1: Detection method, International Organization for Standardization, Geneva, Switzerland.

Jarmoluk, A. and Pietrasik, Z. (2003). Response surface methodology study on the effects of blood plasma, microbial transglutaminase and kappa-carrageenan on pork batter gel properties. Journal of Food Engineering, 60, 327-334.

Korkeala, H. J. and Björkroth, K. J. (1997). Microbiological spoilage and contamination of vacuum-packed cooked sausages. Journal of Food Protection, 60, 724-731.

Korkeala, H., Lindroth, S., Suihko, M., Kuhmonen, A. and Penttilä, P. L. (1985). Microbiological and sensory quality changes in blood pancakes and cooked ring sausage during storage. International Journal of Food Microbiology, 2, 279-292.

Le Maguer, M. (1987). Mechanics and influence of water binding on water activity. In Water activity: theory and application to food. Rockland, L. B. and Beauchat, L. R. (Eds.), 1-25, Marcel Dekker, New York.

Mäkelä, P. and Korkeala, H. (1987). Lactobacillus contamination of cooked ring sausages at sausage processing plants. International Journal of Food Microbiology, 5, 323-330.

McMillin, K. W. (2008). Where is MAP going? A review and future potential of modified atmosphere packaging for meat. Meat Science, 80, 43-65.

139

CHAPTER 5 ______

Mullan, M. and McDowell, D. (2003). Modified atmosphere packaging. In Food packaging technology. Coles, R., McDowell, D. and Kirwan, M. J. (Eds.), 303-338, CRC Press, Boca Raton, Florida, U.S.A.

Santos, E. M., González-Fernández, C., Jaime, I. and Rovira, J. (2003). Physicochemical and sensory characterization of Morcilla de Burgos, a traditional Spanish blood sausage. Meat Science, 65, 893-898.

Santos, E. M., Jaime, I., Rovira, J., Lyhs, U., Korkeala, H. and Björkroth, J. (2005b). Characterization and identification of lactic acid bacteria in “morcilla de Burgos”. International Journal of Food Microbiology, 97, 285-296.

Silva, F. V. M. and Gibbs, P. A. (2010). Non-proteolytic Clostridium botulinum spores in low-acid cold distributed foods and design of pasteurization processes. Trends in Food Science & Technology, 21, 95-105.

Summo, C., Caponio, F.,Tricarico, F., Pasqualone, A. and Gomes, T. (2010). Evolution of the volatile compounds of ripened sausages as a function of both storage time and composition of packaging atmosphere. Meat Science, 86, 839-844.

140

CHAPTER 6

Pereira, J. A., Dionísio, L., Patarata, L. and Matos, T. J. S. Multivariate shelf life modelling in a cooked blood and rice sausage during aerobic and vacuum package storage. To be submitted for publication

141

CHAPTER 6 ______

6. MULTIVARIATE SHELF LIFE MODELLING IN A COOKED BLOOD AND RICE SAUSAGE

DURING AEROBIC AND VACUUM PACKAGE STORAGE

Pereira, J. A.a,b, Dionísio, L.c,d, Patarata, L.e and Matos, T. J. S.b,f

aDepartment of Food Engineering, Instituto Superior de Engenharia, University of Algarve, Campus da Penha, Estrada da Penha, 8005-139 Faro, Portugal bCEER - Biosystems Engineering, Tapada da Ajuda, 1349-017 Lisbon, Portugal cDepartment of Biological Sciences and Bioengineering, Faculdade de Ciências e Tecnologia, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal dICAAM - Institute of Mediterranean Agricultural and Environmental Sciences, P.O. box 94, University of Évora, 702-554 Évora, Portugal eCECAV - Centro de Ciência Animal e Veterinária, Universidade de Trás-os-Montes e Alto Douro, Quinta de Prados, 5000-801 Vila Real, Portugal fDepartment of Science and Biosystems Engineering, Instituto Superior de Agronomia, University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal

6.1 ABSTRACT

Cooked blood sausages are meat products of regular consumption in several countries. These sausages are very susceptible to spoilage and therefore have a limited shelf life. Morcela de Arroz (MA) is among these meat products and can be considered a safe cooked meat product (cooking core product temperature > 90 ºC/30 min), but it is susceptible to microbiological spoilage due to its high pH (6.0) and water activity (0.97). Little is known about the nature and relative magnitude of the spoilage events (microbiological, physicochemical and/or sensory) during MA shelf life. The aim of this study was to model the spoilage in MA during aerobic (without package - WP) and vacuum package (VP) storage (4 ± 1 ºC) by a multivariate approach, using quantitative descriptive analysis, chemical and microbiological indicators of spoilage and the consumer acceptability, in order to determine MA shelf life. Sensory descriptors associated to spoilage were generated by a trained panel and were used to characterise the sensory modifications. Pseudomonas spp., lactic acid bacteria, Enterobacteriaceae, tiobarbituric acid-reactive substances, total basic volatile nitrogen, lactic acid, were

143 CHAPTER 6 ______determined and a consumer test was performed on MA WP and VP during storage. The consumer test, considering a 50% probability of product consumer rejection, indicates that WP has a shelf life of 11.6 days (95% CI 10.4-12.9) and VP of 27.8 days (95% CI 26.0-29.5). The good predicting ability of the multivariate model, combining principal component analysis of spoilage indicators with logistic regression to predict the acceptability by consumers when compared to the apparent uselessness of using individual spoilage indicators to predict shelf life, using the same prediction tool, highlights the behaviour of the consumer choice based on an ensemble of attributes rather than in single ones.

6.2 PRACTICAL APPLICATIONS

The study provides a multivariate approach to better predict the shelf life of a ready-to- eat (RTE) cooked blood sausage and highlights that the attempt to use a panel based on microbiological and sensory data to predict shelf life has limited usefulness. This work contributes with valuable information for the better understanding of the phenomena underlying the spoilage and their relationship in RTE cooked blood sausages and to define strategies of spoilage control, as vacuum packaging, which increases shelf life of these specific meat products.

6.3 INTRODUCTION

Ready-to-eat cooked blood sausages are traditional meat products of popular consumption in several countries, namely morcela in Portugal, morcilla in Spain (Cachaldora et al., 2013) and Argentina (Oteiza et al., 2006), and krvavica in Slovenia (Gašperlin et al., 2012). Despite the different ingredients used in their formulation according to the producing region, cooked blood sausages are susceptible to spoilage and therefore present a very limited shelf life. In foods that are not shelf stable, sensory shelf life determination can potentially be predicted through chemical and/or microbial mediated metabolites as indicators of spoilage. Speciﬁc spoilage organisms (SSO) associated to microbial spoilage of cooked blood sausages may include lactic-acid bacteria (LAB) and Enterobacteriaceae when the product is vacuum packaged and also

144 CHAPTER 6 ______

Pseudomonas spp. if product has access to oxygen (Borch et al., 1996; Santos et al., 2005a). Morcela de Arroz is a traditional blood and rice cooked sausage produced in the south of Portugal (Monchique). This RTE cooked meat product is produced without chemical preservatives, unpacked and with restricted commercialization period (one week at 0-5ºC). Despite the little information available about MA, it can be considered a safe cooked meat product, considering its cooking process (core product temperature > 90 ºC for at least 30 min). Although, MA, like other blood sausages (Santos et al., 2003, 2005a), has a potential short shelf life by presenting high pH (about 6) and water activity (around 0.97), that makes it susceptible to microbiological spoilage (Diez et al., 2009), an important hurdle to production increase through the diversification of more distant markets. At present, little is known about the nature of the spoilage events (microbiological, physicochemical and/or sensory) and their relative magnitude during MA shelf life. In cooked RTE meat products shelf life determination, the chronological rejection events may not start by the rise of the microbial load during refrigerated storage, especially when using cost-effective and easy to apply packaging technologies like vacuum packaging (VP) (McMillin, 2008), that may extend considerably the MA shelf life (Santos et al., 2005a) and commercialization adequacy to the modern food chain systems (Lee, 2010). The growth of SSO in MA may result in the accumulation of compounds from theirs metabolism. Both LAB and Enterobacteriaceae are lactic acid producers. Members of LAB and Enterobacteriaceae are producers of both D- and L-lactate (Garvie, 1980; Stiles and Holzapfel, 1997). Among the LAB species isolated from blood sausages during storage, Weissella spp. and Leuconostoc spp. were identified as the predominant SSO (Santos et al., 2005a, b; Diez et al., 2008). Considering that these LAB are producers mainly of D-lactate, their contribution for the sensory detectable spoilage might be important, once the L- isomer in the cooked sausage can be originated by microbial fermentation or from the raw materials, meat and blood (Benninga, 1990; Borch et al., 1996)- D-lactate is claimed to confer a more pronounced sour flavour to the product than L-lactate (Bucharles et al., 1984; Marchesini et al., 1992). The catabolism of amino acids plays an important role in the spoilage of cooked sausages. Enterobacteriaceae and Pseudomonas spp. are the SSO usually involved in the catabolism of amino acids with production of abnormal odours, which includes ammonia, amines and are odour active compounds that are recognised as putrid odours

145 CHAPTER 6 ______

(Remenant et al., 2014). The estimation of these modifications can be assessed by the quantification of total volatile basic nitrogen present in the product (Kim et al., 2014). The main chemical event that is associated to spoilage is lipid oxidation, resulting in a variety of breakdown products which originate off-odours and ﬂavours (Kanner, 1994; Byrne et al., 2001). Cooking increases the rate and extent of oxidation (Kingston et al., 1998; Sen et al., 2011) and cooked blood meat products are especially susceptible to oxidation because of high iron content (Igene et al., 1979; Chen et al., 1984; Vercellotti, 1988). The estimation of the extension of lipid oxidation is usually assessed using the tiobarbituric acid reactive substances (TBARS) (Byrne et al., 2001; Gómez and Lorenzo, 2012). Although microbial products, the microbial population and indicators of chemical spoilage could be used to assess the degree of spoilage of this type of products (Mataragas et al., 2007). In the vast majority of the cases, shelf life is established based on a direct assessment of sensory characteristics (Hough and Garitta, 2012). Several methods can be used to achieve it, all with advantages/disadvantages (Giménez, Ares and Ares, 2012). Sensory shelf life validation is clearly dependent on consumer acceptance of the food that in turn will determine purchase intention. Recently, survival analysis was considered, both methodological and statistically, the most sound (Hough and Garitta, 2012). Nevertheless, results obtained from descriptive analysis are necessary for understanding product changes during storage. Furthermore allows linking the product microbiological and chemical characteristics with the sensory perception (Moussaoui and Varela, 2010; Varela and Ares, 2012), The aim of this study was to model the spoilage in MA during aerobic (WP) and vacuum package (VP) storage at 4 ± 1 ºC by a multivariate approach, using quantitative descriptive analysis, microbiological and chemical indicators of spoilage and the consumer acceptability, in order to determine MA shelf life and an attempt to find spoilage indicators that might be used to predict shelf life.

146 CHAPTER 6 ______

6.4 MATERIALS AND METHODS

6.4.1 Samples of MA

Samples of Morcela de Arroz from Monchique region were processed in local producers facilities according to the traditional method. Formulation included Alentejano pig breed meat (50-55%), rice (10-15%), pork blood (20-25%), lard (2-6%), grinded onion (2.5-5%), salt (about 1%), ground spices (ground cumin, ground allspice and sometimes ground clove and black pepper) and spearmint. Meat was minced and mixed with raw rice, pork blood, grinded onion, lard, salt, ground spices and spearmint. The mixture was stuffed into 55-75 mm natural casings and tied with cotton thread. MA were then transferred into a cooking vessel and cooked for about 2 h in molho de moura, a broth prepared with water, whole onions, whole garlic, salt and laurel leaves. After cooking, products were cooled at room temperature for about 1.5 h, transferred into a refrigerated chamber (4 ± 1 ºC) until the next day for product temperature stabilization and transported to the laboratory under controlled temperature (max. 4 ± 1 ºC) in sealed polyethylene bags. This experiment included three sets of samples, one for a first approach to develop lexicon to perform descriptive sensory analysis (refers to 6.3.2 section), other to study the sensory, microbiological and chemical modifications involved in the spoilage of MA stored aerobically and vacuum packaged and another to perform the consumer test. Samples for lexicon development and consumer test were previously frozen in order to have the samples with specific storage times available in the defined sensory session. For all the other determinations, samples were immediately prepared to the experiment after production. For lexicon development, MA were stored (first phase) at 4 ± 1 ºC during 8, 15 and 22, 30 days in non-sealed polyethylene bags. For the second phase the final storage time was suppressed (Table 6.3.1). For the consumer test trial, reverse storage procedure was used in order to have samples within all the experimental conditions available in a defined date (Corradini and Peleg, 2007): MA samples were frozen (until core temperature of -20 ºC was reached) in an air blast freezer (Armﬁeld FT 36, Hampshire, England). The freeze MA were vacuum packaged (VP) in a packaging machine (Interdibipack S. p. A., Milan, Italy) in oxygen and water impermeable bags Combitherm XX (dimensions of 230 x 286 mm), 115 µm

147 CHAPTER 6 ______thickness (Wipak Walsrode GmbH & Co.KG, Walsrode, Germany), with an oxygen transmission rate of 3.0 cm3/m2.24 h.atm at 23 ºC and 75% RH and a water vapour transmission rate of 1.0 g/m2.24 h at 23 ºC and 85% RH. Samples were stored at -20 ºC (Haier HF-248, Germany) until consumer test trial. Samples were thawed at 4 ± 1 ºC during 36 h progressively (45.5, 31.5 23.5, 16.5, 9.5, 1.5 days before the consumer test), randomly distributed for the experimental conditions: stored without any package (WP), to simulate the conditions of traditional commercialization, and VP for 0, 8, 15, 22, 30, 44 days of storage at 4 ± 1 ºC. To study the sensory, microbiological and chemical modifications involved in the spoilage, MA were divided in two groups, as indicated for the consumer test.

6.4.2 Lexicon development to sensory analysis

All sensory analysis sessions took place in a sensory analysis laboratory equipped with individual assessment booths and uniform lighting, equivalent to daylight type, and temperature conditions (ISO 8589, 2007). Commercial spring water and unsalted crackers were provided to clean the mouth between each sample. The descriptor set was generated according to ISO 11035 (1994). The overview of the procedure is resumed in Table 6.1. Four sessions involving free generation of vocabulary, where assessors used the descriptors they considered most accurate for describing samples of MA freshness/spoilage with 5 different periods of aerobic storage at 4 ± 1 ºC. Panellists were instructed to refer only characteristics they considered related to freshness or spoilage of the product. Once the approach involved some samples already spoiled, the texture traits were suppressed, the flavour traits were kept, once it was expected that modifications could occur in taste characteristics, namely involving the acid and bitter taste. Instructions were given to panellists to do not perform the mouth part of the analysis when they felt uncomfortable with the spoilage of the sample. The list of 36 identified terms was first screened by the panel leader for inadequately used terms. Hedonistic and quantitative terms, terms referring to the product itself or referring to technological characteristics of MA, such as amount of meat or a spice specific odour, and those indicated in the incorrect sensorial dimension were eliminated. This preliminary sorting of descriptors resulted in a shortlist of 21 terms. The terms were discussed with the panellists to make sure that

148 CHAPTER 6 ______they have well understood the meaning of each descriptor. The doubts were clarified using samples with different spoilage levels available for that effect and discussed by the group. In order to evaluate the pertinence of the terms retained, a new group for the four session were conducted with a set of samples similar to the one used in the first phase, but without the longer storage time, which corresponded to samples already very spoiled (3 MA, 4 storage times). This group of tests was made with the 21 terms identified and they were associated to a six point numeric unipolar discrete scale, with zero representing the absence of the characteristic, and 1 to 5 progressively more intensity in the characteristic (ISO 4121, 2003). The data were analyzed first by determination of frequency (F) of use and the relative intensity (I) of each descriptor in order to calculate the geometric mean (M), where M = √. , which allowed to reduce the number of terms.

Table 6.1 Procedure used for identification and selection of descriptors (adapted from ISO 11035, 1994). Phase Samples Nº S1 Nº T2

Identification of descriptive terms individually by the 15 samples 4 36 panellists 3 producers 5 storage times (4±1ºC) (0, 8, 15, 22 and 30 days)

Screening for incorrectly used terms leader 21 12 samples Testing with a 5-point+zero scale 4 3 producers 4 storage times (4±1ºC) (0, 8, 15 and 22 days) Reduction based on geometric mean – M < 0.30 leader 17 Reduction of terms based on PCA leader 14 Establishment of definitions and references 1 15a

1 - Session; 2 - Terms; a - multifactorial descriptor of global freshness introduced by the panel leader.

Descriptors presenting geometric mean values lower than 0.30 were discarded, resulting in the retention of 17 descriptors. In order to study the relationship between the terms, data was analyzed through principal component analysis. The final list included 15 descriptors for aspect, odour and flavour. A multifactorial descriptor of global freshness was introduced by the panel leader. For that final list of descriptors, definitions and references were established, proposed by the panel leader, and discussed with the panellists. It was used as references the freshly cooked MA and

149 CHAPTER 6 ______unpackaged refrigerated MA stored for 7 to 44 days to achieve the desired sensory parameter intensity. The choice of using the product under study rather than other matrixes as a reference was considered to be more useful to generalize sensory concepts, as suggested by Braghieri et al. (2012).

6.4.3 Consumer test

The consumer study was carried out with 120 subjects (48.3% female and 51.7% male), ranging in age 18-35 (18.3%), 36-50 (55.8%) and 51-65 (25.9%) years and education level: primary school (2.5%), high school (39.2%) and superior (58.3%). Subjects were recruited from Faro town (Algarve, Portugal) considering their liking of cooked blood sausages, as well as their interest and availability to participate in the study. They were presented with the 11 MA samples from Monchique region (0, 8, 15, 22, 30, and 44 days of storage at 4 ± 1 ºC, WP and VP) monadically in random order. Samples were served in slices of approximately 1 cm thick in glass covered Petri dishes and assigned with a random three-digit number, at room temperature (20 ± 3 ºC, tempered for at least 1 h before sensory analysis). For each sample, subjects were asked to evaluate overall liking using unstructured 9-point scale anchored with “liking it” at the left and “detest it” at the right and to answer the question whether they would normally consume such a MA or not.

6.4.4 Sensory analysis - quantitative descriptive analysis

The panel was composed by 18 assessors (fourteen men and four women) selected according to the recommendations in ISO 8586-1 (1993). Samples were presented monodically, in random order and served in slices of approximately 1 cm thick in glass covered Petri dishes and assigned with a random three-digit number, at room temperature (20 ± 2 ºC, tempered for at least 1 h before sensory analysis). The assessors were asked to individually evaluate 3 to 4 samples over 40-min sessions. Sessions took place once a week. Assessors were asked to score each descriptor in each sample on a scale from 1 (low intensity) to 5 (intense).

150 CHAPTER 6 ______

6.4.5. Microbiological analysis

Twenty-five grams of MA were aseptically collected and homogenized with 225 ml of sterile Ringer´s solution (Merck, Darmstadt, Germany) for 90 sec in a lab blender (Masticator, IUL Instruments, Barcelona, Spain): Pseudomonas spp. was determined on Pseudomonads Agar Base (Oxoid, Basingstoke, UK) with Cetrimide, Fucidin and Cephalosporin supplement, incubated at 25 ºC for 48 h; lactic acid bacteria (LAB) on Man Rogosa and Sharpe (Merck) and incubated on anaerobic conditions at 30 ºC for 2-3 days; Enterobacteriaceae on Violet Red Bile Glucose Agar (Merck), incubated at 37 ºC for 24 h. The microbiological counts are presented as log CFU/g. In order to verify the safety of MA to be presented to the panellists and consumers, the presence of L. monocytogenes and Salmonella spp. was screened in five samples from each batch one day after production, according to the ISO 11290-1 (1996) and ISO 6579 (2002), respectively.

6.4.6 Chemical indicators of spoilage

Tiobarbituric acid (TBA)-reactive substances were measured according to method of Tarladgis et al. (1960) modified by Shahidi et al. (1987). Total basic volatile nitrogen (TBVT) analysis was performed according to the Portuguese Standard NP 1848 (1987) using the Conway microdiffusion plates. Lactic acid quantification was performed with an enzymatic method using L- and D-lactate dehydrogenase to quantify both the L- and D-lactate, using the determination kit Nzytech AK00141 (Nzytech, Lisbon), according to the manufacturer instructions.

6.4.7 Data analysis

Data from sensory analysis in lexicon development and data from sensory, microbiological and chemical spoilage parameters were evaluated by principal component analysis (PCA). Survival analysis methodology was used to the consumer data in order to determine the cut-off point of acceptability for sliced MA stored for

151 CHAPTER 6 ______different storage times in the two packaging conditions. The analysis was done by adjusting the distribution of censored data to the Weibull analysis. The cut-off point of MA acceptability was defined considering a 50% probability of a consumer rejecting the product. Data analysis was performed using XLSTAT (Addinsoft 2014). The effect of storage time in the hedonic appreciation of MA by the consumers and the comparison of spoilage sensory, microbiological and chemical parameters between samples of the last storage time accepted and the first storage time rejected by the consumers was studied by one-way Anova. When necessary, Tukey test was used to locate the differences between the means. Logistic regression was used to search for predictors for acceptability by the consumers among the sensory, microbiological and chemical spoilage parameters and to evaluate the ability of the PCA solution to predict the acceptability. Statistical significance was set at p values less than 0.05. Data analysis was performed using XLSTAT 2014 (Addinsoft, New York).

6.5 RESULTS AND DISCUSSION

Neither L. monocytogenes nor Salmonella spp. were detected in 25 g of any of the 5 analyzed MA in the day after batch production, in any of the batches used in the experiment. Lexicon development Panellists freely generated 36 initial terms that could describe samples of MA freshness/spoilage with 5 different periods of storage at 4 ± 1 ºC. After the preliminary sorting, 21 terms were retained (Table 6.2) to be used in further testing sessions using a 6-point scale. Geometric means calculated from the frequency of use and the relative intensity of each descriptor are presented in Table 6.2. The term of aspect “sticky casing” and “cohesion” and of odour “putrid fish” and “salty fish” were discarded due to the low M. The relationship between the seventeen terms retained was analysed through principal component analysis (data not presented). It was observed that the rancid flavour was redundant with the same characteristic analysed by smelling, and that the persistency of flavour and piquant were poorly adequate to evaluate the spoilage, and were discarded. Considering the difficulty of having panellists tasting the samples during spoilage, an effort was made to reduce to the maximum the number of characteristics to be analysed in mouth. Typical, acid and bitter flavours were kept due

152 CHAPTER 6 ______to its good evolution during storage and due to the theoretical expectancy of having modifications detected by those descriptors (Labadie, 2007).

Table 6.2 List of terms identified by the panellists in the first open sessions and respective geometric mean obtained after analysis with a 6-point scale. Aspect Odour Flavour Terms M Terms M Terms M Typical colour 0.64 Typical odour 0.64 Typical flavour 0.46 Slime 0.62 Off-odours 0.58 Persistency2 0.44 Moisture 0.57 Sour/fermented 0.47 Acid 0.36 Brightness 0.51 Rancid 0.48 Piquant2 0.32 Sticky casing1 < 0.3 Putrid 0.43 Rancid2 0.32 Cohesion1 < 0.3 spoiled wine/vinegary 0.32 Bitter 0.30 mouldy/mildew 0.31 Putrid fish1 < 0.3 Salty fish1 < 0.3 1 - Term eliminated due to the low M; 2 - Term eliminated in PCA due to redundancy or inability to discriminate spoiled and fresh samples.

The final 15 descriptor list included: Aspect- moisture: moistness surface property relating to the perception of the amount of water; slime: presence of viscous substance in the surface; typical colour: characteristic dark brown colour; brightness: light reflection due to melting of fat, greasy surface; Odour- typical odour: characteristic meaty and spicy odour; off-odours: any atypical odour considered not expected in the product; rancid: associated with oxidized fat and oils, putrid: the pungent ammonic like odour in advanced spoiled meat or fish; sour/fermented: pungent odour associated to spoiled cooked rice; spoiled wine/vinegary: acetic odour associated to vinegar; mouldy/mildew: characteristic odour of moulds; Flavour- typical flavour: characteristic meaty, salted and slightly sweet flavour; acid: taste like vinegar or lemon juice; bitter: acrid and harsh taste, as the peel of an orange or coffee dregs; global evaluation of freshness: multifactorial evaluation of freshness, by opposition to spoilage. The consumer test included two aspects, a hedonic evaluation of the samples in a 9-point scale, and one binary question about the consumption intention was used as indicator for shelf life establishment, as recommended by Hough et al. (2003); Hough and Garitta (2012) and Giménez, Ares and Ares (2012). Among the 120 consumers that performed the test, 18 indicated no intention to consume MA at the first day of

153 CHAPTER 6 ______storage tested, thus they were eliminated from the data analysis. The results of the hedonic evaluation of MA WP and VP during the storage are presented in Figure 6.1.

9 Without package Vacuum package 8 shelf life limit shelf life limit 7 A 6 A B B B 5 B C C C 4 C Hedonic scale Hedonic C 3

1 0 8 15 22 30 44 Storage time (days)

Figure 6.1 Hedonic evaluation of MA WP (dark grey) and VP (light grey) for 0, 8, 15, 22, 30, 44 days of storage at 4 ± 1 ºC (n = 102 consumers). VP at day zero does not exist; for comparison purposes fresh unpacked samples were used. A, B, C - bars with different letters are different (p < 0.05) between storage times for each package.

It was observed that fresh samples had a mean hedonic appreciation above 6, and it decreased during the storage time. The decreasing was faster when samples were not packed. After 8 days of storage, WP samples remained with a similar (p ≥ 0.05) hedonic appreciation as the fresh , but it decreases significantly (p < 0.05) when it was stored for more 7 days and, after 22 days, it reaches a low level of appreciation that remained constant until the end of the storage time studied. The pattern observed for the VP samples was slightly different. The samples stored for 8 days were less appreciated (p < 0.05) than the fresh samples, but that appreciation, nearly 6 in the 9- point scale, was maintained until the 22nd day of storage. Only after that storage time a significant (p < 0.05) decrease was observed. The arrows presented in Figure 6.1, that indicate the end of shelf life, were not defined based on this hedonic scale, but rather based on the survival analysis of the binary data of intention to consume question (Table 6.3). That methodology has been applied to estimate food shelf life (Hough et al., 2003; Hough, Garitta and Gómez, 2006) based on the probability of a consumer accepting the product beyond the estimated time. To use the data obtained from the

154 CHAPTER 6 ______consumer test in estimating shelf life it is useful to adjust the response data to an appropriate model, after censoring. Censoring of data is a procedure necessary to define its nature, once three types of response can be obtained with consumer tests. The most common is the consumer that accepts the product until, for example, 8 days of storage, and rejects it from the day 15 until the end of the experiment. In this case the exact moment between 8 and 15 days when consumer reject the product is unknown. It is the case of interval censoring. If the consumer accepts the product until the end of the experiment, it is unknown exactly when will he reject it after 44 days of storage. These results are right censored. These two situations are associated to consumers that have a regular behaviour. However, in practice, there are situations that the consumer does not accept the sample at a defined storage time, but they will accept it again with a longer shelf life. In this case, it was used a left censoring, using the last time of acceptance. Results obtained with the group of 102 consumers used in the present work are presented in Table 6.3. Eighteen from the initial group of 120 consumers were eliminated from the data analysis, once they indicate to do not consume fresh MA. The majority of data was interval or right censored, indicating a predicable behaviour of consumers, but, particularly in samples WP, 21 consumers had a more erratic behaviour, as observed by their left censored results. The fitting of data from consumers tests to perform survival analysis is usually made with the Weibull or log-normal distributions (Hough et al., 2003; Giménez et al., 2007). The data from the present work presented a good fitting to the Weibull distribution. To define shelf-life, it is generally accepted to consider the time of storage corresponding to the acceptance of samples by 50% of the consumers (Table 6.3) (Hough, Garitta and Gómez, 2006; Østli et al., 2013). Using this criterion, WP samples have a shelf life of 11.6 days (95% CI 10.4-12.9) and those VP stored, 27.8 days (95% CI 26.0-29.5). In order to study the relationship between the consumer acceptability and the sensory, microbiological and chemical parameters associated to the spoilage, and considering that they were measured in the defined times of the experiment, the acceptability of the consumers was limited to the storage time immediately before the estimated shelf-life - 8 days for WP and 22 for WP. These storage times corresponded also to a shift in the hedonic appreciation (Figure 6.1).

155 CHAPTER 6 ______

Table 6.3 Censoring of data and Weibull distribution parameters µ and σ and the estimated shelf life (50% of probability of consumers’ acceptance) for MA stored WP and VP at 4 ± 1 ºC.

Censoring Weibull distribution parameters Shelf life (days) Packaging (95% CI) Left Interval Right µ (95% CI) σ (95% CI) WP 21 77 4 2.72 (2.56-2.88) 0.72 (0.60-0.86) 11.6 (10.4-12.9) VP 8 80 14 3.48 (3.39-3.58) 0.44 (0.36–0.53) 27.8 (26.0-29.5)

Using the vocabulary developed by the panel and microbiological and chemical indicators of spoilage, an experiment was conducted to study the spoilage of MA until 44 days of storage without packaging and vacuum packed. Samples were analysed at 0, 8, 15, 22, 30 and 44 days. Longer storage times, particularly in MA without packaging, were used mostly to have samples highly spoiled to improve our knowledge involved on the spoilage of this blood sausage, namely the growth of SSO - LAB, Enterobacteriaceae and Pseudomonas spp., the increase in chemical indicators of lipid oxidation (TBA), nitrogen fraction catabolism (TVBN) and fermentation of carbohydrates (lactic acid and proportion of D- lactate). Viewing the selection of the parameters associated to the spoilage that could be used to predict shelf life, a logistic model was considered, using the acceptability by the consumers as a dependent variable and the spoilage parameters as predictors. It was observed that the ensemble of variables that characterise the spoilage were individually poor predictors for the consumer acceptability, either when the data for both packages were considered, or when they were used separately, assuming a different pattern of spoilage in each packaging (results not presented). Corroborating these findings, when data from stored samples was compared by analysis of variance between the last storage time when the samples were accepted and the first when they were rejected (Table 6.4) it was observed that, for the vast majority of the variables, there were no significant differences between accepted and rejected samples. In both packages, only the total lactic acid was higher (p < 0.05) in samples rejected and in VP samples sensory evaluation of off-odours and mouldy/mildew odours. These results indicate that the use of cut-off points based on spoilage parameters to define shelf life, defined on an arbitrary base, are risky and may not have correspondence on the acceptability of the consumers (Giménez, Ares and Ares, 2012; Hough and Garitta, 2012).

156 CHAPTER 6 ______

Table 6.4 Sensory, microbiological and chemical characteristic (mean ± standard deviation) of MA in the last storage time where samples are accepted and the first where the samples are rejected by the consumers, assessed by survival analysis on consumer data, in MA stored without package and vacuum packed. Without package Vacuum packaged Accepted1 Rejected2 Accepted1 Rejected2

8 days 15 days 22 days 30 days

Sensory (5-point scale) A. Moisten 2.90 ± 0.18 3.11 ± 0.63 3.15 ± 0.17 3.24 ± 0.27

A. Slime 1.13 ± 0.12 1.21 ± 0.17 1.20 ± 0.20 1.21 ± 0.16

A. Typical colour 3.54 ± 0.53 3.18 ± 0.33 3.76 ± 0.21 3.61 ± 0.09

A. Brightness 2.96 ± 0.27 3.03 ± 0.46 2.98 ± 0.09 2.88 ± 0.30

O. Typical 3.46 ± 0.77 3.40 ± 0.10 3.80 ± 0.35 3.32 ± 0.09

O. Off-odours 1.38 ± 0.18 1.60 ± 0.22 1.19 ± 0.07a 1.58 ± 0.22b

O. Rancid 1.32 ± 0.23 1.54 ± 0.16 1.11 ± 0.04 1.27 ± 0.09

O. Putrid 1.04 ± 0.03 1.00 ± 0.00 1.08 ± 0.03 1.11 ± 0.10

O. Sour/fermented 1.11 ± 0.06 1.14 ± 0.07 1.21 ± 0.08 1.37 ± 0.25

O. Spoiled wine/vinegary 1.09 ± 0.03 1.14 ± 0.12 1.08 ± 0.04 1.24 ± 0.13

O. Mouldy/mildew 1.04 ± 0.06 1.04 ± 0.07 1.04 ± 0.04a 1.11 ± 0.00b

F. Typical 3.27 ± 0.60 2.98 ± 0.15 3.68 ± 0.40 3.32 ± 0.11

F. Acid 1.23 ± 0.06 1.27 ± 0.09 1.23 ± 0.07 1.24 ± 0.08

F. Bitter 1.12 ± 0.10 1.25 ± 0.14 1.14 ± 0.03 1.19 ± 0.03

Global freshness 3.10 ± 0.63 2.79 ± 0.30 3.48 ± 0.38 3.21 ± 0.16

Microbiological (log CFU/g)

LAB 1.48 ± 1.74 3.79 ± 0.62 4.85 ± 0.55 4.92 ± 0.45

Enterobacteriaceae 1.00 ± 1.25 1.56 ± 0.88 2.83 ± 0.57 3.52 ± 0.47

Pseudomonas spp. 1.38 ± 0.00 2.04 ± 0.00 1.54 ± 0.00 2.07 ± 0.00

Chemical

TBA (mg MDA/kg) 5.64 ± 0.66 7.03 ± 0.88 0.70 ± 0.07 0.84 ± 0.41

TBVN (mg/100g) 37.94 ± 2.26 41.39 ± 2.56 45.34 ± 2.26 47.31 ± 2.96

Total lactate (mg/100g) 21.12 ± 0.99a 24.54 ± 0.77b 24.25 ± 0.47a 25.25 ± 0.06b

D-lactate (%) 4.26 ± 0.50 4.58 ± 0.54 4.29 ± 0.43 4.57 ± 0.27

1, 2 - Samples were compared for the storage times immediately adjacent to the shelf life estimated by survival analysis; a, b - means in the same package followed by different letters are different (p < 0.05).

Considering that consumer acceptability is mostly a choice made in a multifactorial approach to several characteristics, and not only based in specific traits of the product, a multifactorial approach was done by principal component analysis (Figure 6.2), in order to understand these relationships. In the upper part of the figure it is presented the projection of the variables on the space defined by the first and the

157 CHAPTER 6 ______second principal components after varimax rotation, that together account to 82.57% of the explained variance. The original data was adequate to perform PCA, as revealed the Kaiser-Meyer-Olkin measure of sampling adequacy of 0.85 (Reis, 1997). Principal component one includes 59.4% of the information of the original data, and it represents well the pattern of MA spoilage. In the left side of the plan, it was grouped the variables that are associated to freshness, all of them of sensory nature, namely the global evaluation of freshness, moisten and bright aspect and typical colour, odour and flavour. On the opposite side of the plan are grouped those variables linked to spoilage. The sensory evaluation of slime amount in the surface, the spoiled odours like off- odours, sour/fermented, spoiled wine/vinegary, mouldy/mildew and rancid, the bitter and acid flavour were grouped with the TBA. The second PC was determined by the counts of LAB and Enterobacteriaceae, the amount of TBVN, lactic acid amount and proportion of D-lactate. The association of variables to the PC was made based in a factorial load of 0.7 or higher for variables that saturate mainly in one PC. The counts of Pseudomonas spp., the sensory evaluation of putrid odour, the TBVN and the total amount of lactate and proportion of D-isomer have factor loadings distributed by the two PCs. From the associations observed in the first PC, it is possible to infer that the spoilage process is determined by several factors that might be interrelated, namely the drying, that occurs in samples aerobically stored (WP), and results in samples with an aspect less moisten and less bright. The spoilage due to lipid oxidation was not discriminated from other phenomena involved in the spoilage, as TBA and sensory evaluation of rancid odour are grouped with several other spoilage indicators, indicating a concomitant evolution of different spoilage phenomena. Although LAB and Enterobacteriaceae were mainly associated to PC2, the chemical indicators of its metabolism are distributed by the two PC. Both of these spoilage microorganisms are lactic acid producers (Garvie, 1980; Stiles and Holzapfel, 1997) that had a contribution to the sour/fermented odour and acid flavour perceived by the panellists. Both L-and D-lactate were determined, but once the proportion of both is collinear, the L- isomer was not included in the analysis because its origin might be both from the meat and blood used to manufacture MA or microbial metabolism, while the D- isomer is associated only to the microbial action (Benninga, 1990; Pablo et al., 2008). Moreover, the impact on the odour of both isomers is slightly different. It is attributed to the D- lactate a more pungent sensory perception (Labadie, 2007). LAB and Enterobacteriaceae might produce concomitantly with lactic acid acetic acid (Garvie,

158 CHAPTER 6 ______

1980; Stiles and Holzapfel, 1997), which produces the odour recognized by the panellists as spoiled wine/vinegary. There is a clear association between the counts of Pseudomonas spp., the amount of TBVN and the sensory evaluation of putrid odour, that might be justified by the strong catabolic activity of these spoilage microorganisms have on the nitrogen fraction of the food (Huis in't Veld, 1996), resulting in ammonia and other volatile compounds that are usually sensory recognized as putrid (Nychas et al., 2008). The bitter flavour might be also associated to the proteolytic activity of Pseudomonas spp. and other bacteria, once the breakdown of proteins might result in the release of hydrophobic amino acids, or peptides with these amino acids at the end of the peptide chain, which produces that taste sensation (Careri et al., 1993; Hagen et al., 1996). The abundance of slime on the MA surface detected by the panellists was determined by the multiplication of microorganisms, namely Pseudomonas spp. (Labadie, 2007), that are aerobic and reputedly producers of exopolysaccharides (Myszka and Czaczyk, 2009). The results of the PCA also contributes to validate the adequacy of the lexicon generated in the first part of this work to study the spoilage of MA, as that vocabulary is highly discriminate among fresh and spoiled MA, and have a clear relationship to the microbiological and chemical spoilage phenomena that occurs during the storage. The projection of the observations of MA sample means WP and VP are presented in lower part of Figure 6.2. The identification of samples includes inside de parenthesis the results of the survival analysis with the results of consumer test, accept to consume (1) do not consume (0). There is a very clear relationship between the area of the plan associated to spoilage - right part of PC1 and upper part of PC2 and the samples not accepted by the consumers. The dashed line separates the vast majority of the samples that were accepted or not in the consumer test. Among not accepted samples, there is discrimination between those stored aerobically and those stored under vacuum. The absence of package resulted in samples characterized by the variables allocated in the right part of PC1 - high scores in sensory evaluation of spoiled odours and flavours and presence of slime and high levels of TBA, indicating that the rejection of these samples with lipid oxidation had an important contribution.

159 CHAPTER 6 ______

1 LAB Enterobacteriaceae TBVN Total lactate D-lactate (%) Pseudomonas spp. 0.5 O-putrid O-fermented O-vinegary A-slime TBA

F-Bitter 0 O-mouldy

PC2(23,17 %) F-Acid A-moisten A-brightness Off-odours O-typical Global Evaluation of O-rancid -0.5 freshness F-typical A-typical colour

-1 -1 -0.5 0 0.5 1 PC1 (59,40 %)

V-44 (0) V-44 (0) V-30 (0)

A-30 (0) V-30 (0) A-22 (0) A-44 (0) 1 V-44 (0) V-22 (1) V-22 (1) V-30 (0) A-22 (0)

V-15 (1) A-30 (0) V-22 (1) V-15 (1) A-15 (0) A-15 (0) A-22 (0) 0 V-15 (1) A-8 (1) A-15 (0) A-44 (0) V-8 (1) PC2(23,17 %)

V-8 (1) A-30 (0) A-44 (0) A-8 (1) -1 V-8 (1)

A/V-0 (1) A-8 (1)

A/V-0 (1) A/V (1) -2 -3 -2 -1 0 1 2 3 PC1 (59,40 %)

Figure 6.2 Loadings of the variables (upper part) and scores of the samples WP and VP with different storage times (lower part) for the PC1-PC2 dimensions, after varimax rotation, of the 22 variables used to study the spoilage of MA.

160 CHAPTER 6 ______

MA stored under vacuum with longer storage times, not accepted by the consumers, are located in the upper part of the plan, in the region defined by the presence of high counts of LAB and Enterobacteriaceae, and the respective metabolites produced by these bacteria. The PCA model synthesises well the phenomena of spoilage of MA stored aerobically and VP. The prediction ability of this model was tested by logistic regression, using the consumer acceptability as independent variable and the scores of the variables in the first two components of PCA as predictors. It was observed that the model (Table 6.5) is an excellent predictor for the shelf life of MA.

Table 6.5 Results of the logistic regression model for the acceptability of MA by the consumers considering the scores PC1 and PC2 as predictors. Variables Beta SE1 p Odds ratio 95% CI2

PC1 -8.853 4.205 < 0.035 0.0001 0.000-0.543 PC2 -5.428 2.597 < 0.037 0.0044 0.000-0.712

Log likelihood = 11.56, p < 0.001 1 SE - Standard Error, 2 CI - Confidence Interval.

It was confirmed that PC1 and PC2 are excellent predictors of MA acceptability by the consumers, as revealed by the very low odds ratios observed, indicating that if 1 unit is increased on the score of samples in PC1, samples are ten thousand times less likely to be accepted by the consumer. When the score in the second PC increases 1 unit, the probability of being accepted by the consumers is reduced 0.0044 times. This logistic model exhibited a high performance of well-classified events (92%), as indicated by the high area under the curve (AUC = 0.98) determined in the ROC (Receiver Operating Characteristic) plot. As AUC approaches 1.0 it represent a perfect separation of the test values of the two groups, while at 0.5 there are no apparent distributional difference between them (Zweig and Campbell, 1993). The good predicting ability of the multivariate model, when compared to the apparent uselessness of using individual spoilage parameters to predict shelf life highlights the behaviour of the consumer choice based on an ensemble of attributes rather than in single ones. The attempt to use panel based sensory, microbiological and sensory data to predict shelf life is frequently worthless. However, that information is

161 CHAPTER 6 ______valuable, once it allows to understanding of the phenomena underlying the spoilage and their relationship that is important, under the production point of view, to define strategies to control the spoilage and increase the shelf life.

6.6 CONCLUSIONS

The multivariate model shelf life approach through PCA represents well the pattern of MA spoilage and contributed to validate the adequacy of the lexicon generated to study the spoilage of MA, as that vocabulary is highly discriminate among fresh and spoiled MA, and have a clear relationship to the microbiological and chemical spoilage phenomena that occurs during storage at 4 ± 1 ºC. The consumer test, considering consumption intention and a 50% probability of a consumer rejection of the product, indicates that WP have a shelf life of 11.6 days (95% CI 10.4-12.9) and VP 27.8 days (95% CI 26.0-29.5). The variables that were associated to MA freshness were all of sensory nature: global evaluation of freshness, moisten and bright aspect and typical colour, odour and flavour, and were grouped separately from those linked to spoilage. The sensory evaluation of slime amount in the surface, the spoiled odours like off-odours, sour/fermented, spoiled wine/vinegary, mouldy/mildew and rancid, bitter and acid flavour were grouped with the TBA, and a clear association was found between the counts of Pseudomonas spp., the amount of TBVN and the sensory evaluation of putrid odour. Among not accepted samples, there was discrimination between those WP stored and VP stored. The absence of package resulted in samples characterized by high scores in sensory evaluation of spoiled odours and flavours and presence of slime and high levels of TBA, indicating that lipid oxidation had an important contribution to the rejection of MA WP. In turn, MA VP stored for longer storage times and not accepted by the consumers, were associated to higher counts of LAB and Enterobacteriaceae, and the respective metabolites produced by these bacteria. The good predicting ability of the multivariate model used, combining principal component analysis of spoilage indicators with logistic regression to predict the acceptability by consumers when compared to the apparent uselessness of using individual spoilage indicators to predict shelf life, using the same prediction tool, and corroborated by analysis of variance, highlights the behaviour of the consumer choice

162 CHAPTER 6 ______based on an multifactorial assessment of an ensemble of attributes rather than in single ones.

6.7 ACKNOWLEDGEMENTS

This work received financial support from FCT, Fundação para a Ciência e a Tecnologia, Portugal, through the PhD grant SFRH/BD/63341/2009. Authors gratefully acknowledge all members of the sensory panel and consumers for their participation.

6.8 REFERENCES

Benninga, H. (1990). A history of lactic acid making: A chapter in the history of biotechnology. Kluwer Academic Publishers, Dordrecht, Netherlands.

Borch, E., Kant-Muermans, M.-L. and Blixt, Y. (1996). Bacterial spoilage of meat and cured meat products. International Journal of Microbiology, 33, 103-120.

Bucharles, C., Girard, J. P., Sirami, J. and Pascal, G. (1984). Essais de caractérisation d'un saucisson acide. Science des Aliments, 3, 137-143.

Byrne, D. V., Bredie, W. L. P., Bak, L. S., Bertelsen, G., Martens, H. and Martens, M. (2001). Sensory and chemical analysis of cooked porcine meat patties in relation to warmed-over flavour and pre-slaughter stress. Meat Science, 59, 229-249.

Careri, M., Mangia, A., Barbieri, G., Bolzoni, L., Virgilli, R. and Parolari, G. (1993). Sensory property relationships to chemical data of Italian-type dry-cured ham. Journal of Food Science, 58, 968-972

163 CHAPTER 6 ______

Corradini, M. G. and Peleg, M. (2007). Shelf-life estimation from accelerated storage data. Trends in Food Science & Technology, 18, 37-47.

Diez, A. M., Björkroth, J., Jaime, I. and Rovira, J. (2009). Microbial, sensory and volatile changes during the anaerobic cold storage of morcilla de Burgos previously inoculated with Weissella viridescens and Leuconostoc mesenteroides. International Journal of Food Microbiology, 131, 168-177.

Garvie, E. (1980). Bacterial Lactate Dehydrogenases. Microbiological Reviews, 44, 106-139.

Giménez, A., Ares, F. and Ares, G. (2012). Sensory shelf-life estimation: A review of current methodological approaches. Food Research International, 49,311-325.

Giménez, A., Varela, P., Salvador, A., Ares, G., Fiszman, S. and Garitta, L. (2007). Shelf life estimation of brown pan bread: A consumer approach. Food Quality and Preference, 18, 196-204.

Gómez, M. and Lorenzo, J. M. (2012). Effect of packaging conditions on shelf-life of fresh foal meat. Meat Science, 91, 513-520.

Hagen, B. F., Berdagué, J.-L., Holck, A., Naes, H. and Blom, H. (1996). Bacterial proteinase reduces maturation time of fry fermented sausage. Journal of Food Science 61, 1024-1029.

Hough, G. and Garitta, L. (2012). Methodology for sensory shelf-life estimation. A review. Journal of Sensory Studies, 27, 137-147.

Hough, G., Garitta, L. and Gómez, G. (2006). Sensory shelf-life predictions by survival analysis accelerated storage models. Food Quality and Preference, 17, 468-473.

Hough, G., Langohr, K., Gómez, G. and Curia, A. (2003). Survival analysis applied to sensory shelf life of foods. Journal of Food Science, 68, 359-362.

164 CHAPTER 6 ______

Huis in't Veld, J. H. J. (1996). Microbial and biochemical spoilage of foods: an overview. International Journal of Food Microbiology, 33, 1-18.

International Organization for Standardization (ISO) 6579 (2002). Microbiology of food and animal feeding stuffs. Horizontal method for the detection of Salmonella spp., International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 8589 (2007). Sensory analysis. General guidance for the design of test rooms, International Organization for Standardization, Geneva, Switzerland.

Kanner, J. (1994). Oxidative processes in meat and meat products: Quality implications. Meat Science, 36, 169-189.

Kim, H. W., Choi, J. H., Choi, Y. S., Kim, H. Y., Lee, M. A., Hwang, K. E., Song, D. H., Lee, J. W. and Kim, C. J. (2014). Effects of kimchi and smoking on quality characteristics and shelf life of cooked sausages prepared with irradiated pork. Meat Science, 96, 548-553.

Kingston, E. R., Monahan, F. J., Buckley, D. J. and Lynch, P. B. (1998). Lipid oxidation in cooked pork as affected by vitamin E, cooking and storage conditions. Journal of Food Science, 63, 386-389.

Labadie, J. (2007). Spoilage microorganisms: Risks and control. In Hanbook of fermented meat and poultry. Toldrá, F. (Ed.), 421-426, Blackwell Publishing, Iowa, U.S.A.

165 CHAPTER 6 ______

Lee, K. T. (2010). Quality and safety aspects of meat products as affected by various physical manipulations of packaging materials. Review article. Meat Science, 86, 138-150.

Marchesini, B., Bruttin, A., Romailler, N. and Moreton, R. S. (1992). Microbiological events during commercial meat fermentations. Journal of Applied Bacteriology, 73, 203-209.

Mataragas, M., Skandamis, P., Nychas, G.-H. and Drosinos, E. (2007). Modeling and predicting spoilage of cooked, cured meat products by multivariate analysis. Meat Science, 77, 348-356.

McMillin, K. W. (2008). Where is MAP going? A review and future potential of modified atmosphere packaging for meat. Meat Science, 80, 43-65.

Moussaoui, K. A. and Varela, P. (2010). Exploring consumer product proﬁling techniques and their linkage to a quantitative descriptive analysis. Food Quality and Preference, 21, 1088-1099.

Myszka, K. and Czaczyk, K. (2009). Characterization of adhesive exopolysaccharide (EPS) produced by Pseudomonas aeruginosa under starvation conditions. Current Microbiology, 58, 541-546.

Norma Portuguesa (NP) 1848 (1987). Carnes, derivados e produtos cárneos. Determinação do teor em azoto básico volátil total, Instituto Português da Qualidade, Lisboa, Portugal.

Nychas, G-J. E., Skandamis, P. N., Tassou, C. C. and Koutsoumanis, K. P. (2008). Meat spoilage during distribution. Meat Science, 78, 77-89.

Østli, J., Esaiassen, M., Garitta, L., Nøstvold, B. and Hough, G. (2013). How fresh is fresh? Perceptions and experience when buying and consuming fresh cod ﬁllets. Food Quality and Preference, 27, 26-34.

Pablo, B., Asensio, M. A., Sanz, B. and Ordonez, J. A. (2008). The D(-)lactic acid and acetoin/diacetyl as potential indicators of the microbial quality of vacuum-packed pork and meat products. Journal of Applied Bacteriology, 66, 185-190.

Reis, E. (1997). Estatística multivariada aplicada. Sílabo, Lisboa, Portugal.

166 CHAPTER 6 ______

Remenant, B., Jaffrès, E., Dousset, X., Pilet, M.-F. and Zagorec, M. (2014). Bacterial spoilers of food: Behavior, fitness and functional properties. Food Microbiology, xxx, 1-9. http://dx.doi.org/10.1016/j.fm.2014.03.009.

Shahidi, F., Rubin, L. J. and Wood, D. F. (1987). Control of lipid oxidation in cooked ground pork with antioxidants and dinitrosyl ferrohemochrome. Journal of Food Science, 52, 564-565.

Stiles, M. E. and Holzapfel, W. H. (1997). Lactic acid bacteria of foods and their current taxonomy. International Journal of Food Microbiology, 36, 1-29.

Vercellotti, J. R. (1988). Current status of research on warmed-over flavour (WOF) in meat. Food Technology, 42, 101-143.

Zweig, M. H. and Campbell, G. (1993). Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry, 39, 561-577.

167

CHAPTER 7

Pereira, J. A., Silva, P., Matos, T. J. S. and Patarata, L. Shelf life determination of sliced Portuguese traditional blood sausage - Morcela de Arroz de Monchique through microbiological challenge and consumer test. Submitted for publication in Journal of Food Science

169

CHAPTER 7 ______

7. SHELF LIFE DETERMINATION OF SLICED PORTUGUESE TRADITIONAL BLOOD

SAUSAGE - MORCELA DE ARROZ DE MONCHIQUE THROUGH MICROBIOLOGICAL

CHALLENGE AND CONSUMER TEST

Pereira, J. A.a,b, Silva, P.c, Matos, T. J. S. b,d and Patarata, L.c

aDepartment of Food Engineering, Instituto Superior de Engenharia, University of Algarve, Campus da Penha, Estrada da Penha, 8005-139 Faro, Portugal bCEER - Biosystems Engineering, Tapada da Ajuda, 1349-017 Lisbon, Portugal cCECAV, Centro de Ciência Animal e Veterinária, Universidade de Trás-os-Montes e Alto Douro, Quinta de Prados, 5000-801 Vila Real, Portugal dDepartment of Science and Biosystems Engineering, Instituto Superior de Agronomia, University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal

7.1 ABSTRACT

Morcela de Arroz (MA), a traditional ready-to-eat blood and rice cooked sausage produced without preservatives in Monchique, Portugal, supports the economy of several meat industries. MA presents pH of 6.0 and aw of 0.97 and is traditionally commercialized whole, not packed, with restricted period (one week/0-5ºC) and commercialization area. MA commercialization in thick slices could be an interesting strategy to diversify the target consumers, but with safety and sensory concerns. The objective of this work was to establish sliced MA shelf life considering both the behaviour of L. monocytogenes through a microbiological challenge test and the consumer acceptability of MA stored: vacuum packed (VP), modified atmosphere packed (MAP: 80% CO2/20% N2) and aerobic packed (AP). At 3 ± 1 ºC, L. monocytogenes behaviour was not affected by packaging or storage time, highlighting the importance of the rigorous control of storage temperature. At temperature abuse (7 ± 1 ºC), the pathogen increased nearly 1 log UFC/g in the first 4 days. At 2, 7 and 12 days of storage, counts of L. monocytogenes in AP were higher (p < 0.05) than in MAP. VP presented intermediate counts (p ≥ 0.05). An eventual post-heating contamination will accelerate pathogen growth to consumer hazardous levels in the first days of storage if a temperature abuse occurs. Consumer test indicated that sliced MA sensory shelf life at 3 ± 1 ºC is 4.4 (CI: 3.8-5.1); 8.1 (CI: 6.8-9.5) and 10.4 (CI: 8.8-12) days for AP, VP and MAP, respectively, which precedes safety sliced MA shelf life.

171 CHAPTER 7 ______

7.2 PRACTICAL APPLICATIONS

This study defines shelf life for sliced Morcela de Arroz based on behaviour of Listeria monocytogens and sensory spoilage evaluated by a consumer panel and highlights the importance of the rigorous control of storage temperature in sliced cooked blood sausages.

7.3 INTRODUCTION

Morcela de Arroz (MA) is a popular ready-to-eat (RTE) cooked blood and rice sausage, produced in the Monchique region of Portugal that supports the economy of several industries. It is made with grinded pork meats and blood, rice, grinded onion, ground spices (cumin, clove, allspice, pepper and spearmint) and salt all stuffed in natural casings (pork stomach, pork or cattle intestine), boiled in a seasoned broth. This heat process (above 90 ºC/30 min in the core of the product), largely complies with the safe harbour heat treatment recommendation for the safe production of low acid refrigerated processed foods of extended durability (REPFEDs): 2 min/70 ºC allowing a 6 log reduction of L. monocytogenes (ICMSF, 2002). MA is produced without preservatives and presents both high pH (6.0) and water activity (0.97). It is traditionally commercialized whole and not packed restricting the commercialization (one week at 0-5 ºC) to the region of production. This blood sausage has cooking applications in complex meals, but it is frequently consumed as snack or as appetizer, in moderate amounts. Considering the MA weight may be as high as 400 g (Pereira, Ferreira-Dias and Matos, 2008) and its short shelf life, the introduction of the product in the market in small portions, namely thick slices, could be an interesting strategy to diversify the target consumers and consequently increase sales. However, that modification in the commercialisation pattern presents risks, namely at safety and sensory level (Feiner, 2006). MA characteristics makes it a food that potentially supports growth of Listeria monocytogenes, one of the most frequent and heat resistant psychrotrophic vegetative pathogen of concern in RTE foods (Gkogka et al., 2013, Gandhi and Chikindas, 2007; Walls, 2006; Doyle et al., 2001).

172 CHAPTER 7 ______

Microbiological challenge tests (MCT) are reported to be very useful in REPFEDs, being part of code of hygienic practice of international guidelines (CAC, 1999). Although foods shelf life determination can be based on the detection of microbial alteration, as well as physicochemical and sensorial changes, for some foods, when a recontamination is probable to occur, as during slicing and packaging of cooked RTE meat products (Mena et al., 2004; Zhu et al., 2005; Henriques, Telo da Gama and Fraqueza, 2014), shelf life should be determined first based on the period of time in which the pathogen of concern will remain in harmless level to consumers (Zwietering et al, 2010). Nevertheless, final food product shelf life validation, even for products susceptible to microbiological growth, is still dependent on consumer acceptance (Hough and Garitta, 2012). Considering, first, that cooked RTE meat products safety and quality rely on a combination of a heat treatment, storage/distribution at recommended refrigeration temperature, which are considered the primary means of preservation, and in a restricted shelf life, from few days not packaged, to several weeks, packaged in low-oxygen modified atmosphere/vacuum (Uyttendaele et al., 2009), and second, that the majority of home refrigerators show temperatures well above the recommended 4 ºC (James, Evans and James, 2008; Garrido, Carcía-Jalón and Vitas, 2010), the use of safety shelf life tests, that simulates real consumer behaviour (e. g. chilling temperature abuse scenario) are more realistic of a food use (worst case scenario). Microbiological hazard(s) in cooked RTE meat products stored under recommended conditions may not be associated or precede the sensorial deterioration. The objective of this work was to establish sliced MA shelf life considering both the behaviour of L. monocytogenes through a microbiological challenge test and the consumer acceptability of MA stored: vacuum packed (VP), modified atmosphere packed (MAP, 80% CO2/20% N2) and aerobic packed (AP).

7.4 MATERIALS AND METHODS

7.4.1 Samples and experimental design

Samples of MA were acquired in a local producer facility. The formulation included pork (loin, shoulder, neck and bacon) (54%); rice (13%), pork blood (6%), water (17%), lard (4%), grinded onion (3%), salt (1%) and ground spices (cumin,

173 CHAPTER 7 ______clove, pepper, allspice and spearmint) (2%). Meats were minced and mixed with raw rice, pork blood, water, grinded onion, lard, salt and ground spices The mixture was stuffed into 60-65 mm diameter natural beef casings and the MA were cooked (above 90 ºC/30 min in the core of the product) in a broth prepared with water, whole onions, whole garlic, salt and laurel leaves. After cooking, products were cooled to 4 ± 1 ºC. MA were transported to the laboratory under refrigeration (4 ± 1 ºC) in sterile polyethylene bags. On arrival to the laboratory five samples were randomly collected for microbiological and physicochemical analysis. The batch was then divided for the MCT and the consumer test (Figure 7.1).

MA production

MICROBIOLOGICAL AND Freezing, vacuum PHYSICOCHEMICAL package and storage at ANALYSIS -20ºC (whole MA)

Adequacy for MCT (low counts of natural microflora; absence of L. monocytogenes) and for consumer test (absence of L. monocytogenes and Salmonella spp.) MICROBIOLOGICAL CONSUMER TEST CHALLENGE TEST Reverse storage Thawing during 36 h at 3 ºC Thawing during 36 h at 3 ºC at (days before testing) 3 6 9 14 22 32

Slicing and inoculation with L. monocytogenes Slicing and packaging AP VP MAP Packaging Storage at 3±1ºC during (days) AP VP MAP

1 4 7 12 20 30 Storage temperature recommended abusive Consumer test 3 ± 1 ºC 7 ± 1 ºC 2 sessions in consecutive days

9 samples/session randomly Sampling at (days) distributed 0 1 2 4 7 12 20

Selection of the more adequate conditions for the consumer test

Figure 7.1 Experimental design for shelf life determination based on a microbiological challenge test (MCT) and a consumer test.

174 CHAPTER 7 ______

MA samples were frozen in an air blast freezer (Armﬁeld FT 36, Hampshire, England) until the core temperature of -20 ºC was reached. The freeze MA were than vacuum packaged in oxygen and water impermeable bags Combitherm XX, 115 µm thickness (Wipak Walsrode GmbH & Co.KG, Walsrode, Germany) and stored at -20 ºC (Haier HF-248, Germany) until beginning of the MCT (one week) and the consumer test study (five weeks). For the MCT, MA were thawed at 3 ± 1 ºC during 36 h, sliced (1 cm thick), inoculated with L. monocytogenes and randomly distributed for the three packaging conditions (AP, VP and MAP: 80% CO2/20% N2) and storage times (in triplicate).

7.4.2 Microbiological and physicochemical analysis of fresh MA

The natural microflora of the MA were counted to detect any biases that their previous overgrowth could produce in the experiment. Twenty-five grams of MA were aseptically collected and homogenized with 225 ml of sterile Ringer´s solution (Merck, Darmstadt, Germany) for 90 sec in a lab blender (Masticator, IUL Instruments, Barcelona, Spain): Mesophilic total viable count (MTVC) was determined on Plate Count Agar (PCA, Merck), incubated at 30 ºC for 72 h; psychrotrophics (PTVC) were determined on PCA, incubated at 6.5 ºC for 10 days; lactic acid bacteria (LAB) on Man Rogosa and Sharpe (Merck) and incubated on anaerobic conditions at 30 ºC for 2- 3 days; Enterobacteriaceae on Violet Red Bile Glucose Agar (Merck), incubated at 37 ºC for 24 h; Pseudomonas spp. on Pseudomonads Agar Base (Oxoid, Basingstoke, UK) with Cetrimide, Fucidin and Cephalosporin supplement, incubated at 25 ºC for 48 h. The microbiological counts are presented as log CFU/g. In order to verify the safety of MA to be presented to the consumers, the presence of L. monocytogenes and Salmonella spp. was screened according to the ISO 11290-1 (1996) and ISO 6579 (2002), respectively. Moreover, the absence of L. monocytogenes in the samples is a prerequisite for the correct execution of the MTC. Samples for physicochemical analysis were grounded and homogenized in a domestic food processor. Chemical composition was determined according to ISO recommended procedures: moisture (ISO 1442, 1997), total fat (ISO 1443, 1973), protein (Kjeldahl; N x 6.25) (ISO 937, 1978), NaCl (ISO 1841-1, 1996) and ash (ISO 936, 1998) contents. Carbohydrates content was estimated by subtracting to 100% the moisture, protein, fat and ash percentages. Water activity

175 CHAPTER 7 ______

(aw) was determined in a Rotronic Hygroskop DT (Rotronic, HygroLab, Bassersdorf, Switzerland) at 25.0 ± 0.01 ºC. pH was determined with a digital pH meter (GLP 21, Crison, Barcelona, Spain) and by inserting the electrode directly into the sample. Determinations were made in triplicate.

7.4.3 Microbiological challenge test

Three L. monocytogenes strains isolated from the environment of meat products industry (laboratory collection) and one strain from ATCC (35152) were used in this study. The strains, maintained at -18ºC, were subcultured twice in Brain Heart Infusion (BHI, Biokar. Beauvais, France) and incubated for 24 h at 30ºC before use. Cultures for inoculation were grown individually overnight in 30 mL of BHI, harvested by centrifugation, washed twice and resuspended in 0.85% NaCl. A cocktail of the four strains was prepared to achieve the level of inoculation desired, starting with a suspension with a turbidity similar to the Macfarland standard number 2 (Biomerrieux, Marcy l'Etoile, France). Slices of MA (25 ± 3g each) were individually inoculated with 0.1 ml of L. monocytogenes suspension previously diluted to achieve an inoculation level of about 3 log UFC/g and were left for 30 min at refrigeration temperature to allow the inoculum absorption. The inoculated slices (3 slices for each experimental condition) were than randomly distributed for the 3 packaging conditions (AP, VP and MAP: 80% CO2/20%

N2), the 7 storage times (0, 1, 2, 4, 7, 12 and 20 days) and the two storage temperatures: 3 ± 1 ºC, and 7 ± 1 ºC. For AP and MAP three slices were distributed in polystyrene trays. For the first, trays were putted in polyethylene bags closed with a clip and for the second they were packed in MAP (80% CO2/20% N2). Due to the contact of inoculated slices surface with the packaging film, VP samples were packaged individually and were homogenized in the original package bag before L. monocytogenes counting. Both MAP and VP samples were packaged in oxygen and water impermeable bags Combitherm XX, 115 µm thickness (Wipak Walsrode GmbH & Co.KG, Walsrode, Germany). In each sampling moment, slices were unpacked and each was individually taken, weighted and homogenized for 90 sec in sterile plastic bag in a lab blender with the needed volume of NaCl 0.85% to achieve the initial dilution of 0.1 g/ml. VP samples were homogenised in the original package bag (weighted in the bag and subtracting the bag weigh). When

176 CHAPTER 7 ______necessary, serial decimal dilutions were prepared with the same solution and 0.1 ml of appropriate dilutions were spread onto selective Oxford Agar (Biokar BK110+ BS003, Biokar Diagnostics, Beauvais, France) followed by incubation at 30 ºC for 24 h.

7.4.4 Consumer test - sample preparation and analysis

Considering the results of the microbiological challenge test, the consumer test was conducted only with MA stored at 3 ± 1 ºC, once the abusive temperature tested, beside the theoretical and legal impropriety, resulted in a very short shelf life based on the growth of L. monocytogenes. The storage periods used in the consumer test were adjusted, eliminating the time zero and two days of storage, once no significant modifications were expected in such short intervals. Once the results of the microbiological challenge test allowed keeping it for at least 20 days, an extra storage time of 30 days was included to avoid the eventual situation of a massive acceptability of MA at 20 days of storage which might resulted in difficulties of shelf life determination based on survival analysis of consumer data. To perform the consumer test it was used a reverse storage strategy, in order to have samples within all the experimental conditions available in a defined date (Corradini and Peleg, 2007). Once each consumer had to evaluate a total of 18 samples (6 storage periods for 3 packaging conditions), the reverse storage was designed to have the samples available in two consecutive days. Samples were distributed for the two sessions randomly. The randomisation was previous to the thawing of samples to guaranty that samples had the correct storage time at the moment of the testing. The consumer study was carried out with 70 consumers (47% female and 53% male), ranging in age from 18 to 65 years. Subjects were recruited considering their liking of cooked blood sausages, as well as their interest and availability to participate in the study. Each consumer received 18 sliced samples of MA (9 in each session). Samples were served in glass Petri dishes identified with a random three-digit number, at room temperature (20 ± 2 ºC). Consumers were asked to answer “yes” or “no” to the question “Would you normally consume this product?”. Commercial spring water and unsalted crackers were provided to rinse the mouth between samples. Consumers evaluated the samples on a sensory analysis laboratory equipped with individual assessment booths and uniform lightening and temperature conditions.

177 CHAPTER 7 ______

7.4.5 Statistical analysis

The effects of packaging and time of storage on the behaviour of L. monocytogenes were individually tested by comparing the mean counts by one-way ANOVA. The Tukey test was used to determine the significant differences (p < 0.05) between group means. Survival analysis methodology was used to the consumer data in order to determine the cut-off point of acceptability for sliced MA stored for different storage times at 3 ± 1 ºC in the three packaging conditions. According to Hough et al. (2003), only data from consumers that not rejected the freshest sample (1 day) were included in the analysis (55 consumers). The cut-off point of the sliced MA refrigerated stored at 3 ± 1 ºC was estimated considering a 50% probability of a consumer rejecting the product. Data analysis was performed using XLSTAT (Addinsoft 2014).

7.5 RESULTS AND DISCUSSION

The natural microflora of the fresh MA (Table 7.1) was generally low. Only for MTVC and PTVC small counts were observed. LAB presented low counts in only two of the 5 analysed samples. Enterobacteriaceae and Pseudomonas spp. were below the detection limit. These low microbial counts were expected, once the product was cooked (above 90 ºC/30 min), and allowed performing the microbiological challenge test without the risk of having strong interactions between MA natural microflora and the inoculated pathogen. Neither L. monocytogenes nor Salmonella spp. were detected in 25 g of any of the 5 analysed MA in the day after batch production.

Table 7.1 Microbiological and physicochemical analysis of fresh MA (Mean ± standard deviation), n = 5. Microbiological Physicochemical

MTVC (log CFU/g) 2.04 ± 0.02 Water (g/100 g) 54.15 ± 1.16

PTVC (log CFU/g) 1.47 ± 0.69 Protein (g/100 g) 16.01 ± 0.72 LAB (log CFU/g) 0.44 ± 0.36 Fat (g/100 g) 13.74 ± 1.37 Enterobacteriaceae (log CFU/g) < DLa Carbohydrate (g/100 g) 14.14 ± 0.91 Pseudomonas spp. (log CFU/g) < DLa Ash (g/100 g) 1.96 ± 0.10 NaCl (g/100 g) 1.91 ± 0.10

L. monocytogenes (presence in 25g) absent n = 5 aw 0.97 ± 0.01 Salmonella spp. (presence in 25g) absent n = 5 pH 6.06 ± 0.04 a - Detection limit.

178 CHAPTER 7 ______

The proximate composition of MA (Table 7.1) shows that it is a nutritious meat product that, conjugated with the high aw (0.97 ± 0.01) and high pH (6.06 ± 0.04), indicates that MA is highly susceptible to support well the growth of L. monocytogenes. The counts of L. monocytogenes are presented in Figures 7.2 and 7.3, for the storage temperatures of 3 ± 1 ºC and 7 ± 1 ºC for each packaging condition during the 20 days of storage, respectively.

7 AP 6 VP MAP 5

4 Log CFU/g Log 3

2 0 2 4 6 8 10 12 14 16 18 20 Stoarge time (days) Figure 7.2 Counts of L. monocytogenes (Log CFU/g) in MA slices packaged in aerobic condition (AP), vacuum packaging (VP) and modified atmosphere packaging (MAP, 80% CO2 and 20% N2) stored at 3 ± 1 ºC. Results are presented as mean ± sd (n = 3).

7 aE E 6 D abCD aD D C 5 aBC abBC AB AB abAB 4 AB bBC bC Log CFU/g Log A A AP 3 ABC VP MAP A A bA 2 0 2 4 6 8 10 12 14 16 18 20 Stoarge time (days) Figure 7.3 Counts of L. monocytogenes (Log CFU/g) in MA slices packaged in aerobic condition (AP), vacuum packaging (VP) and modified atmosphere packaging(MAP, 80% CO2 and 20 % N2) stored at 7 ± 1 ºC. Results are presented as mean ± sd (n = 3). a-b, means with different lowercase letters are different (p < 0.05) (between packaging technologies at the indicated storage time). A-E, means with different uppercase letters are different (p < 0.05) (between storage periods for each packaging condition) (p < 0.05).

179 CHAPTER 7 ______

During the storage of sliced MA at the recommended storage temperature (3 ± 1 ºC), L. monocytogenes counts slightly increased from a count of 2.97 ± 0.16 log CFU/g to 3.68 ± 0.68 log CFU/g in the AP condition, 3.25 ± 0.24 log CFU/g in the VP and 3.04 ± 0.15 log CFU/g for the packaged MA conditions. In any storage condition the mean counts of L. monocytogenes reached 1 log CFU/g of increase, usually used as indicator to define the limit of the shelf life of the product based on pathogen growth (NACMCF, 2010). The behaviour of L. monocytogenes at that storage temperature was not affected by the packaging condition or by the storage time, revealing that the rigorous control of the recommended storage temperature of MA is enough to control L. monocytogenes growth. When temperature abuse was tested (7 ± 1 ºC) (Figure 7.3), it was observed that the growth of the pathogen was, as expected, faster and reached an increase of one logarithmic unit in the first few days of storage. At that abusive temperature, the behaviour of L. monocytogenes was significantly affected by the packaging condition and the storage time. At the second day of storage and thereafter, the counts of L. monocytogenes observed in AP samples were higher (p < 0.05, at 2, 7 and 12 days) than those observed in MAP. Generally, VP samples presented similar intermediate counts (p ≥ 0.05) to those observed in AP and MAP slices. This is an expected result considering the physiology of this microorganism, being a facultative anaerobe, the presence of O2 increases its growth rate when compared with atmospheres with reduced O2 levels (Fernández et al., 1997). It is accepted that CO2 increases the lag phase and generation time of microorganisms, being more effective as

CO2 proportion increases in the gas mixture, even those that can growth in the absence of oxygen, namely L. monocytogenes (García de Fernando et al., 1995). Uyttendaele et al. (2009) also found an inhibition effect of MAP containing 50% CO2 on the growth of L. monocytogenes in sliced cooked meat products, namely in luncheon meat and pâté. This inhibitory effect is due to the reduction of pH, inhibition of certain enzymes involved in energy production and damage of the cell membrane. Thus, the reduction of L. monocytogenes growth in MAP sliced MA can be attributed to the direct effect of carbon dioxide by extension of the lag phase and generation time of this bacteria. Additionally, recommended refrigerated storage temperature must be controlled to ensure the effectiveness of the MAP gas mixture on microbial growth inhibition. When

CO2 is used in MAP it is absorbed by water and lipid portions of meat until saturation or equilibrium is reached (Jakobsen and Bertelsen, 2002; Sivertsvik and Jensen, 2005).

180 CHAPTER 7 ______

Product pH and lower temperature increases CO2 absorption and its effectiveness as a bacteriostatic (Carr and Marchello, 1986). The international guidelines (CAC, 2007) for storage temperature of RTE foods is ≤ 6 ºC (preferably 2 ºC to 4 ºC), considering that L. monocytogenes is a psychrotrophic pathogen, with a minimum growth temperature in the range of 0 to 5 ºC (Abee and Wouters, 1999), our results are, in this sense, expected. Present European recommendations (European Regulation No. 1441/2007) on microbiological criteria, in RTE meat products that might support the growth of L. monocytogenes, this pathogen must be absent in 25 g of meat product before it lefts the immediate control of the Food Business Operator and maximum concentration must be ≤ 100 CFU/g in the product placed on the market during its shelf life. According to American recommendations (FDA, 2001), in these RTE foods, 1 log count increase was considered appropriate for control of L. monocytogenes providing an appropriate level of protection for consumers. According to our results, at the recommended storage temperature (Figure 7.2), sliced MA, for all the packaging conditions, did not exceed this limit during the entire storage period (20 days), determining that sliced MA was fit for consumption until the end of the storage period studied. Considering the temperature abuse scenario of 7 ± 1 ºC (Figure 7.3), the limit of 1.0 log CFU/g increase was exceeded in all the packaging conditions. The aerobic condition revealed to be the worst option by being unsafe for consumers after 2 days of storage (pathogen reached 4.05 log CFU/g at 4 days of storage) comparatively to the packaged MA, that were unfit for consumption after 4 days (at 7 days, L. monocytogenes reached 4.33 and 4.09 log CFU/g in the VP and MAP condition, respectively) of storage in a temperature abuse condition. This results are in agree with the previous results on meat and cooked meat products, with respect the package effect (Liserre et al., 2002; Mataragas, Drosinos and Metaxopoulos, 2003; Uyttendaele et al., 2009) and the storage temperature effect (Vermeiren et al., 2006; Garrido, Carcía-Jalón and Vitas, 2010) on the growth of L. monocytogenes. Based on the MA L. monocytogenes challenge test at both refrigeration storage temperatures studied, the consumer test was performed only with the sliced MA in the three packaging conditions stored at 3 ± 1 ºC for 1, 4, 7, 12, 20, 30 days. In order to determine the end of the sensory shelf life of sliced MA, survival analysis methodology was used to determine the cut-off point and it was considered as 50% probability of a consumer rejecting the product (Hough et al., 2003; Hough, Garitta

181 CHAPTER 7 ______and Gómez, 2006). To determine the end of shelf life by survival analysis, the most used Weibull distribution was selected, by being a very flexible right-skewed distribution and specific appropriate for modeling survival data (Calle et al., 2006) (Table 7.2).

Table 7.2 Weibull distribution parameters µ and σ and the estimated shelf life (50% of probability of consumers’ acceptance) for sliced MA stored at 3 ± 1 ºC in different packaging conditions.

Censoring Weibull distribution parameters Shelf life (days) Packaging (95% CI) Left Interval Right µ (95% CI) σ (95% CI) AP 16 39 0 1.75 (1.53-1.93) 0.71 (0.56-0.91) 4.4 (3.8-5.1) VP 16 37 2 2.41 (2.16-2.66) 0.85 (0.65–1.11) 8.1 (6.8-9.5) MAP 13 2 3 2.63 (2.40-2.86) 0.79 (0.61-1.03) 10.4 (8.8-12)

According to Hough et al. (2003), only data from consumers that not rejected the freshest sample (1 day) were included in the analysis. 15 consumers within the initial 70 were abandoned because they rejected the freshest sample. In the AP condition and has expected, no consumer accepted the sliced MA at 30 days of storage (right censoring), 39 consumers accepted the samples with shorter storage times and then rejected all the AP samples thereafter with longer storage times (interval censoring) and 16 consumers, that accepted the freshest sample but along storage time accepted/rejected AP samples with different storage times (left censoring). For the VP condition, also 16 consumers presented this behaviour (left censoring), only two consumers accepted the VP samples at 30 days of storage, most of the consumers (37) presented interval censored data. For the MAP condition, 13 consumers presented left censored data, 3 consumers presented right censored data and 2 interval censored data. Thus, 55 consumer data were considered for the survival analysis in order to determine the shelf life of sliced MA. According to Hough (2006), fifty to hundred consumers evaluating acceptability of samples with different storage times is sufficient to estimate the shelf life of a product. In our study, the consumers tested 18 samples in two sessions. Generally, this is the number of samples presented to consumer panels for the survival analysis shelf life estimation (Hough et al. 2003, 2006). MA sensory shelf life at recommended refrigerated storage temperature and considering 50% probability of

182 CHAPTER 7 ______consumer rejection was estimated at 4.4 (3.8-5.1) days for AP, 8.1 (6.8-9.5) days for VP and 10.4 (8.8-12) days for MAP with 95% confidence interval.

7.6 CONCLUSIONS

At recommended refrigeration storage temperature (3 ± 1 ºC) the behaviour of L. monocytogenes was not affected by the packaging condition or by the storage time, revealing that the rigorous control of the recommended storage temperature of MA is enough to control L. monocytogenes growth. However, in a temperature abuse scenario (7 ± 1 ºC), as expected, the pathogen presented faster growth rates and reached an increase of nearly 1 log CFU/g in the first 4 days of storage. At 2 days of storage, counts of L. monocytogenes in AP samples were higher (p < 0.05) than those observed in MAP (80% CO2/20% N2) and, VP samples, presented similar intermediate counts (p ≥ 0.05). Indicating that an eventual post-heating contamination will accelerate pathogen growth to consumer hazardous levels in the first days of storage if a temperature abuse occurs. The consumer test revealed that the sensory shelf life of sliced MA at recommended refrigeration storage temperature (3 ± 1 ºC) is limited to 4.4 (3.8-5.1) days for AP, 8.1 (6.8-9.5) days for VP and 10.4 (8.8-12) days for MAP with 95% confidence interval. When sliced MA is stored at recommended temperature (3 ± 1 ºC) shelf life should be established based on sensory spoilage, once sensory rejection by the consumers occurs long before the product is considered hazardous due to the possible growth of L. monocytogenes.

7.7 ACKNOWLEDGMENTS

183 CHAPTER 7 ______

7.8 REFERENCES

Abee, T. and Wouters, J. A. (1999). Microbial stress response in minimal processing. International Journal Food Microbiology, 50, 65-91.

CAC (Codex Alimentarius Commission) (2007). Guidelines on the application of general principles of food hygiene to the control of Listeria monocytogenes in foods. CAC/GL 61. Retrieved from http://www.codexalimentarius.net/input/download/standards/10740/CXG_061e.pdf Accessed on 03-04-2013.

Calle, M. L., Hough, G., Curia, A. and Gómez, G. (2006). Bayesian survival analysis modeling applied to sensory shelf life of foods. Food Quality and Preference, 17, 307-312.

Carr, T. P. and Marchello, J. A. (1986). Microbial changes of precooked beef slices as affected by packaging procedure. Journal of Food Protection, 49, 534-536.

Corradini, M. G. and Peleg, M. (2007). Shelf-life estimation from accelerated storage data. Trends in Food Science & Technology, 18, 37-47.

Doyle, M. E., Mazzota, A. S., Wang, T., Wiseman, D. W. and Scott, V. N. (2001). Heat resistance of Listeria monocytogenes. Journal of Food Protection, 64, 410-429.

European Regulation No 1441 (2007). Commission Regulation of 5th December 2007. Official Journal of the European Union.2007, L322, 12-29.

Feiner, G. (2006). Meat products handbook: Practical science and technology. Woodhead Publishing, Cambridge, U.K.

FDA (Food and Drug Administration) (2001). Evaluation and Definition of Potentially Hazardous Foods: Chapter 6. Microbiological Challenge Testing. A Report of the Institute of Food Technologists for the Food and Drug Administration of the U.S. Department of Health and Human Services. Retrieved from http://www.fda.gov/Food/FoodScienceResearch/SafePracticesforFoodProcesses/ucm 094154.htm Accessed on 6-01-2014.

184 CHAPTER 7 ______

Fernández, P. S., George, S. M., Sills, C. C. and Peck, M. W. (1997). Predictive model to the effect of CO2, pH, temperature and NaCl on the growth of Listeria monocytogenes. International Journal of Food Microbiology, 37, 37-45.

Gandhi, M. and Chikindas, M. L. (2007). Listeria: A foodborne pathogen that knows how to survive. International Journal of Food Microbiology, 113, 1-15.

García de Fernando, G. D., Nychas, G. J. E., Peck, M. W. and Ordóñez, J. A. (1995). Growth/survival of psychrotrophic pathogens on meat packaged under modiﬁed atmospheres. International Journal of Food Microbiology, 28, 221-231.

Garrido, V., Carcía-Jalón, I. and Vitas, A. I. (2010). Temperature distribution in Spanish domestic refrigerators and its effect on Listeria monocytogenes growth in sliced ready-to-eat ham. Food Control, 21, 896-901.

Gkogka, E., Reij, M. W., Gorris, L. G. M. and Zwietering, M. H. (2013). The application of the Appropriate Level of Protection (ALOP) and Food Safety Objective (FSO) concepts in food safety management, using Listeria monocytogenes in deli meats as a case study. Food Control, 29, 382-393.

Henriques, A. R., Telo da Gama, L. and Fraqueza, M. J. (2014). Assessing Listeria monocytogenes presence in Portuguese ready-to-eat meat processing industries based on hygienic and safety audit. Food Research International, 63, 81-88.

Hough, G. (2006). How does survival analysis help us in estimating the probability of a consumer rejecting a stored product? Workshop summary: Sensory shelf-life testing. Food Quality and Preference, 17, 640-645.

Hough, G. and Garitta, L. (2012). Methodology for sensory shelf-life estimation: a review. Journal of Sensory Studies, 27, 137-147.

Hough, G., Garitta, L. and Gómez, G. (2006). Sensory shelf-life predictions by survival analysis accelerated storage models. Food Quality and Preference, 17, 468-473.

Hough, G., Langohr, K., Gómez, G. and Curia, A. (2003). Survival analysis applied to sensory shelf life of foods. Journal of Food Science, 68, 359-362.

International Organization for Standardization (ISO) 936 (1998). Meat and meat products. Determination of ash content, International Organization for Standardization, Geneva, Switzerland.

185 CHAPTER 7 ______

International Organization for Standardization (ISO) 937 (1978). Meat and meat products. Determination of nitrogen content, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 1442 (1997). Meat and meat products. Determination of moisture content, International Organization for Standardization, Geneva, Switzerland.

International Organization for Standardization (ISO) 1443 (1973). Meat and meat products. Determination of total fat content, International Organization for Standardization, Geneva, Switzerland.

Jakobsen, M. and Bertelsen, G. (2002). The use of CO2 in packaging of fresh red meats and its effect on chemical quality changes in the meat: A review. Journal of Muscle Foods, 13, 143-168.

James, S. J., Evans, J. and James, C. (2008). A review of the performance of domestic refrigerators. Journal of Food Engineering, 87, 2-10.

Liserre, A. M., Landgraf, M., Destro, M. T. and Franco, B. D. G. M. (2002). Inhibition of Listeria monocytogenes by a bacteriocinogenic Lactobacillus sake strain in modified atmosphere-packaged Brazilian sausage. Meat Science, 61, 449-455.

Mataragas, M., Drosinos, E. H. and Metaxopoulos, J. (2003). Antagonistic activity of lactic acid bacteria against Listeria monocytogenes in sliced, cooked cured pork shoulder stored under vacuum or modiﬁed atmosphere at 4 ± 2 ºC. Food Microbiology, 20, 259-265.

Mena, C., Almeida, G., Carneiro, L., Teixeira, P., Hogg, T. and Gibs, P. A. (2004). Incidence of Listeria monocytogenes in different food products commercialized in Portugal. Food Microbiology, 21, 213-216.

186 CHAPTER 7 ______

NACMCF (National Advisory Committee on Microbiological Criteria for Foods) (2010). Parameters for determining inoculated pack/challenge study protocols. Journal of Food Protection, 73, 140-202.

Sivertsvik, M. and Jensen, J. S. (2005). Solubility and absorption rate of carbon dioxide into non-respiring foods. Part 3: Cooked meat products. Journal of Food Engineering, 70, 499-505.

Uyttendaele, M., Busschaert, P., Valero, A., Geeraerd, A. H., Vermeulen, A., Jacxsens, L., Goh, K. K., De Loy, A., Van Impe, J. F. and Devlieghere, F. (2009). Prevalence and challenge tests of Listeria monocytogenes in Belgian produced and retailed mayonnaise-based deli-salads, cooked meat products and smoked ﬁsh between 2005 and 2007. International Journal of Food Microbiology, 133, 94-104.

Vermeiren, L., Devlieghere, F., Vandekinderen, I. and Debevere, J. (2006). The interaction of the non-bacteriocinogenic Lactobacillus sakei 10A and lactocin S producing Lactobacillus sakei 148 towards Listeria monocytogenes on a model cooked ham. Food Microbiology, 23, 511-518.

Walls, I. (2006). Role of quantitative risk assessment and food safety objectives in managing Listeria monocytogenes on ready-to-eat meats. Meat Science, 74, 66-75.

Zhu, M., Du, M., Cordray, J. and Ahn, D. U. (2005). Control of Listeria monocytogenes contamination in ready-to-eat meat products. Comprehensive Reviews in Food Science and Food Safety, 4, 34-42.

187

CHAPTER 8

FINAL CONSIDERATIONS

189

CHAPTER 8 ______

8. FINAL CONSIDERATIONS

8.1 CONCLUSIONS

Traditional Morcela de Arroz (MA) from Monchique region is both a nutritional food (protein content of 16%, fat and carbohydrate about 14%, g/100 g) and a product of difficult preservation with pH around 6.0 and high aw reaching values higher than 0.97.

The traditional thermal process (cooking) applied to MA is long, involves temperatures above 95 ºC for about 38 min in the core of MA and allows to safely extend the refrigerated shelf life of this product by ensuring a reduction of the most heat resistant facultative psychrotrophic vegetative pathogen of concern in ready-to-eat (RTE) meat products, Listeria monocytogenes, largely above the 6D recommended (ICMSF, 2002) for this type of meat product. The MA thermal process also ensures a 6D reduction of the number of spores of psychrotrophic non proteolytic C. botulinum strains (ACMSF, 1992).

The unsteady-state heat transfer equations model (USHTE) used allowed the adjustment of external conditions and product dimensions without the need for the continuous monitoring of MA internal temperature. The model can easily be used to predict the adequacy of MA thermal process in terms of food safety. Average cooking loss of 4 batches was 20.3%.

Generally, MA complies with the FSA (2008) recommendations for the safe production of VP and MAP chilled foods with a shelf life longer than 10 days, with respect to the growth and toxin production by non proteolytic C. botulinum, opening good perspectives for the application of these packaging technologies.

The traditional thermal process of MA cannot theoretically destroy the most heat resistant proteolytic strains of C. botulinum. That might be a problem only if a temperature abuse occurs.

MA cooling and chilling complies with the FDA (2009) recommendations for cooked potential hazardous foods.

Post-cooking contamination on the surface of the product is caused by handling of the product during cooling, refrigeration and post processing operations. In this way,

191

CHAPTER 8 ______

Good Hygienic and Manufacturing Practices must continue to be included in the control of the process, with particular attention to the thermal process and storage temperature.

The study allowed to establish the lexicon for sensory characteristics of Monchique MA which can be used in quantitative descriptive analysis of this traditional product, allowing MA characterization, particularly to discriminate the origin of the product, in order to apply for one of the EC special labels certification (PDO or PGI) and also to improve sensory quality control in this type of meat products.

Recently, European Regulation No 510/2006 was replaced by the European Regulation No 1151/2012 in which the scope for DOP and IGP is still limited to foodstuffs for which an intrinsic link exists between the foodstuff characteristics and geographical origin. This recent European Regulation aims to help producers of agricultural products and foodstuffs to communicate the product characteristics and farming attributes of those products and foodstuffs to buyers and consumers. In Article 7, referring to product specification: “A PDO or IGP shall comply with a specification which shall include at least: 1.(b) a description of the product, including the raw materials, if appropriate, as well as the principal physical, chemical, microbiological or organoleptic characteristics of the product”; 1.e) a description of the method of obtaining the product and, where appropriate, the authentic and unvarying local methods as well as information concerning packaging, if the applicant group so determines and gives sufficient product-specific justification as to why the packaging must take place in the defined geographical area to safeguard quality, to ensure the origin or to ensure control, taking into account Union law, in particular that on the free movement of goods and the free provision of services”.

The introduction of references for the selected descriptors and the use of a wider scale increased the performance of the panellists. Additionally, the list of selected descriptors and references should help researchers to elucidate about the factors which influence typical appearance, odour, flavour and texture in this type of meat products.

Monchique MA has a less moisten surface, the pieces of meat and the rice are bigger, but the size of fat pieces is smaller, than those from other origins. Monchique MA has a weaker smoke odour, but presents a clear higher odour note to cumin and spearmint, which can be assumed as a distinctive trait. The overall flavour and the

192

CHAPTER 8 ______spices flavours are slightly higher. Due to the amount of meat, and to the size of its pieces, Monchique MA is harder and firmer, less pasty and adhesive.

Considering the different packaging technologies employed and accordingly to the results of the microbial groups behaviour in each packaging technology during the storage period, it is possible to conclude that Enterobacteriacea was associated particularly to the loss of freshness in VP, and Pseudomonas spp. in WP samples.

Presence of B. cereus, although not compromising the safety of MA, once the infective dose for both diarrheal and emetic syndrome is clearly higher, being the reference dose reported of >104 cells g-1 food to be considered hazardous (Arnesen, Fagerlund and Granum, 2008), was probably associated to the contamination of the raw materials and/or to the insufficient heat treatment to inactivate the spores. However, the safety of the product is still not compromised, once these bacteria failed to grow at temperatures as low as 4 ºC. Therefore, rigorous hygienic conditions during and post-heat treatment followed by refrigeration storage temperature control during commercialization, including information to consumers, must be implemented to avoid temperature abuse and consequent potential growth of psychrotrophic and, specially, mesophilic pathogenic bacteria. This measure should be reinforced with others in a preventive way, namely a rigorous control in rice and spices used in the product formulation.

The traditional storage and distribution procedure of MA from Monchique region, without packaging, not only compromises the shelf life of this difficult to preserve RTE blood sausage, which was evidenced by the sensory and microbiological results, but also contributes to considerable economic losses for the producers not only due to product recall but also by compromising the commercialization period and the conquest of more distant markets.

The introduction of VP represents an improvement in the shelf life: MA maintained good sensorial characteristics during storage, however, it losses the freshness mainly due to the growth of LAB and Enterobacteriaceae. MAP using 80%

CO2 and 20% N2, by its inhibition effect on the growth of LAB, Enterobacteriaceae and Pseudomonas spp. and on sensorial rancidity revealed to be the most efficient on extending the shelf life of Morcela de Arroz.

193

CHAPTER 8 ______

From the sensory point of view, VP and MAP had a similar performance for the most of the tested attributes. However, MAP samples were generally more bright (p < 0.05 after day 8), and, for longer storage periods (22 days and after), it presented a higher (p < 0.05) typical flavour and global freshness (slices or packaged).

The use of MAP revealed to be the best solution to increase shelf life of MA, once almost all the sensory traits were maintained during the 44 days of storage.

The package, Combitherm XX, 115 µm thickness film, a polyamide/ethylene- vinyl-alcohol/polyethylene multilayer complex film, indicated for MAP of cooked meat 2 products, with an O2 transmission rate of 3.0 cm3/m .24 h.atm at 23 ºC and 75% of relative humidity (RH), carbon dioxide transmission rate of 13.5 cm3/m2.24 h.atm at 23 ºC and 75% RH and a water vapour transmission rate of 1.0 g/m2.24 h at 23 ºC and 85%

RH (Wipak Walsrode GmbH & Co., Walsrode, Germany) showed a good control of O2,

CO2 and humidity, being a good technological solution for both MAP and VP of MA.

The multivariate model shelf life approach through PCA represented well the pattern of MA spoilage and contributed to validate the adequacy of the lexicon generated to study the spoilage of MA, as that vocabulary is highly discriminate among fresh and spoiled MA, and have a clear relationship to the microbiological and chemical spoilage phenomena that occurs during storage at 4 ± 1 ºC.

The consumer test, considering consumption intention and a 50% probability of a consumer rejecting the product, indicated that WP have a shelf life of 11.6 days (95% CI 10.4-12.9) and VP 27.8 days (95% CI 26.0-29.5).

The variables that were associated to MA freshness were all of sensory nature: global evaluation of freshness, moisten and bright aspect and typical colour, odour and flavour, and were grouped separately from those linked to spoilage. The sensory evaluation of slime amount in the surface, the spoiled odours like off-odours, sour/fermented, spoiled wine/vinegary, mouldy/mildew and rancid, bitter and acid flavour were grouped with the TBA, and a clear association was found between the counts of Pseudomonas spp., the amount of TBVN and the sensory evaluation of putrid odour. Among not accepted samples, there was discrimination between those WP stored and VP stored. The absence of package resulted in samples characterized by high scores in sensory evaluation of spoiled odours and flavours and presence of slime and high levels of TBA, indicating that lipid oxidation had an important contribution to the

194

CHAPTER 8 ______rejection of MA WP. In turn, MA VP stored for longer storage times and not accepted by the consumers, were associated to higher counts of LAB and Enterobacteriaceae, and the respective metabolites produced by these bacteria.

The good predicting ability of the multivariate model used, combining principal component analysis of spoilage indicators with logistic regression to predict the acceptability by consumers when compared to the apparent uselessness of using individual spoilage indicators to predict shelf life, applying the same prediction tool, and corroborated by analysis of variance, highlights the behaviour of the consumer choice based on a multifactorial assessment of an ensemble of attributes rather than in single ones.

Considering the interesting strategy to diversify the target consumers and consequently increase sales, the scenario of the introduction of MA in the market in small portions as thick slices aerobic packed (AP), VP and MAP, was studied considering safety (Listeria monocytogenes challenge test at recommended temperature 3 ± 1 ºC, and refrigeration storage temperature abuse, 7 ± 1 ºC) and sensorial (consumer test) risks.

However, in a temperature abuse scenario (7 ± 1 ºC), as expected, the pathogen presented faster growth rates and reached an increase of nearly 1 log CFU/g in the first 4 days of storage. At 2 days of storage, counts of L. monocytogenes in AP samples were higher (p < 0.05) than those observed in MAP (80% CO2/20% N2) and, VP samples, presented similar intermediate counts (p ≥ 0.05) indicating that an eventual post-heating contamination will accelerate pathogen growth to consumer hazardous levels in the first days of storage if a temperature abuse occurs.

The consumer test revealed that the sensory shelf life of sliced MA at recommended refrigeration storage temperature (3 ± 1 ºC) is limited to 4.4 (3.8-5.1) days for AP, 8.1 (6.8-9.5) days for VP and 10.4 (8.8-12) days for MAP with 95% confidence interval. Therefore, when sliced MA is stored at 3 ± 1 ºC, shelf life should be established based on sensory spoilage once sensory rejection by the consumers

195

CHAPTER 8 ______occurs long before the product is considered hazardous due to the possible growth of L. monocytogenes.

Further studies are, however, necessary, to provide LAB species identification in order to know more about the so called specific spoilage microorganisms and a better understanding of the spoilage process involved in this traditional cooked blood sausage. Additionally, the possibility of application of natural preservatives, alone or together with less conventional preservative processes (new mild technologies), e.g. high pressure (Diez et al., 2008a, 2008b, 2009), which may be more effective in MA preservation without negative changes in sensory attributes, should also be explored.

The new EC Regulation No 1169/2011 of the provision of food information to consumers changes food labeling and includes mandatory nutrition information on processed foods. It shall be applied from 13th December of 2016, and includes energy value, amounts of fat, saturates, carbohydrate, sugars, protein and salt, expressed per 100 g or 100 ml. In addition, the energy value and the amounts of nutrients should be also expressed per portion and/or per consumption unit, easily recognisable by the consumer. This mandatory nutrition declaration may be supplemented with an indication of the amounts of one or more of the following nutrients: mono-unsaturates, polyunsaturates, polyols, starch, fiber and any of the vitamins or minerals listed in point 1 of Part A of Annex XIII and which are present in significant amounts as defined in point 2 of Part A of Annex XIII. This is an important step with respect to more complete and adequate information to consumers about food, enabling them to identify and make appropriate choices that suite their individual dietary needs, thus ensuring a higher health protection. Therefore, future work is needed in order to determine this mandatory nutrition information on Morcela de Arroz of Monchique.

We believe that the objectives of this thesis were accomplished; allowing us to characterize the processing technology, sensory, microbiological and physicochemical aspects of this Portuguese traditional cooked blood and rice sausage and to improve its shelf life through the application of different packaging technologies. MA, by its formulation and processing, is unique among other traditional cooked blood sausages and is part of a historic and cultural patrimony of this region that is important to preserve and about which so little was known.

196

CHAPTER 8 ______

Finally, other meat products of Monchique region need to be studied, namely Farinheira de Milho also named Morcela de Farinha, another type of cooked blood sausage produced in this region.

8.2 REFERENCES

Arnesen, L. P. S., Fagerlund, A. and Granum, P. E. (2008). From soil to gut: Bacillus cereus and its food poisoning toxins. Review article. FEMS Microbiology Reviews, 32, 579-606.

Diez, A. M., Santos, E. M., Jaime, I. and Rovira, J. (2009). Effectiveness of combined preservation methods to extend the shelf life of Morcilla de Burgos. Meat Science, 81, 171-177.

European Regulation No 510 (2006). Council Regulation of 20th March 2006. Official Journal of the European Union.2006, L93, 12-25.

European Regulation No 1151 (2012). Regulation of the European Parliament and of the Council of 21st November 2012. Official Journal of the European Union.2012, L343, 1-29.

European Regulation No 1169 (2011). Regulation of the European Parliament and of the Council of 25th October 2011. Official Journal of the European Union.2011, L304, 18-63.

197

CHAPTER 8 ______

198