Relationships Between Volatile Flavor Compounds, Sensory Descriptors and Consumer Acceptability of American Dry-Cured Ham

Mississippi State University Scholars Junction

Theses and Dissertations Theses and Dissertations

1-1-2007

Relationships Between Volatile Flavor Compounds, Sensory Descriptors And Consumer Acceptability Of American Dry-Cured Ham

Alessandra Julian Pham-Mondala

Follow this and additional works at: https://scholarsjunction.msstate.edu/td

Recommended Citation Pham-Mondala, Alessandra Julian, "Relationships Between Volatile Flavor Compounds, Sensory Descriptors And Consumer Acceptability Of American Dry-Cured Ham" (2007). Theses and Dissertations. 3776. https://scholarsjunction.msstate.edu/td/3776

This Graduate Thesis - Open Access is brought to you for free and open access by the Theses and Dissertations at Scholars Junction. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Scholars Junction. For more information, please contact [email protected]. RELATIONSHIPS BETWEEN VOLATILE FLAVOR COMPOUNDS,

SENSORY DESCRIPTORS AND CONSUMER ACCEPTABILITY

OF AMERICAN DRY-CURED HAM

Alessandra Julian Pham

A Thesis Submitted to the Faculty of Mississippi State University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Food Science and Technology in the Department of Food Science, Nutrition and Health Promotion

Mississippi State, Mississippi

December 2007

RELATIONSHIPS BETWEEN VOLATILE FLAVOR COMPOUNDS,

SENSORY DESCRIPTORS AND CONSUMER ACCEPTABILITY

OF AMERICAN DRY-CURED HAM

Alessandra Julian Pham

Approved:

______Mark W. Schilling Patti C. Coggins Assistant Professor of Food Science Assistant Professor of Food Science (Director of Thesis) (Committee Member)

______J. Mike Martin J. Byron Williams Assistant Professor of Animal and Dairy Science Assistant Professor of Food Science (Committee Member) (Committee Member)

______William B. Mikel Vance H. Watson Graduate Coordinator in the Dean of the College of Agriculture Department of Food Science, and Life Sciences Nutrition and Health Promotion (Committee Member)

Name: Alessandra Julian Pham

Date of Degree: December 15th, 2007

Institution: Mississippi State University

Major Field: Food Science and Technology

Major Professor: Dr. Mark W. Schilling

Title of Study: RELATIONSHIPS BETWEEN VOLATILE FLAVOR COMPOUNDS, SENSORY DESCRIPTORS AND CONSUMER ACCEPTABILITY OF AMERICAN DRY-CURED HAM

Pages in Study: 101

Candidate for Degree of Master of Science

The relationships between volatile flavor compounds, sensory descriptors and consumer acceptability were determined for eight commercial American dry-cured hams using external preference and flavor mapping. The majority of consumers preferred (p<0.05) hams that had more intense caramelized, smoky, savory and molasses aromas as well as more intense sweet and savory flavors. Sixteen aroma impact compounds were identified from the headspace volatiles of dry-cured hams.

The consumers with the highest acceptability scores preferred (p<0.05) hams that were characterized by 4-methyl-2-methoxyphenol (sweet ham), 4-ethyl-2-methoxyphenol

(sweet ham), 2-methoxyphenol (smoky, cocoa), 2,6-dimethoxyphenol (smoky ham, savory) and 2-furanmethanol (burnt meat, vitamin). Fourteen percent of consumers

preferred (p<0.05) two hams that were characterized by methional (baked potato).

Consumer acceptability scores were lower for hams either characterized by methanethiol (sulfur), carbon disulfide (sulfur), 2-butanone (sweet), 3-methylbutanal

(malty, fermented), 2-heptanone (burnt meat, vitamin), hexanal (cut grass), benzeneacetaldehyde (floral), 1-octen-3-ol (mushroom) or characterized by benzaldehyde (burnt meat, cooked meat) and limonene (citrus).

DEDICATION

To my parents, Dr. Chay and Dr. Laura Pham, whose love, guidance and support have made me who I am today. To God, Almighty and All Knowing, for without Him I am nothing, to Him belongs all glory forever.

ACKNOWLEDGEMENTS

My deepest gratitude goes to the following people and institutions for making this thesis possible. To my advisor, mentor and one of the best persons I have ever known, Dr. Wes Schilling. Your guidance, support and encouragement have made this scholarly journey such a rewarding experience and your thoroughness and dedication have made this research what it is. Thank you for always believing in me and for your words of appreciation even in the smallest things accomplished. I hope everyone gets to be blessed with the kind of advisor that I have had in you. Thank you Dr. Benjy Mikel for the time you have allotted in reviewing my work and for the constructive criticisms in the improvement of my results, your expertise as one of the best dry-cured ham experts has been truly invaluable to this research work. Dr. Patti Coggins has added strength to our research through her expertise in sensory evaluation, I appreciate the time she has graciously given throughout this undertaking. Special thanks go to Dr.

Mike Martin for his helpful suggestions on my manuscript, for lending his experience to this endeavor and for teaching me to solve problems with more practicality. I am also grateful to Dr. Byron Williams for his meticulous editing of my manuscript and valuable suggestions in improving the understanding of my results.

To my esteemed colleagues at the Food Chemistry/Muscle Foods Laboratory, it is with gratitude that I acknowledge the various contributions of Dr. Youngmo Yoon,

iii

Vi Jackson, Vijay Radhakrishnan, Raman Sekhon-Grewal, Vimal Kamadia and Vamsi

Battula. Thank you all for your commitment and dedication to our research group.

Appreciation is given to Timothy Armstrong, Joseph Andol, Mary Andol, Coy Farley and Donna Bland for making sure I had all the resources I needed to carry out this research work.

Deep gratitude is given to the Department of Food Science, Nutrition and Health

Promotion, my fellow graduate students, wonderful people I’ve gotten to know in

Starkville, dear friends Stellar and Hyejin and best friends Pier, Maebeth and Tina for never failing to cheer me on. To my extended family Timi, Aunt Alma, Tita Lina, Tito

Andy, Anna and Adrian for your constant encouragement and support.

I would like to thank Andro, my inspiration, for all the love, support, sacrifice and encouragement you have shown me as I completed this research work. You have filled my life with happiness and I can never thank God enough for giving you to me. I would also like to thank my sister Trixie, my idol, who I followed all the way to graduate school for all the love, support and guidance she has shown me. I would never have survived this journey if we weren’t in it together and for that alone I will forever be indebted to her. To my parents, Dr. Chay and Dr. Laura Pham, no amount of words can express my gratitude for all the love, support, sacrifice, patience and guidance you have shown me. Thank you for instilling in me the values which you have and for making me who I am today. I am proud to be your daughter and thankful for having the best parents anyone could ever hope for. To the Lord God Almighty, who has made everything possible, I will always be humbled by His tremendous insurmountable love.

TABLE OF CONTENTS

DEDICATION...... ii

ACKNOWLEDGEMENTS...... iii

LIST OF TABLES...... vii

LIST OF FIGURES ...... viii

CHAPTER

I. INTRODUCTION...... 1

II. LITERATURE REVIEW...... 5 History of American Dry-cured Ham ...... 5 Production of American Dry-cured Ham...... 6 Volatile Flavor Compounds in American Dry-cured Ham...... 8 Generation of Volatile Flavor Compounds in Dry-cured Ham ... 14 Volatile Flavor Compounds Generated by Degradation of Amino Acids ...... 15 Volatile Flavor Compounds Generated by Lipid Oxidation...... 17 Volatile Flavor Compounds Generated by Reactions Between Amino Acids and Other Compounds...... 18 Volatile Flavor Compounds Generated from the Manufacturing Process and Raw Materials ...... 19 Descriptive Language for Sensory Evaluation of Dry-cured Ham...... 20 Extraction of Volatile Flavor Compounds in Dry-cured Ham .... 23 Analysis of Volatile Flavor Compounds in Dry-cured Ham ...... 24 Gas Chromatography Olfactometry (GCO) ...... 24 Gas Chromatography Mass Spectrometry (GCMS) ...... 26

Relationships Between Volatile Flavor Compounds, Sensory Descriptors and Consumer Acceptability ...... 27 Principal Components Analysis (PCA) ...... 29 Cluster Analysis ...... 30 External Preference Mapping ...... 31

III. MATERIALS AND METHODS...... 33

Materials ...... 33 Sample...... 33 Chemicals...... 34 SPME (Solid Phase Microextraction) Fiber ...... 35 Methods...... 35 Sample Preparation ...... 35 Descriptive Sensory Analysis ...... 36 Consumer Acceptability...... 38 Conditioning of SPME Fiber for Chromatographic Analysis...... 39 Isolation of Volatile Compounds...... 39 Gas Chromatography Mass Spectrometry (GCMS) ...... 40 Gas Chromatography Olfactometry (GCO)...... 41 Identification of Aroma Impact Compounds ...... 42 Statistical Analysis...... 43

IV. RESULTS AND DISCUSSION...... 45

Descriptive Analysis ...... 45 Aroma Impact Compounds...... 52 Consumer Acceptability...... 62 External Preference and External Flavor Mapping: Relating aroma impact compounds, sensory descriptors and consumer acceptability…….……...... 68

V. SUMMARY AND CONCLUSIONS ...... 79

REFERENCES ...... 81

APPENDIX...... 93

LIST OF TABLES

TABLE Page

1 Volatile compounds previously identified in American dry-cured ham.……………………………………………………………….. 11 2 Main groups of volatile compounds generated during the processing of dry-cured ham (Toldrá, 2004).……………………………………….. 16 3 Dry-cured hams from different regional locations and with varying aging periods……………………………………………………… 34 4 Sensory descriptors and definitions for American dry-cured ham.……. 37

5 Mean sensory ratings for the flavor, aroma and textural attributes of eight commercial brands of American dry-cured hams…...……… 46 6 Aroma impact compounds identified and quantified in the headspace of American dry-cured hams using SPME-GCMS, SPME-Osme- GCO and SPME-GCFID………………………………………….. 53

7 Retention indices for aroma impact compounds detected in the headspace of American dry-cured ham during SPME-GCMS, SPME-Osme-GCO and SPME-GCFID…………………………... 55

8 Mean scores for consumer acceptability (N=71) of eight American dry-cured ham treatments with varied aging periods and processing conditions …………………………………………….. 63

9 Mean scores for overall consumer acceptability of eight American dry- cured ham treatments according to different clusters of consumer segments using a hedonic scale…………………………………… 67

vii

LIST OF FIGURES

FIGURE Page

1 Principal Component Analysis Biplot of descriptive analysis data for flavor attributes of American dry-cured ham……………………... 47

2 Principal Component Analysis Biplot of descriptive analysis data for aroma attributes of American dry-cured ham…………………….. 49

3 Principal Component Analysis Biplot of descriptive analysis data for textural attributes of American dry-cured ham…………………… 51

4 Comparison of the GCFID chromatogram (top) with the identification of some aroma impact compounds against GCO Osme time- intensity Osmegram (bottom) for American dry-cured ham……… 56

5 Principal Component Analysis Biplot of the aroma impact compounds of eight American dry-cured hams by SPME-GCMS.. 61

6 External Preference Map of combined consumer data with descriptive analysis data for flavor attributes of eight American dry-cured hams...... 64

7 External Preference Map of combined consumer data with descriptive analysis data for aroma attributes of eight American dry-cured hams………………………………………………………………. 65

8 External Preference Map of combined consumer data with descriptive analysis data for textural attributes of eight American dry-cured hams. ……………………………………………………………... 70

9 External Preference Map of combined consumer data with the volatile composition of eight American dry-cured hams…………. 72

A.1 Score Sheet for Descriptive Sensory Analysis………………………... 94

viii

A.2 Score Sheet for Consumer Acceptability Tests……………………….. 99

A.3 Dendrogram showing how panelists were grouped together into clusters based on dissimilarity of consumer acceptability scores for American dry-cured ham……………………………………… 101

CHAPTER I

INTRODUCTION

Country style or country-cured ham refers to American dry-cured ham, a value- added product characterized by unique and aged flavors. This product is a valued tradition in the Southern United States with regional variations in its curing, smoking and aging stages that result in subtle differences in the final flavor of the meat product

(Marriott and Ockerman, 2004). It is during the aging period that the hams are subjected to various time-temperature-humidity treatments which can last between 90 to 360 days, depending on the final quality and desired product flavor. Sufficient aging time is necessary for the development of characteristic flavors that enhance the product's consumer appeal (Lowery, 1986; Marriott and Ockerman, 2004).

Flavor and aroma are essential parameters for the overall acceptance of dry- cured hams and are markedly affected by raw material, processing techniques, and aging time (Ockerman et al., 1964; Shahidi, 1998; Dirinck et al., 1997; Sánchez-Peña et

al., 2005). The flavor and aroma of dry-cured ham can be determined by sensory

descriptive analysis and the composition of aroma impact compounds, most of which

are produced post-mortem by chemical and enzymatic mechanisms (Flores et al., 1997).

Several studies have been conducted to identify and quantify the volatile compounds in

different types of dry-cured hams such as Iberian (Timón et al., 1998; Ruiz et al., 2001;

Carrapiso et al., 2002), Serrano (Flores et al., 1997; 1998), Parma (Barbieri et al., 1992;

Careri et al., 1993; Bolzoni et al., 1996), Black Forest (Wittkowski et al., 1992), Jinhua

(Huan et al., 2005; Zhang et al., 2006) and French hams (Berdagué et al., 1993;

Buscailhon et al., 1994). Compared to the aforementioned dry-cured ham products, no

recent studies (Ockerman et al., 1964; Lillard and Ayres, 1969; Piotrowski et al., 1970)

have been published on the volatile profile of American dry-cured hams.

A considerable amount of research on dry-cured ham volatiles has focused on

identification and quantification with limited approach to their true contribution to dry-

cured ham aroma and flavor. More than 260 compounds have been identified in dry-

cured hams (Flores et al., 1998). However, it has been established that not all volatile compounds actually contribute to the overall aroma of food. In addition, some volatile compounds are present at very low concentrations such as ppm or ppb levels which may be too low for their identification using gas chromatography-mass spectrometry

(GCMS) alone (Qian et al., 2007). Gas chromatography olfactometry (GCO) is an

instrumental tool that is utilized for the identification of aroma impact compounds in

foods (Marsili, 2007). This instrument has been successful at characterizing the aroma

impact compounds in Serrano (Flores et al., 1997), Iberian (Carrapiso et al., 2002) and

Parma (Blank et al., 2001) dry-cured hams through identification of the volatile compounds that contribute to the overall flavor of these products. A study on the volatile composition of American dry-cured ham is important because it can help relate flavor compounds to sensory descriptors and consumer preference as well as help monitor flavor quality.

To fully comprehend the nature of dry-cured ham flavor, a sensory language

(lexicon) must be established through sensory descriptive analysis in order to differentiate and describe dry-cured ham products based on their flavor, aroma, and textural attributes (Flores et al., 1997; Ruiz et al., 1998; Armero et al., 1999). Previous studies have related chemical information to the sensory properties of dry-cured ham flavor (Berdagué et al., 1993; Careri et al., 1993; Buscailhon et al., 1994; Bolzoni et al.,

1996; Flores et al., 1997). However, none of these studies identified which volatile flavor compounds or sensory attributes characterize certain dry-cured ham products and how intensity and presence of volatile compounds and/or sensory attributes relate to consumer acceptability. Aging time was identified as the major determinant of the flavor and aroma attributes in Serrano (Flores et al., 1997; 1998) and Iberian (Ruiz et al,

1998) dry-cured hams. A lexicon used to describe Serrano ham was grouped into three

factors such as “cured flavor”, “off-flavor” and “pork flavor” and indicated higher

intensities for these factors in its longer aged hams. Iberian hams that were ripened by

the longer aging process also had higher intensities for “cured flavor”, “aftertaste” and

“flavor strength” when compared to short aged hams.

The objective of this research was to determine the relationships between

volatile flavor compound composition, objective sensory descriptors and consumer

acceptability of American dry-cured ham. This objective was met through identifying

the volatile and the aroma impact compounds in American dry-cured ham using a

combination of instrumental methods including solid phase microextraction (SPME)-

GCMS and SPME-Osme-GCO, identifying the sensory characteristics that describe and

differentiate American dry-cured ham, and utilizing preference and flavor mapping to explain the relationships between volatile flavor compounds, common sensory descriptors, and consumer acceptability.

CHAPTER II

LITERATURE REVIEW

History of American Dry-cured Ham

The origin of curing meat is lost in antiquity. It is believed that meat preservation was first practiced by the Sumerians in about 3000 B.C. (Jay, 1970). The original purpose of curing was to preserve meat when food was scarce. Dry-cured hams are still consumed in the United States because they have a distinct flavor and aroma that can only be obtained from the dry curing process.

In the United States, country style or country-cured ham is also known as

American dry-cured ham which is a traditionally popular meat product characterized by unique flavors that meets niche markets. By definition, true American dry-cured ham is cured with a dry salt cure, loses at least 18% of its original weight in curing and contains a minimum of 4% salt (USDA, 1999). It is undoubtedly one of the oldest traditions in the nation, and is still practiced across the South where dry-curing is considered a way of life (Lowery, 1986). The American ham-curing process has its roots in Virginia where the first hams were said to be cured in the Jamestown colony when the colonists adopted the curing process for preserving game used by the

Warascoyak Indians (Marriott and Ockerman, 2004). What started out as family tradition has evolved into a business venture for most of the nation’s dry-cured ham

processing facilities where techniques have been passed from generation to generation.

Some of the established ham-curing states include Virginia, Kentucky, Tennessee,

North Carolina, Georgia, and Missouri where the year round weather conditions meet

the requirements for curing ham (Marriott and Ockerman, 2004).

The commercial economic importance of this product is continually increasing

with over 5.5 million hams processed by the U.S. dry-cured ham industry in 1998 with a

retail value exceeding 200 million dollars (Stalder et al., 2003). According to the

National Country Ham Association (NCHA), it is not easy to track hams manufactured or sold by non-members which may outnumber those marketed by members of the association (Voltz and Harvell, 1999).

Production of American Dry-cured Ham

The dry-curing process consists of applying the cure ingredients in their dry, granular form on the ham, picnic or belly surface (Marriott and Schilling, 2004). The dry-curing process involves several stages: cure application, cure equalization, smoking which is an optional process and aging (Marriott and Ockerman, 2004). The most important curing ingredients are salt and nitrates and/or nitrites. Optional ingredients such as sugar and pepper can be included to impart distinct sensory properties to the end product. The typical ham curing procedure results in the application of about 8.0% salt,

3% sugar and 2,180 ppm of sodium nitrate where processors can choose to supplement sodium nitrate with up to 140 ppm of sodium nitrite (Marriott and Schilling, 2004). The main role of salt in the dry-curing process is to act as a bacteriostatic agent by inhibiting the growth of spoilage microorganisms through reduction of the available water (aw) 6

(Toldrá et al., 1997). Moreover, it also affects the flavor of the product by imparting a characteristic salty taste and increases the solubility of myofibrillar proteins. Nitrate is reduced to nitrite by nitrate reductase which is a bacterial enzyme present in the natural flora of ham (Toldrá, 2002). Nitrite functions as a preservative by inhibiting the growth of undesirable microorganisms, specifically Clostridium botulinum (Cassens, 1995).

Further reduction to nitric oxide, which is supported by the slightly acidic pH found in

the ham, results in the development of the cured meat color when it reacts with the meat

pigment myoglobin. Sugar improves the meat flavor and can accelerate the cure

process. It helps counteract the harshness of salt and caramelizes with the heat during

cooking to enhance the flavor.

Application of the cure ingredients is carried out for approximately two days per

pound of uncured weight (40-50 days) at 2-4oC. The external Semimembranosus muscle

is first penetrated by the cure ingredients followed by their slow diffusion through the

rest of the whole ham (Toldrá and Flores, 1998). The temperature is maintained below

o 4 C until a water activity (aw) of less than 0.96 is attained to ensure that microbial growth is retarded. The residual ingredients are wiped, brushed or washed from the product. This is followed by cure equalization where the product is kept for 14 days at

10-13oC to enable a more uniform distribution of the cure (Marriott and Ockerman,

2004). During cure equalization, salt content decreases on the surface because it

diffuses toward the inner part of the ham. Salt crystallization and the growth of

microorganisms which can produce off flavors are prevented by maintaining a relative

humidity that is slightly higher than 75%. The “sniff test” is conducted to check for

rotten or “bone-sour” hams where an experienced worker punctures the ham with an ice

pick, then sniffs the pick to detect any “off” odor. Cure equalization is followed by

smoking which is an optional step. Smoking (1 day, 31-33oC) is one of the oldest preservation techniques and may be applied continuously or intermittently in a smokehouse. It improves the color and flavor of the ham and protects its surface from mold or yeast growth due to the bactericidal and bacteriostatic effect of smoke compounds (Toldrá, 2002). The trademark hardwoods for smoking among ham producers include apple, maple, pecan, hickory and even sassafras. The last step of the dry-curing process is aging at 29-32oC for 90 to 360 days. A key contributor to the development of American dry-cured ham’s unique flavor is the length of processing time where an extended aging period is needed to have a stronger and more “cured” flavor and aftertaste. During this step, the temperature is slowly increased and the relative humidity is lowered to below 75% to diminish mold growth and to protect against excess drying by facilitating oil drip as well as to enhance flavor development

(Marriott and Schilling, 2004). Higher quality hams are usually sold as an entire piece or boned. Vacuum packed boned hams are distributed through retailers for final cutting into pieces or slices. Commercial distribution of vacuum packed slices is increasing

(Marriott and Ockerman, 2004).

Volatile Flavor Compounds in American Dry-cured Ham

Flavor and aroma are valuable characteristics that impact the overall acceptance of American dry-cured ham and are determined by volatile compound composition.

Although there are only a few reports dealing with the volatile constituents of American 8

dry-cured ham (Ockerman et al., 1964; Lillard and Ayres, 1969; Piotrowski et al.,

1970), these published works were instrumental in determining the compounds responsible for dry-cured ham flavor. Much of the research by these scientists was carried out before the development of gas chromatography mass spectrometry (GCMS).

In the last five decades most of the extensive research published on this product focused on their nonvolatile constituents and their relation to the flavor of the product (McCain et al., 1968; Lillard and Ayres, 1969), influence of different quality factors on flavor

(Eakes and Blumer, 1975a; 1975b; Kemp et al., 1979), acceleration of the dry-curing process (Marriott et al., 1983; 1987; 1992) and shelf-life and stability (Fang et al., 1997;

Portocarrero et al., 2002). Compared to these research works, relatively few studies have been published on the volatile profile of American dry-cured ham and analyses performed to identify its odorants have been scarce.

Several studies have been conducted to identify and to quantify the volatile compounds in different types of dry-cured hams such as Iberian ham (Timón et al.

1998; Ruiz et al., 2001; Carrapiso et al., 2002), Serrano ham (Spanier et al., 1997;

Flores et al., 1998), Parma ham (Barbieri et al., 1992; Careri et al., 1993; Bolzoni et al.,

1996), French ham (Berdagué et al., 1993; Buscailhon et al., 1994), Black Forest ham

(Wittkowski et al., 1992) and Jinhua ham (Huan et al., 2005; Zhang et al., 2006). More

than 260 compounds have been identified and aldehydes, alcohols and esters were the

most important organic classes that can be utilized to differentiate between Italian,

French and Spanish hams (Flores et al., 1998). Compared to the aforementioned dry-

cured ham products, relatively few studies (Ockerman et al., 1964; Lillard and Ayres,

1969; Piotrowski et al., 1970) have been published on the volatile profiles of American

dry-cured hams. Studies on other dry-cured ham products are not automatically

applicable to American dry-cured ham due to differences in processing conditions,

antemortem and postmortem factors and use of various analytical techniques in the

elucidation of their volatile composition. In addition, American dry-cured ham may or

may not undergo the smoking process and is usually consumed cooked. A study on the

volatile composition of American dry-cured ham is essential since it can help relate

flavor compounds to sensory descriptors and consumer preference as well as help

monitor flavor quality.

Ockerman and coworkers (1964) isolated a number of aldehydes, ketones,

alcohols, carboxylic acids, bases and sulfur compounds (Table 1) in an attempt to

characterize the volatile flavor compounds of American dry-cured ham. Extraction was

performed by vacuum distillation at 60oC for 24 h with precautions that a higher distillate temperature would produce different compounds than what have been identified. Tentative identification of the compounds was performed by comparison of their gas chromatography retention time with those of standards and further verification of the compounds by infrared spectroscopy. The researchers noted the importance of the identified aldehydes and ketones to the aged flavor of the product as their levels increased with aging time. A notable increase was found for 2-butanone especially in the later part of the aging period. The total quantity of carbonyl derivatives obtained from the hams increased with aging time. This is in agreement with other works which reported that conditions that were favorable to oxidation yielded an increase in carbonyl

Table 1. Volatile compounds previously identified in American dry-cured ham.

Aldehydes 2,4-heptadienalc Carboxylic acids Ethyl formateb Methanalac 2,6-nonadienalc Formic acida Ethyl butyrateb Ethanalac 2,4-decadienalc Acetic acida Ethyl hexanoateb Propanalac 2,4-undecadienalc 2-propenoic acida Propenala 2,4-dodecadienalc Propanoic acida Sulfur compounds 2-methylpropanala Isobutyric acida Hydrogen sulfidesa Butanalabc Ketones Trimethylacetic acida 3-methylbutanala Acetonea Methacrylic acida Miscellaneous 2-Butenala 2,3-butaedioneab Butyric acida Ammoniaa Pentanalabc 2-butanoneab Isovaleric acida Methylaminea Heptanalbc 3-methyl-2-butanonea Crotonic acida Hexanalc 2-pentanoneab Valeric acida Octanalc 3-pentanonea Dimethylacrylic acida Nonanalc 2,6-hexanedioneb Isocaproic acida Decanalc Caproic acida Undecanalc Alcohols Dodecanalc Ethanola Esters 2-hexenalc Methanolb Methyl acetateb 2-heptenalc Propanolb Methyl propionateb 2-octenalc Hexanolb Methyl hexanoateb 2-nonenalc Heptanolb Methyl octanoateb 2-decenalc Octanolb Methyl decanoateb 2-undecenalc 2-octanolb Methyl laurateb 2-dodecenalc a Volatile flavor compounds identified by Ockerman et al. (1964) using vacuum distillation and a 2-meter diisodecyl phthalate gas chromatographic column. b Volatile flavor compounds identified by Lillard and Ayres (1969) in ham diffusates using steam distillation under reduced pressure and a 1.8-meter gas chromatographic Carbowax 20M column. c Volatile flavor compounds identified by Lillard and Ayres (1969) in ham lipids using steam distillation under reduced pressure and a 1.8-meter gas chromatographic Carbowax 20M column.

compounds. Similarly, the number and quantity of carboxylic acids isolated were found to increase with the length of the aging period, except for formic acid. Ockerman and coworkers attributed this to the fact that formic acid is a constituent of wood smoke, and some decrease in acid from this source may have occurred especially since wood smoke

residue is deposited primarily on the meat surface. Selective trapping allowed for the identification of hydrogen sulfide and trace amounts of disulfides and/or monosulfides.

Methylamine and ammonia were also reported as basic compounds that contribute to the flavor of American dry-cured ham. Analysis of the wood smoke revealed retention times characteristic of methanal, ethanal, acetone, 2-butanone, 3-methylutanal, pentanal and formic ester.

Lillard and Ayres (1969) showed the presence of various aldehydes, ketones, alcohols and esters in the volatile fraction and lipids of American dry-cured ham (Table

1). The volatiles were isolated by steam distillation under reduced pressure, collected on a trap cooled with dry ice and analyzed by injecting the ether extract of the distillate into a gas chromatograph. The authors noted that some of the volatile carbonyls reported by Ockerman et al. (1964) were not found due to the treatment of the distillate before gas chromatographic analysis. The carbonyl compounds identified in the lipid fraction of American dry-cured ham have been previously associated with the oxidation products of pork fat or have been found in the autooxidation studies of other lipids. The researchers have attributed these compounds to the fatty acid composition of the lipids which are influenced by the diet of the pigs. Further analysis of the free fatty acids and free amino acids showed that the degree of proteolysis and lipolysis varied among hams produced by different processors and were dependent upon temperature and aging period. Lillard and Ayres (1969) also showed that the distillate had an aroma typical of

American dry-cured ham and further proposed that the flavor and flavor precursors of this product were not water soluble. Cooking also intensified the aroma and flavor of

the hams suggesting that the heating process contributes to the reactions which produce the American dry-cured ham flavor.

Piotrowski et al. (1970) analyzed various fractions of American dry-cured ham to identify the cured aroma and to look into the changes occurring in pork during curing, cooking and smoking. Extraction of the ham samples with a mixture of chloroform-methanol (2:1, v/v) yielded a lipid fraction with cured, smoky aromas.

Although the aqueous extracts yielded cured, hammy aromas, these aromas were more intense in the chloroform-methanol soluble fractions, suggesting that the cured aroma components or precursors were essentially in the lipid phase. Their results indicated that the precursors for the basic meaty aroma were water soluble, while the cured aroma was related to the lipid fraction. The volatile compounds derived from American dry-cured ham diffusates and lipid extracts were separated by gas chromatography and their aromas were described as they emerged from the chromatographic column. According to the authors, most compounds of intermediate volatility which eluted between 10 to

25 minutes had green, plant-like aromas whereas the volatile components of the lipid fraction were described as waxy, fatty, fried, green and aromatic. As in the case of the ham diffusates, no single compound was responsible for the cured, meaty or smoky aroma.

Smoking is an optional process in the manufacture of American dry-cured ham but may be responsible for some of the sensory properties that make it a valued product.

Baloga et al. (1990) developed a technique for utilizing an atomic emission detector for the selective detection of nitrogen-, oxygen- and sulfur- containing compounds in an

extensive gas chromatographic analysis for ham flavor. However, their research was conducted on a cured, precooked premium ham sample where they used a Likens-

Nickerson apparatus to isolate the volatile flavor compounds. They have detected and associated several phenolic compounds with the smoky aroma of cured ham which demonstrates the contribution of smoking to the dry-cured ham flavor. Eighteen phenolic compounds were identified where 2-methoxyphenol and 4-methyl-2- methoxyphenol were the predominant compounds and most likely make a major contribution to the smoky character of cured ham. These two compounds along with o- cresol, m-cresol, 4-ethyl-2-methoxyphenol and 2, 6-dimethylphenol contributed 72% of the phenolic content in the ham samples or 7.5% of the total volatile composition on the basis of the Flame Ionization Detection (FID) chromatogram.

Generation of Volatile Flavor Compounds in Dry-cured Ham

Dry-cured ham is a value-added meat product dependent upon its distinct flavor and aroma. Some of the most valued dry-cured ham products include the Spanish

Iberian and Serrano hams, Italian Parma and San Danielle hams, French Bayonne and

Corsican hams, German Katenschinken, Black Forest and Westphalia hams, Chinese

Jinhua and Yunnan hams, Finnish Sauna ham and U.S.A. Country style ham (Toldrá,

2002). These hams are usually eaten raw requiring no additional smoking or cooking except for the Country style and Westphalia hams. Various factors affect and modify the overall quality of these products. The raw materials are affected by pig breed, age of slaughter, feed type and processing conditions such as aging period, salt content and

smoking and impact the final flavor, aroma and texture as well as the homogeneity of

the whole ham (Carrapiso and Carrapiso, 2005).

The development of flavor and aroma in dry-cured ham is a complicated process

which involves a number of reactions such as chemical or enzymatic oxidation of

unsaturated fatty acids and further interactions with proteins, peptides, and free amino

acids (Flores et al., 1998; Flores et al., 1997; Toldrá, 2002). These reactions generate a variety of volatiles and precursors (Table 2). More than 260 volatile compounds have been reported in dry-cured ham products and some of the most important are hydrocarbons, aldehydes, alcohols, ketones, esters, furans, pyrazines, lactones and sulfur compounds. The characteristic aroma and odor threshold of each compound contributes to the final flavor and aroma of the product. Differences in the quantified volatiles among dry-cured hams have been attributed to the fact that this is not a homogenous product and that the muscles and subcutaneous fat of the samples may differ (Sánchez-Peña et al., 2005).

Volatiles Flavor Compounds Generated by Degradation of Amino Acids

In addition to being known as meat flavor precursors, free amino acids and

peptides are also important sources of a number of volatile compounds found in dry-

cured hams. These precursors give rise to volatile flavor compounds such as methyl

branched aldehydes, sulfur compounds, alcohols and pyrazines via Strecker degradation

(Ventanas et al., 1992). It has been established that changes in flavor traits can be

related to changes in the amount of amino acids during the aging period. An intense

proteolytic activity results in an increased concentration of free amino acids that serves 15

Table 2. Main groups of volatile compounds generated during the processing of dry- cured ham (Toldrá, 2004).

Groups of volatile Main origin General characteristic compounds aromas Aliphatic Autooxidation of lipids Alkane, crackers hydrocarbons Aldehydes Oxidation of free fatty acids Green, pungent, fatty Branched aldehydes Strecker degradation of amino Roasted cocoa, cheesy- acids green Alcohols Oxidative decomposition of lipids Medicinal, onion, green, alcoholic Ketones -keto acid decarboxylation or Buttery, floral, fruity fatty acid -oxidation Esters Interaction of free carboxylic Fruity acids and alcohols Nitrogen compounds Maillard reaction of amino acids Meaty, nutty, toasted nuts with carbohydrates Sulfur compounds Sulfur-containing amino acids Dirty socks Furans Sulfur-containing amino acids Ham-like, fishy with carbohydrates

as a pool for the Strecker reaction (Toldrá et al., 1997; Flores et al., 1998). Higher levels

of free amino acids during the long curing process in Spanish Serrano dry-cured ham

has been attributed to the action of exopeptidases especially allanyl amino peptidase and

aminopeptidase B which have been characterized from porcine skeletal muscle (Flores

et al., 1996; 1998). In American dry-cured ham, significant increases were observed for

ninhydrin positive material, serine, glutamic acid, threonine, leucine and isoleucine (not

separated), valine, phenylalanine, praline, tyrosine, alanine, glycine and histidine during

successive aging periods (McCain et al., 1968).

Methyl branched aldehydes that have been identified as important contributors

to the flavor development of various dry-cured hams (Dirinck et al., 1997; Sabio et al.,

1998) include 2-methylpropanal, 2-methylbutanal and 3-methylbutanal which arise from the Strecker degradation of their corresponding amino acids such as valine, isoleucine and leucine. Sulfur compounds (thiols) derived from the Strecker degradation of sulfur containing amino acids (methionine and cysteine) (Barbieri et al., 1992) are considered significant flavor contributors due to their low odor thresholds.

Volatile Flavor Compounds Generated by Lipid Oxidation

An intense lipolysis occurs during the dry-curing process, resulting in the production of a high amount of free fatty acids which can undergo further oxidative reactions. During lipolysis, triacylglycerols are broken down by lipases and phospholipids by phospholipases to generate free fatty acids (Toldrá, 2002; 2004). The beginning of lipid oxidation corresponds with the development of the characteristic dry- cured ham flavor although an excess may lead to the formation of off-flavors

(Buscailhon et al., 1994). Aliphatic hydrocarbons, aldehydes, ketones and alcohols are formed from the lipid oxidation of unsaturated fatty acids while esters arise from the interaction of the free fatty acids with alcohols.

Aliphatic hydrocarbons are secondary products from the autooxidation of lipids, but have not been identified as contributors to dry-cured ham flavor. Alcohols may be derived from the oxidative decomposition of certain lipids. For instance, 1-butanol may be derived from myristoleic acid. Because unsaturated alcohols such as 1-penten-3-ol

(onion-toasted) and 1-octen-3-ol (mushroom-like) have lower threshold values when compared to saturated alcohols, these compounds have a greater contribution to the flavor of the product. There were higher concentrations of unsaturated alcohols in 17

French Corsican hams when compared with Serrano, Iberian, Parma, Bayonne and

Light Italian Country hams (Sabio et al., 1998). Aldehydes appear to be the major class of volatiles in the headspace chromatograms of most dry-cured ham products in a study carried out by Dirinck et al. (1997). Linear saturated, unsaturated and polyunsaturated aldehydes are typical fat degradation compounds formed by lipolysis autooxidation mechanisms (Belitz and Grosch, 1987). The low odor thresholds of aldehydes allow them to significantly contribute to the flavor of the product even in trace amounts.

Ketones are -keto acid decarboxylation or -oxidation products of fatty acids

(Berdagué et al., 1991) and have been considered to provide buttery, spicy, floral and fruity aromas. French dry-cured hams have been characterized for their high content of ketones (Buscailhon et al., 1994; Berdagué et al., 1993) compared with other dry-cured ham products. Esters impart fruity and sweet caramel-like aromas to dry-cured ham

(Flores et al., 1998). There are higher concentrations of esters in Italian dry-cured ham

(Barbieri et al., 1992) when compared to Spanish (Lopez et al., 1992) and French

(Berdagué et al., 1991) dry-cured hams. The manufacture of Italian dry-cured ham does not involve the addition of nitrates or nitrites. Therefore, the inhibitory effect on lipid oxidation due to the addition of these cure ingredients may be responsible for the higher concentration of esters in the said dry-cured ham product (Flores et al., 1998).

Volatile Flavor Compounds Generated by Reactions Between Amino Acids and Other Compounds

The reaction of amino acids with other compounds represents another source of volatile flavor compounds. Pyrazines are generated from the Maillard reaction between

amino acids and sugars and have been established as contributors to the nutty and

roasted aromas of cooked foods (McGorrin, 2007). Methylpyrazine and

2,6-dimethylpyrazine have been identified in the headspace of Serrano dry-cured hams

as providing nutty and toasted nut aromas to the product (Flores et al., 1998). Pyrazines

are extensively produced during cooking, but few pyrazines have been identified since

dry-cured ham products are not commonly cooked prior to volatile analysis (Flores et

al., 1998). In the case of American dry-cured ham, the early works of Lillard and Ayres

(1969) have shown that cooking generally resulted in an increase in free amino acids

which intensified cured flavor even though no pyrazines were identified. Furans may

also be generated from the reaction of sulfur containing amino acids with

carbohydrates. Its importance in the unique Iberian ham flavor has been noted by Ruiz

and coworkers (1999) due to its higher content in this product when compared to other

dry-cured hams. Furans have also been identified in the headspace of Serrano dry-cured

ham and as contributors to the overall odor of boiled and roasted meats (Flores et al.,

1997; 1998).

Volatile Flavor Compounds from the Manufacturing Process and Raw Materials

The presence of limonene as well as other terpenes in dry-cured ham products

(Sabio et al., 1998; Sánchez-Peña, 2005) has been associated with the unsaponifiable fraction of vegetable fat. According to Buscailhon et al., (1994), these compounds may

originate from animal feed and accumulate in the fat of the animal. It has also been

suggested by Sabio et al. (1998), that the high concentration of some terpenes in

Bayonne and Corsican dry-cured hams may be due to the black pepper treatment applied on the surface of the these products during processing.

Phenolic compounds have been associated with the smoking of meat products.

Five phenolic compounds have been identified by Poligné and coworkers (2002) in

Boucané which is a smoked cured pork belly from Reunion, France. These researchers have associated the smoked odor with 2,6-dimethoxyphenol and the smoked taste with

2-methoxyphenol. Previous studies have suggested that phenolic compounds tend to give an incomplete smoked, cured aroma and that a complex mixture of substances was necessary for a whole smoke aroma (Daun, 1972).

Descriptive Language for Sensory Evaluation of Dry-cured Ham

Descriptive analysis is a sensory method used by human subjects who have been specifically trained for the purpose of identifying and quantifying the attributes of a food or product (Hootman, 1992; Meilgaard et al., 1999). The analysis may be focused on certain aspects such as flavor, aroma, texture and aftertaste or it may encompass all the parameters of the product. Variations of this method include the Flavor Profile

Method (FPM), Quantitative Descriptive Analysis (QDA)® and SpectrumTM Method. In all these methods, establishing a trained panel to perform descriptive analysis is a critical matter. It is important for the panel to be taught and maintained by a sensory professional with the training and experience in the analytical method being applied

(Hootman, 1992). According to Stone and Sidel (1985), the method of QDA embodies a formal screening and training procedure followed by the development of a sensory language with scoring of products in repeated trials to obtain a complete and 20

quantitative description and perform statistical analyses on the responses. Testing usually starts with 25 individuals and then is reduced to 10-12 qualified discriminators who are able to differentiate and describe the products. In FPM, a minimum of 4 trained panelists is used to identify each characteristic that contributes to the overall sensory impression of the product. This is followed by an assessment of their intensities to build a description of the aroma, flavor and aftertaste of the product (Keane, 1992). This method allows the panel leader to fully participate in the product evaluation as opposed to QDA. The SpectrumTM Method measures the perceived intensities of the identified sensory attributes in relation to absolute or universal scales. This allows the comparison of relative intensities among attributes within a product and among products tested

(Muñoz and Civille, 1992). The scale used in this method is a 15-cm (or 10-cm) line scale and can be anchored at the end with terms such as “none” or “extreme”. The trained panel is usually composed of 12-15 qualified discriminators.

Dry-cured ham flavor is often described as a complex process where the terms used can be very subjective and dependent on the origin of the product (Flores et al.,

1998; Toldrá, 2002). Thus, it is necessary to develop a lexicon of precise sensory descriptors that is easily understood and replicated in order to better define the sensory characteristics of the product. The most representative attributes may be chosen from the developed lexicon for further analysis. Flores and coworkers (1997) developed a lexicon for the dry-cured ham flavor in which they selected the following attributes as most representative of the desirable flavors and off-flavors associated with the product:

fat complex, boar taint, barnyard, haylike/musty, brown spice, pickling spice, smoky,

pork, serum, pungent, sour, salty, bitter, astringent, metallic and mouth filling.

The length of the dry-curing process has a great influence on the development of

the aroma and flavor traits of the product. Sensory analysis of the Serrano dry-cured

ham revealed higher values for boar taint, barnyard, sour and salty attributes in the long

processed ham (12 months) when compared with the shorter-processed ham (7 months)

(Flores et al., 1997; 1998). The authors attributed the higher barnyard (aromatics associated with free fatty acids) attribute to an increase in volatile fatty acids, higher sour taste to an increase in free amino acids such as aspartic and glutamic acids and higher salty taste to a lower water activity, all of which are enhanced in longer aged hams. Aldehydes, branched aldehydes, and dimethyl disulfide (fresh-cured pork flavor) were dominant in the short aged ham while 2-propanol, 2-propanone and heptane characterized the long aged ham (Flores et al., 1998; Toldrá, 2002). In Iberian hams, higher values were observed for flavor strength, cured flavor and aftertaste for longer aged hams (Ruiz et al., 1998). Parma hams, which require a minimum of 12 months in the length of processing time, showed aged, salty and acid tastes as well as aged and fresh pork odors as its most important descriptors (Careri et al., 1993).

Differences in the sensory attributes among dry-cured ham products are mainly due to the raw materials and manufacturing process. Sensory attributes related to lipid content and fatty acid composition are affected by rearing conditions such as utilizing a commercial feed or a free range system (Toldrá, 2002). Dry-cured hams from pigs raised in a free-range system based on acorn and pasture gave higher scores in oiliness,

brightness of the lean, marbling and aroma and flavor traits (Cava et al., 2000; Toldrá,

2002). The location of the sample is also important since differences in muscle composition and variation in the extent of dehydration is dependent on the distance that the muscle is from the ham surface, which affects product texture. In Iberian hams, the slice location affected fat firmness, lean dryness and juiciness (Ruiz et al., 1998). The

external Semimembranosus muscle is usually harder, drier and more fibrous than the

internal Biceps femoris muscle (Cava et al., 2000). A comparison of dry-cured hams from various crossbreeds and sex revealed that hams from females and Danish Duroc- sired pigs received higher values for over-all quality. On the other hand, Belgian

Landrace-sired pigs showed a high rancid aroma even though they were characterized with a low marbling appearance.

Extraction of Volatile Flavor Compounds in Dry-cured Ham

Various methods have been used for the extraction and determination of volatile compounds in dry-cured hams such as vacuum distillation (Berdague et al., 1991), steam distillation (Lillard and Ayres, 1969), simultaneous distillation/extraction (SDE)

(García-Esteban et al., 2004), supercritical fluid extraction (SFE) (Timón et al., 1998), purge and trap (P&T) (Barbieri et al., 1992; Buscailhon et al., 1994; Bolzoni et al.,

1996; Flores et al., 1997; Ruiz et al., 1999; Carrapiso et al., 2002) and solid phase microextraction (SPME) (Ruiz et al., 1998; Gianelli et al., 2002; Andrés et al., 2002;

Pérez-Juan et al., 2006). In food samples, analysis of headspace composition is preferred due to the simplicity of the method and preservation of the natural characteristics of the samples. Methods which efficiently extract target compounds 23

from the sample matrix are continually being sought since this process is generally the step at which analyte loss occurs (Steffen and Pawliszyn, 1996). Solid phase microextraction (SPME) is a rapid isolation technique that is used to extract volatiles in the headspace of a food product and/or solution. SPME is a solvent-free method that is used to extract analytes from various sample matrices through absorption or adsorption by the fiber phase until equilibrium is reached in the system (Pawliszyn, 2001). SPME is a sensitive and reproducible sample preparation technique which has been successful in the extraction of a broad range of dry-cured ham volatiles when compared to other extraction techniques that have been mentioned above (Ruiz et al., 1998). Isolation and concentration techniques that provide an extract with an odor representative of the original product are key to the success of flavor characterization (Qian et al., 2007).

Analysis of Volatile Flavor Compounds in Dry-cured Ham

Gas Chromatography Olfactometry (GCO)

Gas Chromatography Olfactometry (GCO), also known as “GC sniffing”, has emerged as a valuable analytical tool in flavor research because it characterizes the aroma of individual compounds or complex mixtures of volatiles eluting from the sniffing port (Mistry et al., 1997). A number of volatiles may be identified in a food sample but not all of these compounds will contribute to the characteristic aroma and flavor of the product. GCO has been useful in the screening of volatile compounds that play key roles in overall flavor and aroma by identifying the compounds that are present in the sensory threshold range. In this method, the human nose, which has a theoretical

odor detection limit of about 10-19 moles, is used as the instrument’s detector. This makes GCO a very sensitive tool that is effective at determining aroma active volatiles

(Reineccius and Heath, 1994). Another advantage of this method is that volatile compounds with different aromas and those that may vary slightly in their retention indices can be differentiated from one another by GCO (Acree, 1993). GCO methods can be divided into three categories: dilution methods, detection frequency methods and intensity methods.

The two dilution techniques include the aroma extract dilution analysis (AEDA) developed by Grosch and coworkers (1993) and the combined hedonic aroma response measurements (CHARM) analysis by Acree et al. (1984). Both of these dilution techniques are based on the “sequential dilution of an aroma concentrate” (Qian et al.

2002). In AEDA, the sample extract undergoes subsequent dilution and analysis by the

GCO until no odor is perceived at the sniffing port. The results are expressed as the flavor dilution (FD) factor, which is the maximal dilution value that the odorants can still be detected by GCO. In CHARM analysis, sequential dilutions of the original extract is performed until the odorants are no longer detected at the sniffing port. The sensory description and aroma duration of the individual compounds emerging from the

GC column are recorded so that the contribution of the individual aroma peaks to the overall aroma profile can be determined by the time-intensity combination (Qian et al.,

2002).

The detection frequency method was developed by Pollien et al. (1997). In this method, panelists (8-10) sniff the nondiluted extract and record the beginning and end

of elution for each odorant using a computerized device. Olfactometry global analysis, a

relatively new technique developed by Ott et al., (1997), is based on the detection frequency method. The intensities of the odorants are associated with their detection frequencies rather than the peak heights obtained during analysis. An aromagram of the compounds is generated by plotting the detection frequency (DF) on the Y axis and the

Kovats retention index on the X axis.

Osme is a time-intensity approach to GCO that directly estimates the magnitude of an odor by utilizing the human nose as the detector (McDaniel et al., 1990). It makes

use of the cross modal matching technique which is based on Stevens’ power law that

matches the intensity of one attribute (i.e. aroma intensity) to that of another attribute

(i.e. odorant concentration) (Fu et al., 2002). Trained panelists use a 15 point scale to

rate the intensities of the eluting aroma active compounds.

LeGuen and coworkers (2000) compared AEDA, Osme time-intensity and

olfactometry global analysis in the identification of the most potent odorants in cooked

mussels. Results from the three olfactometric techniques were well correlated,

especially for the most potent odorants. The authors also concluded that the choice of a

method depends on the objective of the study, the quality of the panel and the time

scheduled for the analyses.

Gas Chromatography Mass Spectrometry (GCMS)

The combination of gas chromatography (GC) for separation and mass

spectrometry (MS) for detection and identification of volatile compounds has become

one of the most powerful analytical tools available. GCMS has the ability to separate 26

mixtures into their individual components and then provide quantitative and qualitative information on the amounts and chemical structure of each compound (McMaster and

McMaster, 1998). An important application of this instrument in the food industry is the identification of the volatile flavor compounds in various food products. The GCMS is favored for quantitative analyses due to its excellent selectivity of detection and low limits of detection for numerous compounds (Masucci and Caldwell, 2004). The mass spectrometer generates the molecular weight of the fragment ions which will be helpful in deducing the molecular structures of the compounds of interest. The underlying principle of this method is that “after being separated on a capillary column, the height or integrated area of each individual peak is proportional to the amount of each volatile compound in the original mixture ” (Lei, 2001). For quantification of the compounds, the right internal or external standard is selected which is usually a structural analog of the most important volatile compound in the sample matrix (Grob and Kaiser, 2004).

Relationships Between Volatile Flavor Compounds, Sensory Descriptors and Consumer Acceptability

Several studies have been performed to correlate the volatile and nonvolatile components with sensory analysis of various dry-cured ham products. Multivariate statistical methods which have been utilized to relate the instrumental and sensory analyses of dry-cured ham flavor include: Principal Components Analysis (PCA)

(Berdagué et al., 1993; Ruiz et al., 1999), Multiple Factor Analysis (Flores et al., 1997) and Generalized Procrustes Analysis (GPA) (Careri et al., 1993). Until now there has

been no suggestion of a specific compound in dry-cured hams as responsible for the

characteristic “cured” flavor and aroma.

Correlations have been found between the aged flavor of Parma ham and

identified esters, aromatic hydrocarbons and cyclic nitrogen compounds that have been

associated with the high acceptability scores received by the product (Careri et al.,

1993). In the case of French dry-cured hams, several ketones have been related with its pleasant aroma while several aldehydes were associated with the aroma of fresh-cured pork (Buscailhon et al., 1994). Hinrichsen and Pedersen (1995) have established methyl-branched aldehydes, secondary alcohols, methylketones, ethyl esters and dimethyl trisulfide as contributors to the nutty, cheesy and salty attributes of dry-cured ham. Relationships between volatile compounds, sensory descriptors and specific characteristics of the processing condition have also been demonstrated. For instance, the short ripening process for Serrano dry-cured ham has been associated with a fresh- cured pork flavor defined by the volatile compounds hexanal, 3-methylbutanal, 1- penten-3-ol and dimethyl disulfide (Flores et al., 1997). Similarities with the French dry-cured hams were observed especially in the definition of the pleasant aroma of the short ripening process by ketones, esters, pyrazines and aromatic hydrocarbons. In contrast, the longer ripening-drying process for Serrano dry-cured ham “produced an increase in pork, cured and off-flavor that masked the pleasant aroma” (Flores et al.,

1997). Iberian hams characterized by longer processing times had higher amounts of methylketones, aldehydes, branched aldehydes and alcohols. The accumulation of these

compounds in this type of ham has been associated with positive aroma and flavor traits and lower rancidity scores for the product (Ruiz et al., 1999).

Establishing significant correlations between specific chemical compounds with particular flavors requires extensive instrumental and sensory analysis (Marsili, 2007).

Nonetheless an understanding of how differences in sensory properties among a range of samples are caused by variations in chemical composition can generate valuable results.

Principal Components Analysis (PCA)

Principal Components Analysis (PCA) is a statistical analysis technique used “to identify the smallest number of latent variables, called “principal components”, that explain the greatest amount of observed variability” (Meilgaard et al., 1999). According to Dunteman (1989), the goal of this multivariate data analysis method is to reduce the dimensionality of the original data set since a small set of uncorrelated variables is easier to grasp and utilize in further analysis compared with a larger set of correlated variables. This method allows as much as 75% to 90% of the total variability of a data set with 25 to 30 variables to be explained with as few as 2 to 3 principal components.

If a significant portion of the total variation can be accounted for by the first two or three principal components, then the PCA biplots can be used to visually identify clusters. The PCA biplot consists of a graphical representation of the variables and data.

However these representations can only be reliable if the “sum of the variability percentages associated with the axes of the representation space are sufficiently high

(i.e. 80%)” (XLSTAT Tutorial, 2007). 29

Cluster Analysis

Cluster analysis is a multivariate statistical method which groups observations together based on the degree of similarity among their ratings in the same way that PCA groups attributes together based on their degree of correlated behavior (Meilgaard et al.,

1999). An important application of cluster analysis is in sensory acceptance testing where it has been used to identify groups of respondents that have different patterns of liking across products (Meilgaard et al., 1999). Agglomerative Hierarchical Clustering

(AHC) is a cluster analysis algorithm which is used to “make up homogeneous groups of objects (classes) on the basis of their description by a set of variables, or from a matrix describing the similarity or dissimilarity between the objects” (XLSTAT

Tutorial, 2007). In comparison to PCA which has long been used as a stand alone method in sensory and instrumental analyses, AHC is seldom used in sensory studies aside from external preference mapping (Schilling and Coggins, 2007). This method allows a better understanding of the hedonic scaled consumer data by making the results easier to interpret (Ares et al., 2006; Schilling and Coggins, 2007). Another advantage of the AHC is in the graphical presentation of its results using a dendrogram which shows the progressive grouping of the data (XLSTAT Tutorial, 2007). Schilling and

Coggins (2007) coupled AHC with traditionally used analyses in the evaluation of hedonic scaled consumer data relating to chicken nuggets, retorted ham, fluid milk and cooked shrimp. The authors concluded that combining AHC with traditional analyses was successful at grouping consumers together based on product preference and liking.

Once the clusters have been identified, preference mapping can be performed to reveal

the similarity and differences among the clusters in how sensory attributes relate to product acceptability.

External Preference Mapping

External preference mapping is utilized to explain consumer preference data by descriptive sensory information and/or instrumental data (Ebeler, 1999; Schlich, 1995;

McEwan, 1996; Murray and Delahunty, 2000). Individual consumer preferences are regressed onto the first two principal components of the covariance matrix of descriptive or analytical ratings across products such that the dimensions of the descriptive or instrumental analysis space are the predictor variables while consumer acceptability is the response variable (Schlich, 1995; Guinard 1998; 2002). An important contribution of this technique is that it can guide product optimization and development (McEwan, 1996). One limitation of external preference mapping is that some consumers may not be fitted by any of the models. They are clearly consumers whose degree of liking for a set of products does not follow any preset predictive model. Even though some information is lost with “unfitted consumers”, the main market segments and drivers of liking are sorted out with external preference mapping

(Guinard, 2002). External preference mapping has been applied in a number of studies with various products such as chocolate milk (Thompson et al., 2004), cheddar cheese

(Young et al., 2004, Casapia et al. 2006), butter (Krause et al., 2007), dulce de leche

(Ares et al., 2006), commercial lager beers (Guinard et al., 2001) and vanilla ice cream

(Guinard, 2002). Knowing what characterizes a market segment and where its

preference lies in terms of sensory characteristics, makes for a very powerful product optimization tool (Guinard, 2002).

CHAPTER III

MATERIALS AND METHODS

Materials

Sample

Eight dry-cured ham products (Table 3) with varied aging periods and cure

ingredients were purchased from six manufacturers at three different processing times

within the year where seven of the hams were purchased as slices and one ham was

purchased as a whole. These samples were carefully selected by two experienced dry-

cured ham researchers giving a good representation of the variety in dry-cured hams

based on regional locations and flavor preferences. Five hams underwent a processing

period of 90-180 days (short process), which consisted of the typical stages such as

curing (2 days per pound of uncured weight, 2-4oC), cure equalization (14 days, 10-

13oC), smoking (1 day, 31-33oC) and aging (40-130 days, 29-32 oC). The remaining

hams were aged by means of the long process with a processing period of either 180-

270 days or 270-360 days. Only center ham slices (6+1mm) which consisted of the

Biceps femoris, Semitendinosus and Semimembranosus muscles were used. Dry-cured ham slices were separated, individually vacuum packed in high performance bags

(Vacuum Pouches, KOCH Supplies Inc., Kansas City, MO) with a vacuum packaging

Table 3. Dry-cured hams from different regional locations and with varying aging periods.

Ham Processing Period Curing Informationc Smoked 1SHORT 3-6 mos Salt Yes 2SHORTa 4-6 mos Salt Yes 3SHORT 3-6 mos Salt Yes 4SHORTb 4-6 mos Salt Yes 5SHORT 4 mos Salt Yes 1LONGa 10-12 mos Salt Yes 2LONGb 9-12 mos Salt No 3LONG 9 mos Salt No a-b Dry-cured hams with the same letter denote the same manufacturer. c Additional ingredients withheld for proprietary reasons.

machine (Model HVT-30, Hollymatic Corp., Countryside, IL) and kept at 4oC until

subsequent analyses were performed. Stored samples were used within a period of one

month.

Chemicals

The following chemical standards were obtained to verify gas chromatographic

results: methanethiol, carbon disulfide, 2-butanone, 3-methylbutanal, hexanal, 2-

furanmethanol, 2-heptanone, methional, 1-octen-3-ol, benzaldehyde, limonene,

benzeneacetaldehyde, 2-methoxyphenol, 4-methyl-2-methoxyphenol, 4-ethyl-2-

methoxyphenol and 2,6-dimethoxyphenol (Sigma-Aldrich Chemical Co Milwaukee,

WI). An internal standard chlorobenzene (200 ppm) and an n-alkane series C5-C18

(Sigma-Aldrich Chemical Co., Milwaukee, WI) were used to standardize the results and calculate Linear Retention Indices (LRI). Deionized water (Fisher Scientific Company

LLC, Middletown, VA) with a constant pH of 7.0 at room temperature was used for sample preparation of dry-cured ham homogenates.

SPME (Solid Phase Microextraction) Fiber

A three phase solid phase microextraction (SPME) fiber (2 cm-50/30 m

StableFlex Divinylbenzene (DVB)/CarboxenTM (Car)/Polydimethylsiloxane (PDMS),

Supelco, Bellefonte, PA) was used to extract the volatile compounds from the samples.

This fiber was utilized due to its sensitivity for medium and high molecular weight analytes and its ability to extract dry-cured ham volatiles (Gianelli et al., 2002).

Methods

Sample Preparation

Ham slices were equilibrated to room temperature (20oC), wrapped in

ReynoldsTM extra heavy duty foil bags (Alcoa Consumer Products, Alcoa Inc.,

Richmond, VA) and placed on a metal baking pan for support. Ham slices were cooked in an electric oven (Model JBP25DOJ2WH, General Electric, Louisville, KY) at 177oC to a surface temperature of 66+3oC which was monitored using a non-contact infrared thermometer (Model IT-330, Horiba, Kyoto, Japan). This sample preparation method was adapted from the cooking information provided by the manufacturers to consumers to obtain a good representation of the product as it is normally consumed as well as to minimize variation in the cooking process. Skin, subcutaneous and intermuscular fats were trimmed off the oven-baked ham slices prior to analytical and sensory analyses.

Descriptive Se nsory A nalysis

Eight dry-cured ha ms were assessed by a panel of 7 me mbers with greater than

1 0 0 h o urs of e x p eri e n c e p ert ai ni n g t o t h e e v al u ati o n of m e at pr o d u cts. A t ot al of 3

s essi o ns w er e c o n d u ct e d i n a s p a n of 3 m o nt hs w h er e ei g ht h a ms w er e e v al u at e d f or

e v er y s essi o n. T h e Q u a ntit ati v e D es cri pti v e A n al ysis ( Q D A ® ) m et h o d of d es cri pti v e

a n al ysis w as c arri e d o ut in all t h es e s essi o ns ( St o n e a n d Si d el, 1 9 8 5; M eil g a ar d et al.,

1999). The panelists participated in training sessions which totaled more than 20 hours

with a mini mu m of 3 hours for each dry-cured ha m sa mple. Previously identified

descriptors ( Table 4, Ar mero et al., 1 9 9 9; R ui z et al., 1998) and ter ms generated during

training were utilized for the sensory evaluation of dry-cured ha ms. Panelists

i n di vi d u all y f or m e d a d es cri pti v e pr ofil e f or e a c h s a m pl e u n d er fl u or es c e nt li g hti n g

( a p pr o xi m at el y 6 4 c d), i n a w ell v e ntil at e d, p ositi v e pr ess ur e a n d t e m p er at ur e c o ntr oll e d

roo m. This was follo wed by a group discussion to arrive at a consensus on the lexicon

of d es cri pt ors a n d e ns ur e p a n el c o nsist e n c y in t h e u n d erst a n di n g of th es e d es cri pt ors.

T h e d es cri pt ors w er e rat e d usi n g a 1 5- p oi nt int e nsit y li n e sc al e w h er e 0 = n ot d et e ct e d

a n d 1 5 = e xtr e m el y str o n g wit h r es p e ct t o t h e s e ns or y attri b ut es ( Fi g ur e A 1).

O v e n b a k e d, t hi n dr y- c ur e d h a m sli c es w er e c ut i nt o 2. 5 4 c m x 2. 5 4 c m pi e c es,

pl a c e d i n g all o n-si z e d Zi pl o c b a gs ( Zi pl o c br a n d b a gs, S. C. J o h ns o n a n d S o n I n c.,

R a ci n e, WI) a n d st or e d i n a w at er b at h ( 6 0 o C) u ntil s e ns or y e v al u ati o ns c o ul d b e perfor med (10-15 min). Upon serving, dry-cured ha m pieces were placed in 2-oz. pl asti c c o nt ai n ers wit h li ds ( S w e et h e art C u p C o., O wi n g Mills, M D) a n d w er e c o d e d wit h t hr e e- di git r a n d o m n u m b ers. E a c h p a n elist r e c ei v e d 4 pi e c es (first t w o pi e c es w er e

3 6

Table 4. Sensory descriptors and definitions for American dry-cured ham.

Sensory Descriptor Definition Aroma Rancida Aroma associated with extremely oxidized fat or oil Molassesc Aroma associated with molasses; has a sharp, slight sulfur and/or caramelized character Fermentedc Aroma associated with lactic or spoilage bacteria as in soured meat Caramelizedc Sweet aromatic, characteristic of browned sugars and some other carbohydrates Pork complexc Aromatic associated with a live pig or its habitat, or wet pig hair Smokyc Perception of any type of smoke aroma Earthyc Aroma characteristic of damp soil, wet foliage, or slightly undercooked boiled potato Savoryc Aroma associated with savoriness or meatiness Flavor Curedb Intensity of the typical flavor from cured meat products (very low to very high) Rancidb Intensity of the rancid flavor (very low to very high) Saltya The taste on the tongue associated with sodium ions Aftertasteb Intensity and time extension of the flavor after sample is swallowed (very low to very high) Pork complexc Flavor associated with a live pig or its habitat, or wet pig hair Sweeta Level of sweetness ( not to very sweet) Savoryc Flavor associated with savoriness or meatiness Bittera The taste on the tongue associated with caffeine Astringenta The chemical feeling factor on the tongue described as puckering/dry and associated with strong tea Mouthdryingc The chemical feeling factor on the tongue or other skin surfaces of the oral cavity described as puckering/dry and associated with tannins or alum Saltburnc Burning sensation on the tongue Texture Hardnessb Effort required to bite through lean and to convert the sample to a swallowable state Drynessb Amount of juices which is present in the mouth in the first chews Fibrousnessb Extent to which fibers/strands are perceived on chewing Juicinessb Impression of lubricated food during chewing (not to very juicy) Chewiness Identified by chew count and chew rate: sensation of labored mastication due to sustained, elastic resistance from a foodstuff Mushiness Softness; pulpy; lacking in texture Appearance Color Presence of color homogeneity in the different muscles of a transversal cut homogeneitya Marblinga Presence of intramuscular fat a Sensory descriptors and definitions taken from Armero et al. (1999). b Sensory descriptors and definitions taken from Ruiz et al. (1998). c Sensory descriptors generated during training by the trained descriptive panel, definitions taken from Civille and Lyon, 1996.

used to evaluate aroma and appearance attributes and the next two pieces were used for flavor and textural attributes) of each ham for every session. This was carried out to ensure that the dry-cured hams were maintained at the same temperature throughout the evaluation. A factorial rotation was applied for the order of presentation of the samples to consider the effect of rank. Panelists were provided with water (Mountain Spring

Water, Blue Ridge, GA), unsalted crackers (Premium Nabisco, NJ), apple juice (Lucky

Leaf Apple Juice, Knouse Foods Co-op Inc., Peach Glen, PA) and expectorant cups to remove residual flavors between sample evaluations.

Consumer Acceptability

Consumer acceptability of eight dry-cured hams was evaluated by a panel of 71 consumers who had been preselected as regular dry-cured ham consumers. The consumers were recruited from faculty, staff and students in the university through personal communication and email. Three sessions (n=23-25) were conducted in a span of three months and within a week of the descriptive analysis by the trained panel.

Evaluation was conducted at the same time for every session at a sensory evaluation laboratory. Panelists evaluated eight dry-cured ham samples in an eight booth sensory panel, under fluorescent lighting, in a well ventilated and temperature controlled room.

Each panelist received 2 pieces (2.54 cm x 2.54 cm) of each ham in 2-oz. plastic containers with lids bearing three-digit random numbers. Conditions for the preparation of these samples were identical with those of the trained descriptive panel. Panelists were provided with water, unsalted crackers, apple juice and expectorant cups to remove residual flavors between sample evaluations. Panelists were asked to evaluate 38

each ham sample based on overall acceptability and the acceptability of flavor, aroma, texture, and appearance using a 9-point hedonic scale. The scale used was as follows: 1-

Dislike extremely, 2-Dislike very much, 3-Dislike moderately, 4-Dislike slightly, 5-

Neither like nor dislike, 6-Like slightly, 7-Like moderately, 8-Like very much and 9-

Like extremely (Meilgaard et al., 1999) (Figure A2).

Conditioning of SPME Fiber for Chromatographic Analysis

Prior to sampling, fibers were conditioned in a split/splitless GC injector port for

30 min at 250oC, to remove any possible contaminants from the fiber coating. This was followed by desorption at the GC injector port for 5 min, with the GC programmed, to determine the presence of any extraneous peaks. An instrument blank, a fiber blank, a deionized water blank, and a deionized water with sodium chloride blank were run to rule out irrelevant peaks that might appear in the analyses of the samples.

Isolation of Volatile Compounds

Ten grams of dry-cured ham homogenate (Model HC306, Black and Decker,

Corporation, Towson, MD) (1:1 dilution, 50%w/w) were transferred into a 40 ml amber glass vial (O.D. 28 x 98 mm height, Supelco, Bellefonte, PA) with an open center propylene screw cap and Teflon faced silicone septum (O.D. 22 mm diameter, Supelco,

Bellefonte, PA). Internal standard (1l, 200 ppm, chlorobenzene) was added into the vial for quantification of the detected volatile compounds. Sodium chloride (0.5 g;

99%+, ACS reagent, Sigma-Aldrich Chemical Co., Milwaukee, WI) was added to improve the extraction of volatile compounds. The sample was then equilibrated for 30

min at 50oC in the amber glass vial. The SPME fiber (DVB/CAR/PDMS) was exposed

to the generated sample headspace for 1 hr at 50oC in a thermostatic heating block

(Reacti-therm Heating/Stirring Module, Pierce Biotechnology Inc., Rockford, IL) with constant stirring using a magnetic octagonal stirring bar (8 mm diameter x 13 mm length, Fisher, Pittsburgh, PA). This was followed by thermal desorption of the volatiles from the SPME fiber inserted into a split/splitless injector port of a Varian 3900 gas chromatograph (Varian Inc., Walnut Creek, CA). Prior to actual analysis of the samples, optimization of the method was carried out. Effects of different homogenate concentrations (25, 50, 75, 90 and 100%), added salt (0 and 0.5g), equilibration times

(30 and 45min), extraction times (30, 45 and 60min) and equilibration and extraction temperatures (47, 50 and 52oC) were tested. Analysis of each dry-cured ham was repeated in triplicate for three replications to ensure the reproducibility of results.

Gas Chromatography Mass Spectrometry (GCMS)

Analysis of the volatile compounds adsorbed on the fiber was carried out using a Varian 3900 gas chromatograph equipped with a CP-1177 Split/Splitless injector and coupled to a Saturn 2100T ion trap mass selective detector (MSD, Varian Inc., Walnut

Creek, CA). Separation of the volatile compounds was accomplished on a DB-5 capillary column (30 m x 0.25 mm i.d. x 0.25 m film thickness (df), J &W Scientific,

Folsom, CA) with a flow rate of 1.0 mL/min and ultra high purity helium (Airgas,

Columbus, MS) as the carrier gas. The SPME fiber was desorbed in the injector port at

250oC for 5 min which was operated in a splitless mode. The programming sequence

for the GC oven temperature was as follows: an initial temperature of 35oC held for 1 40

min and then raised to 250oC at a rate of 14oC/min and held for 4 min for a total running time of 20.36 min. The MSD conditions were as follows: transfer line maintained at

250oC; electronic ionization mode at an ionization voltage of 70 eV; scan rate of 0.5

sec/scan; m/z range of 35-350 amu and manifold temperature of 180oC. Mass spectral

data of the volatile compounds were determined using the library search algorithm,

NIST02 Mass Spectral Database (NIST, Maryland; purchased from Varian Inc.).

Gas Chromatography Olfactometry (GCO)

Three panelists trained in sensory and GCO analyses evaluated the aroma

properties of the volatile compounds present in the samples. Training sessions consisted

of practice runs by sniffing original samples and prepared samples extracted by SPME

from dry-cured ham homogenates. These sessions were conducted to familiarize the

panelists with the aroma of the compounds, to practice proper breathing techniques, and

to establish verbal descriptors for the samples (Rousseff et al., 2001).

GCO analysis, by the Osme time-intensity method, was carried out using a

modified Varian CP-3800 (Varian Inc., Walnut Creek, CA) gas chromatograph

equipped with a sniffing port (ODO-I, SGE, Kramer Lane, Austin, TX) and a flame

ionization detector (FID). A DB-5 capillary column (30 m x 0.25 mm i.d. x 0.25 m

film thickness (df), J &W Scientific, Folsom, CA) contained in the GC oven extended through a stainless steel sniff port to a custom-made glass nose cone fixed at the end

(SGE, Kramer Lane Austin, TX). In addition, the glass nose cone was purged with humidified air at a flow rate of 30 mL/min to maintain olfactory sensitivity by reducing drying of the mucous membranes in the nasal cavity. The GCO operating conditions 41

such as injector port temperature and oven temperature programming sequence were

identical to those listed for GCMS.

The intensity of the perceived aroma was rated by each panelist using a 0-15

potentiometric sliding scale interfaced to a computer (Osme Software, Starkville, MS)

making it possible for the time-intensity data to be recorded. The retention time,

intensity and verbal description of the aroma were concurrently listed by a coworker.

Retention times and verbal descriptions were recorded to couple aroma descriptors with

computerized aroma intensity plots and to label the generated aroma intensity peaks

with the appropriate aroma descriptors (Rousseff et al., 2001). Authentic standards and

n-alkane series C5-C18 were analyzed with the GCFID. Headspace volumes (10ul) were drawn using a gastight digital syringe (1700 Series GASTIGHTTM Digital Syringe,

Hamilton Company, Reno, Nevada) and immediately injected into the injector. GCFID

operating conditions were the same with those listed for the GCO analysis of the dry-

cured ham samples.

Identification of Aroma Impact Compounds

Identification of the aroma impact compounds was based on comparison of mass

spectra identified with spectra obtained using the NIST02 library, comparison of linear

retention index and aroma quality perceived at sniffing port with an authentic standard

and comparison of linear retention index and the aroma quality perceived at the sniffing

port with literature (Gianelli et al., 2002; García-Esteban et al., 2004; Ramirez and

Cava, 2007) or with retention indices databases (Acree and Arn, 2005; El-Sayed, 2003).

The aroma retention times were converted to linear retention indices using the retention 42

ti m es of a n n -alkane series C 5 - C 1 8 . T h e i d e ntit y a n d ar o m a q u alit y of t h e ar o m a i m p a ct c o m p o u n ds w er e c o nfir m e d b y r u n ni n g a ut h e nti c st a n d ar ds. T h e ar e a r ati o ( ar e a of co mpound/area of internal standard) was multiplied by the concentration of the internal standard after taking into account sa mple weight to deter mine the relative abundance of the co mpound. The dry-cured ha m co mposition was expressed as ng/g ha m (relative abundance ppb).

St atisti c al A n al ysis

A r a n d o mi z e d c o m pl et e bl o c k d esi g n wit h t hr e e r e pli c ati o ns w as utili z e d t o deter mine differences ( P <0.05) a mong dry-cured ha m treat ments in respect to descriptive flavor, aro ma and textural attributes. When significant differences occurred a mong treat ments, the Least Significant Difference ( LS D) test was utilized to separate t h e tr e at m e nt m e a ns. Pri n ci p al C o m p o n e nts A n al ysis ( P C A, St atisti c al A n al ysis

Soft ware, version 9.1, S A S ® I nstit ut e, C ar y, N C) w as c arri e d o ut t o d et er mi n e t h e relationships bet ween descriptive data, dry-cured ha m treat ments and volatile c o m p o u n d c o m p ositi o n. A ra n d o mi z e d c o m pl et e bl o c k d esi g n wit h t hr e e re pli c ati o ns w as utili z e d t o d et er mi n e t h e a v er a g e diff er e n c e ( P < 0. 0 5) i n c o ns u m er a c c e pt a bilit y a mong dry-cured ha m treat ments. Consu mers were clustered together based on their li ki n g f or dr y- c ur e d h a m b y a g gl o m er ati v e hi er ar c hi c al cl ust eri n g usi n g th e E u cli di a n dist a n c e a n d W ar d’s m et h o d as a g gr e g ati o n crit eri o n. D e n dr o gr a m a n d dissi mil arit y pl ots w er e us e d to d et er mi n e h o w m a n y cl ust ers s h o ul d b e us e d to gr o u p c o ns u m ers.

Rando mized co mplete block designs were utilized to differentiate (P <0.05) a mong dry- cured ha m treat ments within each cluster. When significant differences occurred a mong 4 3

treatments within a cluster, the Least Significant Difference (LSD) test was performed to separate the treatment means. External preference and external flavor mapping

(XLSTAT-Pro and XLSTAT-MX version 2007, Addinsoft USA, New York, NY) were performed on compound relative abundance, descriptive data and consumer acceptability scores to establish the relationships between volatile flavor compounds, sensory attributes and consumer acceptability.

CHAPTER IV

RESULTS AND DISCUSSION

Descriptive Analysis

Eight commercial American dry-cured hams were described and differentiated by a trained descriptive panel using 25 sensory descriptors. Differences existed

(P<0.05) among the dry-cured ham products for 19 of the 25 sensory descriptors. The descriptors that differentiated (P<0.05) the hams were “cured”, “rancid”, “aftertaste”,

“pork complex”, “savory”, “astringent”, “mouthdrying”, “saltburn” flavors; “molasses”,

“fermented”, “caramelized”, “smoky”, “savory” aromas and “hard”, “dry”, “fibrous”,

“juicy”, “chewy”, “mushy” textural attributes. The mean intensity values of the sensory descriptors for the dry-cured ham products are shown in Table 5 with a range of 1.0-9.3 for all attributes on a 15-point scale.

The principal components analysis (PCA) biplot for the descriptive flavor attributes illustrates differences between the samples and relationships between the attributes in the first two principal components (PCs) which accounted for 90% of the variation in the data (Figure 1). The PC1 dimension, which explained 68%

Table 5. Mean sensory ratings for the flavor, aroma and textural attributes of eight commercial brands of American dry-cured hams 1.

Flavor Attribute Cured Rancid Salty Aftertaste Porkcomplex Sweet Savory Bitter Astringency Mouthdrying Saltburn 1SHORT 7.4bc 5.0bc 4.8 4.7bc 5.3a 1.4 2.7bcd 1.0 3.8cd 4.4cd 3.9cd 2SHORT 8.0ab 5.4bc 6.1 5.6abc 4.1bc 1.7 3.2ab 1.3 4.6bcd 5.7abc 5.4bc 3SHORT 9.0a 7.4a 7.4 6.8a 5.3a 1.0 2.3d 1.8 6.9a 6.9a 7.7a 4SHORT 6.4c 4.3c 4.9 4.5c 3.8c 1.9 2.9bcd 1.0 3.4d 3.9d 3.5d 5SHORT 8.1ab 5.0bc 5.4 4.7bc 4.0bc 1.4 2.7bcd 1.0 4.8bcd 5.1bcd 4.9bcd 1LONG 9.1a 5.7bc 7.3 6.5a 3.8c 1.7 3.7a 2.1 5.8ab 5.0cd 6.5ab 2LONG 7.8abc 5.1bc 6.5 5.7abc 4.8abc 1.3 2.4cd 1.3 5.0bc 5.3bcd 5.2cd 3LONG 8.3ab 6.4ab 6.4 6.2ab 4.9ab 1.0 3.0abc 1.6 5.4b 6.6ab 5.2bcd Aroma Attribute Rancid Molasses Fermented Caramelized Pork complex Smoky Earthy Savory 1SHORT 5.5 2.7abc 3.2ab 1.9c 6.3 1.6c 2.0 2.7bc 2SHORT 5.2 3.2ab 1.9d 2.6abc 3.9 3.1a 1.9 3.3ab bc a c c c

46 3SHORT 6.5 2.4 3.6 2.1 6.8 1.5 2.5 2.3 4SHORT 5.2 2.4bc 2.3cd 2.8ab 4.2 1.9bc 2.1 3.2ab 5SHORT 5.3 2.8abc 2.5bcd 2.1bc 4.9 2.0bc 2.2 3.1ab 1LONG 5.1 3.3a 2.2cd 3.0a 3.5 2.6ab 2.4 3.5a 2LONG 4.9 2.1c 2.6bcd 1.9c 5.1 1.5c 1.9 2.6bc 3LONG 6.4 2.3c 3.0abc 2.4abc 6.4 1.6c 3.0 2.6bc Texture Attribute Hardness Dryness Fibrousness Juiciness Chewiness Mushiness 1SHORT 4.0e 5.5bc 4.5c 7.5a 3.5c 3.3 2SHORT 6.0bcd 6.8b 5.7bc 6.6ab 4.5bc 4.2 3SHORT 7.9ab 6.7b 6.7b 6.3ab 6.1b 1.4 4SHORT 4.2de 5.2c 5.7bc 6.7ab 4.5bc 3.5 5SHORT 5.6cde 6.5b 6.2b 6.5abc 5.3bc 2.5 1LONG 5.5cde 6.4bc 6.1b 7.1ab 5.6b 3.6 2LONG 6.6bc 6.4bc 6.8b 6.1bc 6.5ab 1.8 3LONG 9.3a 8.3a 8.2a 5.3c 7.9a 1.6 a-e Means in columns with different letters are significantly different (P<0.05). 1 Intensities scored on a 15-point line scale where 0 = none and 15 = very. For American dry-cured ham descriptions, see Table 1.

2 Savory 1LONG

Sweet

Bitter

1 Salty 2SHORT Cured Saltburn 4SHORT Aftertaste Astringency 5SHORT 0 Rancid 47 2LONG 3LONG 3SHORT Mouthdrying

Principal Component 2(22%) 2(22%) Component Component2(22%) Principal Principal 1SHORT - 1

Pork Complex

- 2

-4- 4 -3- 3 --2 2 -1- 1 0 1 2 3 4 - 3 Principal Component 1 (68%) Figure 1. Principal Component Analysis Biplot of descriptive analysis data for flavor attributes of American dry-cured ham.

of the variance, showed that samples were mainly differentiated by “bitter”, “salty”,

“cured”, “saltburn”, “aftertaste”, “astringency”, “rancid” and “mouthdrying” (positive correlation to PC1 dimension). The PC2 dimension explained an additional 22% of the variance and was primarily a function of the “sweet” and “savory” (positive correlation to PC2 dimension) and “pork complex” flavors (negative correlation to PC2 dimension). 2SHORT, 4SHORT, 5SHORT and 1LONG were differentiated from

3SHORT and 3LONG due to their stronger “sweet” and “savory” flavor attributes.

The PCA biplot for the aroma sensory descriptors is shown in Figure 2 where

65% of the variation among products was explained by the PC1 dimension and 20% was explained by the PC2 dimension. The positive PC1 dimension was largely correlated with “caramelized”, “molasses”, “smoky” and “savory” which were dominant aromas in 1LONG, 2SHORT, 4SHORT and 5SHORT. “Rancid”,

“fermented”, “pork complex” (negative correlation to PC1 dimension) and “earthy”

(positive correlation to PC2 dimension) characterized the 3SHORT and 3LONG hams.

Results indicated that the inclusion of the smoking process as a manufacturing technique for the production of American dry-cured ham is a key factor in differentiating the dry-cured ham products based on their “sweet”, “savory”,

“caramelized”, “molasses” and “smoky” attributes. These results were in agreement with findings by Poligné et al. (2002) who reported that the smoking process was critical in determining the aroma quality of boucané pork, a traditionally cured smoked pork-belly product from France. But the results were quite different from previous sensory studies of Parma (Careri et al., 1993), Serrano (Flores et al., 1997; 1998) and

Earthy 2 Caramelized Rancid 3LONG 1LONG Molasses 3SHORT Smoky 1 Fermented Savory Pork Complex 4SHORT 2SHORT

0 1SHORT 5SHORT

49 Principal Component2(20%) Component2(20%) PrincipalComponent2(20%) Principal 2LONG - 1

- 2 -4- 4 -3- 3 -2- 2 -1- 1 0 1 2 3 4 Principal Component 1 (65%) - 3 Figure 2. Principal Component Analysis Biplot of descriptive analysis data for aroma attributes of American dry-cured ham.

Iberian (Ruiz et al., 1998; Toldrá and Flores, 1998) dry-cured hams which identified the length of processing time as the major factor in the strength of the flavor and aroma attributes.

The contribution of the length of processing time was more evident in the

previously mentioned dry-cured hams compared with the American dry-cured ham due

to their longer processing times which range from 9-36 months. Sweet, savory, and

smoky notes in smoked food products have been attributed to the phenolic compounds

in the vapor phase of smoke (Shahidi and Naczk, 2003). The smoking process is an

optional part in the manufacture of American dry-cured ham but it is often opted for due

to the enhanced color and flavor it provides to the product (Toldrá, 2004; Marriott and

Ockerman, 2004). In addition, Zotos et al. (1995) has suggested that the smoke flavor

may mask rancid flavor, causing a greater degree of oxidation to be needed for the

product to be unacceptable to some consumers. Six of the eight commercial dry-cured

ham products evaluated in this study were smoked.

Juicy and mushy textural attributes are located on the negative PC1 (85% of the

variation) and positive PC2 (9% of the variation) dimensions and characterize

1SHORT, 2SHORT, 4SHORT, 5SHORT and 1LONG (Figure 3). Hard, fibrous, chewy

(positive correlation to PC1 dimension) and dry (positive correlation to PC2 dimension)

textural attributes were dominant in 3SHORT, 2LONG and 3LONG. Previous studies

have reported that a longer processing time contributes to a harder texture in dry-cured

ham products. However, 1LONG may differ from 2LONG and 3LONG because it

underwent smoking since some products have a softer texture after being smoked

2 Mushiness

2SHORT Dryness 1 3LONG 1LONG Hardness Fibrousness

Juiciness 5SHORT Chewiness

4SHORT 0 1SHORT 2LONG 51 3SHORT

Principal Component 2(9%) 2(9%) Component PrincipalComponent2(9%) Principal - 1

- 2

-4- 4 -3- 3 -2- 2 -1- 1 0 1 2 3 4

- 3 Principal Component 1 (85%)

Figure 3. Principal Component Analysis Biplot of descriptive analysis data for textural attributes of American dry-cured ham.

(Maga, 1988). According to Cava et al. (2000), slice location is also important since the external Semimembranosus muscle tends to be harder, drier and more fibrous than the internal Biceps femoris muscle. Evaluation of the sensory appearance attributes was not included due to variation in raw material among the different dry-cured ham products.

The color of dry-cured ham mainly depends on the concentration of its natural pigment, myoglobin, which is dependent on the type of muscle and age of animal (Toldrá, 2004).

Dry cured ham is not a homogeneous product so the intramuscular and subcutaneous fat of the samples may differ (Sánchez-Peña et al., 2005)

Aroma Impact Compounds

A total of 18 aroma impact compounds were detected and 16 were tentatively identified. Table 6 shows the estimated concentrations of these compounds in the headspace of the dry-cured hams with essentially similar compounds in both short and long aged hams. All samples had similar volatile compound profiles with each ham containing 12-16 aroma impact compounds where the ham samples only differed in the relative concentrations of these compounds. Thirteen of the identified compounds have not been previously identified in the headspace volatiles of American dry-cured ham

(Ockerman et al, 1964; Lillard and Ayres, 1969; Piotrowski et al., 1970). However, all of these compounds have been detected in different types of dry-cured hams using various extraction techniques. Five aroma impact compounds which were significantly different in their relative concentration among the dry-cured hams were methanethiol,

2-butanone, 2-heptanone, benzaldehyde and 1-octen-3-ol.

Table 6. Aroma impact compounds identified and quantified in the headspace of American dry-cured hams using SPME- GCMSa, SPME-Osme-GCO and SPME-GCFID.

Compound Hamb 1SHORT 2SHORT 3SHORT 4SHORT 5SHORT 1LONG 2LONG 3LONG Methanethiol 4.6bc 3.5c 15.2a 4.0bc 10.3abc 3.0c 3.2c 13.0ab carbon disulfide 66.0 53.1 194.8 110.2 73.0 29.9 34.6 73.4 2-butanone 11.6b 8.4b 174.9a 6.2b 15.9b 8.0b 9.7b 30.8b 3-methylbutanal 23.6 40.8 57.6 22.1 44.5 14.9 25.2 24.8 Hexanal 80.8 110.2 365.9 79.6 236.1 139.3 104.8 230.6 2-furanmethanol 29.6 32.4 0.0 16.6 56.9 25.2 0.0 0.0 2-heptanone 3.1b 29.0b 95.4a 19.3b 39.5b 24.0b 24.2b 20.8b Methional 0.8 4.3 1.6 15.9 12.8 19.8 15.9 26.2 benzaldehyde 35.4b 30.9b 86.1a 34.5b 48.0ab 47.4ab 33.6b 86.0a c abc a c ab bc c abc 53 1-octen-3-ol 38.6 87.6 160.9 47.0 139.1 51.1 38.4 119.9 limonene 18.2 7.0 88.3 45.9 9.3 32.8 17.9 46.4 benzeneacetaldehyde 31.5 29.9 76.1 21.4 50.7 37.5 34.1 33.2 2-methoxyphenol 108.4 44.1 24.6 45.0 99.7 110.8 11.3 0.0 4-methyl-2-methoxyphenol 57.2 62.8 13.8 20.3 58.1 53.7 0.2 0.0 4-ethyl-2-methoxyphenol 12.3 23.7 0.0 1.6 10.3 24.0 0.0 0.0 2,6-dimethoxyphenol 24.3 25.5 16.3 7.1 66.9 34.1 4.5 6.6 a Results expressed as means of nine samples in ng/g ham. b Different letters in the same row (a, b, c) indicate significant statistical difference (P<0.05).

The aroma descriptions of the compounds which were repeatedly sniffed by at least 2 of the 3 trained panelists are summarized in Table 7. In the Osme time-intensity method, aroma intensity is assigned by the sniffer, and high-intensity aroma compounds potentially have greater impact on the aroma of the food and are regarded as the most important contributors to product flavor (McDaniel et al., 1990). The average aroma intensities were between 2.3 (2-butanone) and 12.0 (methional) on a 15-point scale where seven compounds had an average aroma intensity > 7: benzeneacetaldehyde, methional, 1-octen-3-ol, 2-heptanone, 2-furanmethanol, 2-methoxyphenol and unknown

(cooked rice, buttery popcorn). Moreover, methanethiol, carbon disulfide, 3- methylbutanal, hexanal, benzaldehyde, limonene, 4-methyl-2-methoxyphenol, 4-ethyl-

2-methoxyphenol and unknown (burnt meat, coffee) were given intermediate Osme intensity values ranging from 3 to 6. The remaining compounds 2-butanone and 2,6- dimethoxyphenol were also detected, but their contribution to American dry-cured aroma is miniscule based on these results. Figure 4 shows a comparison of the gas chromatography flame ionization detector (GCFID) chromatogram with the identified aroma impact compounds and GCO Osmegram with their respective aroma descriptions. The aroma impact compounds that were identified as important contributors to the distinct aroma of this product based on their assigned aroma intensities were not necessarily the compounds that were present in the highest concentrations, as can be seen by their peak areas. No single compound was associated with the “cured” aroma characteristic of dry-cured ham. Therefore, it appears that the

“cured” aroma is caused by a combination of volatile flavor compounds.

Table 7. Retention indices for aroma impact compounds detected in the headspace of American dry-cured ham during SPME- GCMS, SPME-Osme-GCO and SPME-GCFID.

Compound ab RI c Method of Identification d Aroma Descriptor e Aroma Intensity f Methanethiol <500 a, b, c Sulfury 3.9 Carbon disulfide 516 a, b, c Sulfury, Rotten cabbage 5.2 2-butanone 590 a, b, c Sweet, Alcoholic, Milky 2.3 3-methylbutanal 649 a, b, c Malty, Fermented 6.7 Hexanal 792 a, b, c Fresh cut green grass 4.9 2-furanmethanol 868 a, b, c Burnt meat, Vitamin 8.9 2-heptanone 888 a, b, c Burnt meat, Vitamin 8.9 Methional 912 a, b, c Baked potato 12.0 Unknown 917 d Burnt meat, Coffee 6.4 Unknown 925 d Cooked rice, Buttery popcorn 8.3

55 Benzaldehyde 964 a, b, c Burnt meat, Cooked meat 5.0 1-octen-3-ol 989 a, b, c Mushroom 7.9 Limonene 1032 a, b, c Citrus 5.9 Benzeneacetaldehyde 1065 a, b, c Floral, Rose 7.2 2-methoxyphenol 1126 a, b, c Smoky, Cocoa 9.1 4-methyl-2-methoxyphenol 1121 a, b, c Sweet ham 4.6 4-ethyl-2-methoxyphenol 1287 a, b, c Sweet ham 4.6 2,6-dimethoxyphenol 1359 a, b, c Smoky ham, Savory 2.5 a Compounds correspond to those in Figures 4 and 5, Table 6 . b Aroma impact compounds are presented in order of their elution on the DB-5 capillary column. c Retention indices calculated for DB-5 capillary column (J & W Scientific: 30m x 0.25 mm i.d. x 0.25 m film thickness) on a gas chromatograph equipped with a sniffing port and flame ionization detector. d Compounds were identified by: a - Mass spectrum tentatively identified using NIST02 library, b - Retention index calculated using GCO data and aroma perceived at sniffing port match those of an authentic standard, c - Retention index calculated using GCO data and aroma perceived at sniffing port in agreement with literature, d - Aroma perceived at the sniffing port e Aroma quality perceived by at least two panelists during SPME-Osme-GCO. f Average aroma intensity perceived at the sniffing port during SPME-Osme-GCO.

Figure 4. Comparison of the GCFID chromatogram (top) with the identification of some aroma impact compounds against GCO Osme time-intensity osmegram (bottom) for American dry-cured ham.

The flavor and flavor precursors in American dry-cured ham can be derived from proteolysis and lipolysis, two of the most important biochemical mechanisms that occur during the processing of this product (Toldrá, 2002). Proteolysis generates peptides, free amino acids, Strecker aldehydes and sulfur compounds. Lipolysis produces free fatty acids, hydrocarbons, aldehydes, ketones, alcohols, esters and lactones. An additional source of volatile flavor compounds is the reaction between amino acids and other compounds as in the case of the Maillard reaction which generates pyrazines that are extensively produced during meat cooking (Toldrá and

Flores, 1998; Mottram, 1999). Conversely, processing techniques such as smoking can contribute phenols, furans and pyrazines that impact the flavor development of this product (Wittkowski et al., 1992)

Three sulfur containing compounds (methanethiol, carbon disulfide and methional) were detected in the headspace of American dry-cured ham. The low threshold values of these compounds (0.02-0.2 ppb; Leffingwell and Leffingwell, 1991;

Carrapiso et al., 2002) allowed them to contribute to the flavor development of this product even in minute concentrations. Sulfur-containing volatiles are among the compounds responsible for the meaty note in cooked meats (Carrapiso and García,

2004). Methional was the most potent odorant contributing to a pleasant aroma described as baked potato. This compound may have been produced from the Strecker degradation of methionine during cooking (Le Guen et al., 2001). The Strecker degradation of sulfur containing amino acids, such as methionine and cysteine to thiols,

form sulfur containing compounds such as carbon disulfide and methanethiol (Mottram,

1999).

Four aldehydes were identified in the headspace of the American dry-cured ham.

These compounds are responsible for the following aroma descriptors: 3-methylbutanal

(malty, fermented), hexanal (fresh cut green grass), benzeneacetaldehyde (floral, rose) and benzaldehyde (burnt meat, cooked meat). It has been suggested that aldehydes contribute to the loss of desirable flavor in meats due to their high rate of formation during lipid oxidation and the low odor thresholds of 3-methylbutanal, hexanal, and benzeneacetaldehyde (0.2 ppb, 4.5 ppb, 4 ppb; Buttery et al., 1997) (Careri et al., 1993).

Formation of hexanal and benzaldehyde have been suggested to originate from the lipolysis autooxidation mechanisms while 3-methylbutanal and benzeneacetaldehyde can be derived from the Strecker degradation of amino acids. It is important to note that hexanal had a high concentration in all of the hams, although this compound was higher in a shorter aged ham than the expected longer aged. Hexanal and 3-methylbutanal have been previously identified as impact odorants of various dry-cured meat products such as Serrano (Flores et al., 1997) and Iberian (Carrapiso et al., 2002) hams.

Four odorants were isolated from American dry-cured ham that have not been reported previously: 2-methoxyphenol, 4-ethyl-2-methoxyphenol, 4-methyl-2- methoxyphenol and 2,6-dimethoxyphenol. Some of these phenolic compounds have low threshold values (3-21 ppb, 50 ppb, 90 ppb; Leffingwell and Leffingwell, 1991), thus making a marked contribution to flavor even when present at low concentrations.

Phenolic compounds in smoked food products have been suggested to originate from

the pyrolysis of lignin (Shahidi and Naczk, 2003). According to Varlet et al. (2006), phenolic compounds found in the vapor phase of smoke may contribute to imparting a smoky flavor to foods. Baloga et al. (1990) have attributed the smoky, cured aroma of a precooked premium ham to its phenolic compounds with 2-methoxyphenol and 4- methyl-2-methoxyphenol as the most predominant contributors to its distinct flavor.

Headspace enrichment procedures and subsequent sensory evaluation conducted by

Wittkowski et al. (1992) on Black Forest ham has proven that 2-methoxyphenol and some of its alkyl isomers are the dominant smoke flavor compounds in smoked food products.

2-heptanone and 2-butanone, were described as having burnt meat/vitamin and sweet/alcoholic/milky aromas. 2-heptanone appears to be the more dominant methylketone in the headspace of American dry-cured ham based on the Osme results and estimated concentrations. These lipid oxidation products of fatty acids have been identified as important flavor contributors in several dry-cured hams such as Serrano

(Dirinck et al., 1997) or Parma (Barbieri et al., 1992; Dirinck et al., 1997). The mushroom-like aroma of 1-octen-3-ol was identified as a significant contributor to the distinct flavor of this product due to its frequent detection and high Osme intensity value. Alcohols generally do not contribute to the overall aroma of food because of their high threshold values unless they are present at high concentrations or are unsaturated such as 1-octen-3-ol (Reineccius and Heath, 1994). This compound is regarded as a

‘classical’ high impact aroma chemical with an odor threshold of only 1 ppb (Rowe,

2002). This compound was previously identified as an aroma impact compound in several dry-cured ham products including Iberian ham (Carrapiso et al., 2002).

2-furanmethanol was also detected for the first time in American dry-cured ham aroma. Furans have been associated with the overall odor of broiled and roasted meats although they have been insignificant contributors to the basic meaty taste. Because of its fairly high Osme intensity value, it may contribute significantly to the cooked, cured aroma of this product. Two aroma impact compounds were not identified which may be due to their low concentrations, coelution with or masking by other compounds. The unknown aroma impact compound with a cooked rice/buttery aroma was effectively sniffed but was not successfully identified due to interference of another compound tentatively identified as methoxy phenyl oxime which eluted in the same area of the chromatogram. Methoxy phenyl oxime has been associated with the glue that is used to attach the SPME fiber to the syringe plunger (Grimm et al., 2002). The unknown aroma impact compound with a burnt meat, coffee aroma was tentatively identified by GCMS as an alkylpyrazine, which could be confirmed by its aroma description and comparison of its linear retention index with the available standards.

In order to show which volatiles occurred in a relatively greater concentration in the different dry-cured hams, PCA was performed on the concentrations of 16 aroma impact compounds. The first and second principal components (PCs) explained 52 and

26% of the variation of the data, respectively (Figure 5). The distribution of the aroma impact compounds on the first two PCs shows three separate groups. The dry-cured ham products 1SHORT, 2SHORT, 1LONG and 5SHORT were grouped on the upper

2 5SHORT 2,6-dimethoxyphenol 2-Furanmethanol 4-methyl-2-methoxyphenol 3-methylbutanal 1 Benzeneacetaldehyde 2-methoxyphenol 1-octen-3-ol 4-ethyl-2-methoxyphenol 2-Heptanone Hexanal 2SHORT 2-Butanone 3SHORT 1SHORT Methanethiol 1LONG Carbon disulfide 0 Benzaldehyde Limonene Methional 61 4SHORT - 1 3LONG

Principal Component 2 (26%) (26%) 2 Component 2 Component Principal Principal (26%) 2LONG

- 2

- 3 - 5 -4-3- 4 - 3 -2-1- 2 - 1 0 1 2 3 4 5 Principal Component 1 (52%)

Figure 5. Principal Component Analysis Biplot of the aroma impact compounds of eight American dry-cured hams by SPME- GCMS.

left part of the graph (positive PC2 dimension) and were characterized by 4-methyl-2- methoxyphenol, 4-ethyl-2-methoxyphenol, 2-methoxyphenol, 2,6-dimethoxyphenol and

2-furanmethanol. On the other hand, dry-cured hams (3SHORT, 3LONG) located on the positive PC1 dimension were associated with methanethiol, carbon disulfide, 2- butanone, 3-methylbutanal, 2-heptanone, hexanal, benzeneacetaldehyde, 1-octen-3-ol, benzaldehyde and limonene which can be derived from lipid oxidation, amino acid degradation or Maillard reaction. The third group was comprised of 4SHORT, 2LONG and 3LONG with methional as the main aroma impact compound.

Consumer Acceptability

Differences existed (P<0.05) among dry-cured ham products for overall acceptability, flavor, aroma, texture and appearance (Table 8). On average, consumers preferred (P<0.05) the short-aged and smoked dry-cured ham products, but almost all hams had average consumer scores between 6 and 7, signifying “like slightly” to “like moderately”. According to the descriptive analysis carried out by the trained panel

(Figures 6 and 7), the dry-cured ham products which received higher consumer acceptability scores had more intense “sweet”, “savory”, “smoky”, “caramelized” and

“molasses” flavor and aroma attributes. The dry-cured products which received lower consumer acceptability scores had more intense “rancid”, “cured”, “fermented”, “pork complex”, “earthy”, “bitter”, “salty”, “saltburn”, “aftertaste” flavor and aroma attributes. The mean texture acceptability ranged from 5.2 to 7.2 signifying “neither like nor dislike” to “like moderately”. Hedonic responses for texture acceptability were

Table 8. Mean scores for consumer acceptability (N=71) of eight American dry-cured ham treatments with varied aging periods and processing conditions 1.

American Dry-cured Ham1 Overall Acceptability Flavor Aroma Appearance Texture 1SHORT 6.3ab 6.3ab 6.0cd 7.1a 6.9a 2SHORT 7.0a 7.2a 7.4a 6.8a 7.2a 3SHORT 5.6bc 5.6bc 5.6de 6.6ab 5.8bc 4SHORT 6.8a 7.0a 6.4bc 6.5abc 7.1a 5SHORT 6.8a 6.7a 6.3bc 7.2a 7.1a 1 LONG 6.8a 6.9a 6.9ab 6.5abc 6.8a 2LONG 6.2ab 6.3ab 5.8cde 5.9bc 6.5ab

63 3LONG 5.1c 5.2c 5.3e 5.6c 5.2c a-e Means with different letter within each column are significantly different (P<0.05). 1 Scores were based on a 9-point hedonic scale (1=dislike extremely, 5=neither like nor dislike, 9=like extremely).

2 Savory 1LONG

Sweet Cluster 2 Cluster 3 Cluster 6 Bitter Salty 1 Cluster 5 Cured Cluster 4 2SHORT Saltburn 4SHORT Aftertaste Astringency 5SHORT 0 3LONG Rancid

64 2LONG 3SHORT Mouthdrying Cluster 1

Principal Component 2(22%) 2(22%) Component PrincipalComponent2(22%) Principal 1SHORT - 1

Pork Complex

- 2

- 4 - 3 - 2 - 1 0 1 2 3 4 - 3 Principal Component 1 (68%) -4 -3 -2 -1 0 1 2 3 4 Figure 6. External Preference Map of combined consumer data with descriptive analysis data for flavor attributes of eight American dry-cured hams.

2 Earthy Cluster 3 Rancid 3LONG Caramelized 1LONG Molasses Smoky 1 3SHORT Fermented Savory Pork Complex

4SHORT 2SHORT

0 1SHORT

65 5SHORT

Cluster 2 Principal Component 2(20%) 2(20%) Component PrincipalComponent2(20%) Principal - 1 2LONG Cluster 6 Cluster 1 Cluster 5

Cluster 4

- 2

- 4 - 3 - 2 - 1 0 1 2 3 4

- 3 Principal Component 1 (65%) -4 -3 -2 -1 0 1 2 3 4 Figure 7. External Preference Map of combined consumer data with descriptive analysis data for aroma attributes of eight American dry-cured hams.

different (P<0.05) among the dry-cured ham products. 3SHORT, 2LONG and 3LONG received the lowest mean texture acceptability scores. These samples had higher descriptive intensity scores for hardness, dryness, fibrousness and chewiness which suggests that consumers do not like the texture of dry-cured ham if it is too hard, dry, fibrous or chewy (Figure 3).

Agglomerative hierarchical clustering (AHC) was performed to further elucidate the panelists’ attitude toward the dry-cured ham products. The dendrogram (Figure A3) generated by the AHC indicated the presence of six clusters based on the consumer preference for American dry-cured ham. Randomized complete block designs were performed within each cluster to differentiate among the dry-cured ham products (Table

9). Cluster 1 (10% of the consumers) ranged from neither like nor dislike to like very much in their preference for the dry-cured ham products but were partial to 1SHORT and 5SHORT and 4SHORT and 2LONG, which were manufactured by the same company. Cluster 2 (37% of the consumers) contained consumers who liked all the products between “like slightly” and “like very much”. This consumer group preferred the smoked dry-cured ham products, specifically 2SHORT, 5SHORT and 1LONG.

Cluster 3, which included 17% of the consumers, preferred 2SHORT, 3SHORT,

4SHORT, and 1LONG. This consumer group did not like non-smoked hams with long aging periods or 1SHORT, the sample with the shortest aging time. Cluster 4 (11% of the consumers) ranged from “dislike moderately” to “like moderately” in their preference and were partial to short-aged, smoked dry-cured ham products, specifically

4SHORT and 2SHORT. Cluster 5 (14% of the consumers) preferred 1SHORT, 1LONG

Table 9. Mean scores for overall consumer acceptability of eight American dry-cured ham treatments according to different clusters of consumer segments using a hedonic scale1.

Cluster Consumers American Dry-cured Ham Treatments1 (N) 1SHORT 2SHORT 3SHORT 4SHORT 5SHORT 1LONG 2LONG 3LONG 1 7 8.0a 5.1b 5.1b 7.9a 8.3a 5.6b 8.1a 5.6b 2 26 6.8c 7.7ab 7.0c 7.3bc 7.7ab 8.0a 7.0c 6.0d 3 12 4.9c 7.8a 6.8ab 7.1ab 6.1b 6.6b 3.7d 4.9c 4 8 4.8cde 7.4ab 3.9ef 7.7a 5.4cd 4.4def 6.1bc 3.1f 5 10 6.6a 5.6ab 3.4c 5.7ab 4.9b 6.7a 6.8a 5.5ab 6 8 6.8a 6.9a 4.3bc 4.9b 6.9a 6.9a 4.6bc 3.6c 67 a-f Means with different letter within each row are significantly different (P<0.05). 1 Scores were based on a 9-point hedonic scale (1=dislike extremely, 5=neither like nor dislike, 9=like extremely).

and 2LONG with mean scores ranging from 6.6-6.8 on a 9-point hedonic scale. Cluster

6 which comprised of 11% of the consumers preferred 1SHORT, 2SHORT, 5SHORT and 1LONG. This segment of panelists rated the smoked dry-cured ham products as like slightly. These hams were characterized by phenolic compounds and 2-furanmethanol.

This reveals that these consumers were partial to a strong smoked flavor. The overall acceptability of the dry-cured ham products across all clusters were generally reflected by flavor, aroma and texture liking trends.

External Preference and External Flavor Mapping: Relating aroma impact compounds, sensory descriptors and consumer acceptability

External preference mapping was conducted to relate the descriptive analysis data with the preferences of each consumer cluster and understand the sensory descriptors that drive the consumer liking for American dry-cured ham. This technique maps the consumer hedonic data onto an already existing product map from the PCA biplot of the descriptive analysis data. This map shows the descriptive analysis data as predictor variables and the consumer acceptability data as response variables (Schlich,

1995; Guinard, 1998; 2002). When the consumer hedonic data was overlaid on the

PCA biplot, it was apparent that the dry-cured ham products (2SHORT, 4SHORT,

5SHORT, 1LONG) preferred by a majority of the clusters were also the most “sweet” and “savory” (Figure 6). Dry-cured ham products (3SHORT, 3LONG) characterized by more intense “cured”, “rancid”, “aftertaste”, “pork complex”, “mouthdrying”,

“saltburn”, “astringent”, “bitter” and “salty” flavors received lower acceptability scores.

Clusters 1, 3 and 4 were well explained by their flavor attributes compared with the

other clusters. External preference mapping of aroma attributes (Figure 7) showed that the positive drivers for liking of American dry-cured hams (2SHORT, 4SHORT,

5SHORT, 1LONG) were “caramelized”, “molasses”, “smoky” and “savory” aromas while those dry-cured ham products (3SHORT, 3LONG) which received lower consumer acceptability scores were described as having stronger “fermented”, “rancid”,

“earthy” and “pork complex” aromas. All clusters except for cluster 5 were well described by their aroma sensory descriptors. External preference mapping of textural attributes (Figure 8) revealed that liking for the majority of the clusters was linked to dry-cured ham products (1SHORT, 2SHORT, 4SHORT, 5SHORT, 1LONG) with juicy and mushy textures. The least preferred hams (3SHORT, 3LONG) were characterized by hard, fibrous, chewy and dry textural attributes. Clusters 1 and 6 were more precisely differentiated by their textural attributes compared with the other clusters.

External flavor mapping was performed to relate the concentrations of the aroma impact compounds with the consumer hedonic data and understand which aroma impact compounds contribute to the generation of these flavors and aromas that impact the consumer liking for this product. Only aroma impact compounds which were detected by two out of three trained panelists and present in their sensory threshold range were included. Even though descriptive analysis was not performed to evaluate the dry-cured ham extract, headspace techniques are believed to yield representative odorant isolates as they extract the volatile compounds that surround food and are therefore perceived by the olfactory system (Carrapiso and García, 2004).

2 Mushiness Cluster 6 Cluster 3 Cluster 2 2SHORT Cluster 4 Dryness 3LONG 1 1LONG Hardness Cluster 5 Fibrousness Juiciness 5SHORT Chewiness

4SHORT 0 1SHORT 2LONG 70 3SHORT Cluster 1 Principal Component 2(9%) 2(9%) Component PrincipalComponent2(9%) Principal - 1

- 2

-4- 4 -3- 3 -2- 2 -1- 1 0 1 2 3 4 Principal Component 1 (85%) - 3

Figure 8. External Preference Map of combined consumer data with descriptive analysis data for textural attributes of eight American dry-cured hams.

An external flavor map of the aroma impact compounds with the consumer hedonic data reveals that 52% and 26% of the total variance is explained by the PC1 and PC2 dimensions, respectively (Figure 9). The distribution of the aroma impact compounds on the first two PCs shows three separate groups. The majority of the clusters preferred dry-cured ham products (1SHORT, 2SHORT, 1LONG, 5SHORT) that were characterized by 4-methyl-2-methoxyphenol, 4-ethyl-2-methoxyphenol, 2- methoxyphenol, 2,6-dimethoxyphenol and 2-furanmethanol located in the positive PC2 dimension. In fact, these aroma impact compounds have been previously associated with smoke or smoked food products. Clusters 5 and 6 were better characterized by their aroma impact compounds than the other clusters. Cluster 5 preferred three dry- cured ham products (4SHORT, 2LONG, 3LONG) with methional as the dominant aroma impact compound. The positive PC1 dimension contains the aroma impact compounds methanethiol, carbon disulfide, 2-butanone, 3-methylbutanal, 2-heptanone, hexanal, benzeneacetaldehyde, 1-octen-3-ol, benzaldehyde and limonene which characterized two dry cured ham products (3SHORT, 3LONG) and whose formation pathways have been associated with lipid oxidation or amino acid degradation.

The sweet, smoky and savory flavors and aromas in American dry-cured ham may be attributed to phenolic compounds including 4-methyl-2-methoxyphenol, 4- ethyl-2-methoxyphenol, 2-methoxyphenol, 2,6-dimethoxyphenol and 2-furanmethanol which characterized smoked dry-cured ham products (1SHORT, 2SHORT, 5SHORT,

1LONG) that were preferred by the majority of the clusters (Figures 6, 7 and 9). During olfactometry, phenolic compounds were described by panelists as having sweet ham,

2 Cluster 6 5SHORT 2,6-dimethoxyphenol 2-Furanmethanol 4-methyl-2-methoxyphenol Cluster3-methylbutanal 2 1 Benzeneacetaldehyde 2-methoxyphenol Cluster 31-octen-3-ol 4-ethyl-2-methoxyphenol 2-Heptanone Hexanal 2SHORT 2-Butanone 3SHORT 1SHORT Methanethiol 1LONG Carbon disulfide 0 Cluster 1 Benzaldehyde Cluster 4 Limonene Methional

72 4SHORT - 1 Cluster 5 3LONG

(26%) 2 Component 2 Component (26%) Principal Principal 2LONG

- 2

- 3 - 5 -4-3- 4 - 3 -2-1- 2 - 1 0 1 2 3 4 5 Principal Component 1 (52%)

Figure 9. External Preference Map of combined consumer data with the volatile composition of eight American dry-cured hams.

smoky ham, savory and cocoa aromas (Table 7). Many studies have indicated that

phenolic compounds present in the vapor phase of smoke may contribute to imparting a

smoky flavor to foods (Varlet et al., 2006). Only twenty phenols have been characterized in smoked food products (Toth and Potthast, 1984). Two of these compounds 2-methoxyphenol and 4-methyl-2-methoxyphenol were identified in the headspace of American dry-cured ham. It has been reported that these compounds are strong contributors to pleasant smoke-like aroma (Shahidi and Naczk, 2003). To our knowledge, these phenolic compounds have never been reported in American dry-cured ham, although they have been identified in smoked meats and fish (Baloga et al., 1990;

Wittkowski et al., 1992; Poligné et al., 2002; Varlet et al., 2006). Phenolic compounds may be key contributors to the distinct flavor of American dry-cured ham that impacts the consumer liking for this product. 2-furanmethanol which exhibited burnt meat and vitamin–like aromas has several suggested formation pathways such as the Maillard reaction (Mottram, 1998), during the smoking process by thermal degradation (Rozum,

1998) or as a result of the deamination and dehydration of Amadori products during meat heat treatments (Mottram, 1991). Furans contribute to the burnt or cooked and roasted aromas in food products (Varlet et al., 2006). Their low threshold values and pleasant aroma make these compounds important contributors to the desirable aroma of dry-cured ham products (Ramirez and Cava, 2007).

Methional, characterized by a baked potato odor, was dominant in two dry-cured ham products that were manufactured by the same company (4SHORT, 2LONG) and preferred by Cluster 5 (Table 7) as well as in 1LONG and 3LONG. This compound

was detected by all panelists in all dry-cured ham products. It was also the most potent aroma, as indicated by the highest aroma intensity (Table 7). The Strecker degradation of methionine yields methional and other sulfur-containing intermediates (Mottram,

1991). Even though this sulfur-containing compound was present in relatively low concentrations in the headspace, its low odor threshold makes it an important contributor to the distinct flavor of this product. Although the baked potato aroma of methional was strongly detected by olfactometry (Table 7), it was not perceived during descriptive analysis. This demonstrates the challenge that exists in relating the descriptive analysis data with olfactometry results since aromas are evaluated separately and out of the food matrix during olfactometry while aromas are blended and may interact with other constituents during descriptive analysis (Le Guen et al., 2000).

The fermented/malty and floral/rose - like aromas of 3-methylbutanal and benzeneacetaldehyde are associated with the “cured” and “fermented” descriptors used to describe American dry-cured ham during descriptive analysis (Figures 6, 7 and 9).

Methyl-branched aldehydes such as 3-methylbutanal and benzeneacetaldehyde have been established as major flavor contributors in other dry-cured hams such as Serrano,

Iberian, Bayonne and Corsican due to their low odor thresholds and distinctive aroma characteristics (Dirinck et al., 1997; Dirinck and Opstaele, 1998; Sabio 1998). Careri et al. (1993) reported that 3-methylbutanal is a key contributor to the aged flavor of Parma hams while Ruiz et al. (1999) has linked the high acceptability of Iberian hams to the high proportion of this compound in that particular product. Contrary to the results of

Ruiz et al. (1999), 3-methylbutanal and benzeneactaldehyde characterized a dry-cured

ham product (3SHORT) which received lower consumer acceptability scores in this study. The differences could be due to different manufacturing techniques and raw material that can influence the final sensory quality of the product. Formation of these compounds has been attributed to the oxidative deamination-decarboxylation of amino acids through Strecker degradation and could be related to proteolysis (Dirinck et al.,

1997).

Linear saturated and unsaturated aldehydes with more than five carbon atoms are typical of fat degradation formed by lipolysis autoxidation mechanisms (Belitz and

Grosch, 1987) and have been associated with rancidity notes which develop during the storage of fatty foods (Mottram, 1998). 3SHORT had the highest concentration of hexanal while benzaldehyde was dominant in 3LONG (Figure 9, Table 6). These aldehydes are very potent aroma compounds with low odor thresholds, and these compounds may play a major role in the “rancid” flavors and aromas characterized by the trained panel during descriptive analysis. Hexanal was the most abundant aroma impact compound in the headspace of 3SHORT and 3LONG (Table 6) both of which received lower consumer acceptability scores when compared with other dry-cured ham products. Regarded as a good indicator of the oxidative state of meat lipids, it is the main volatile derived from the oxidation of linoleic and arachidonic acids, the n-6 fatty acids (Shahidi, 1998). High concentrations of hexanal signal flavor deterioration in meat products often resulting in a rancid aroma (Cava et al., 2000; Ramirez and Cava,

2007). A significant volatile constituent in the flavor of meat, benzaldehyde is also one of the main volatile products obtained when defatted meat is heated (Kato et al., 1973;

Macleod and Ames, 1987; Mottram, 1991). 3LONG contained relatively higher concentrations of benzaldehyde compared with the other dry-cured ham products (Table

6) and included extra lean cuts with the skin and fat removed. This compound maybe generated through the Strecker degradation of the amino acid phenylglycine (Mottram,

1991; 1998).

The “pork complex” flavor and aroma attributes may be related to the fruity, spicy and fatty notes of methyl ketones (Dirinck and Opstaele, 1998). 2-butanone and 2- heptanone were detected during olfactometry with sweet/alcoholic/milky and burnt meat/vitamin aromas, respectively (Figure 9, Table 7). They were dominant in 3SHORT which received lower consumer acceptability scores in comparison to many other short aged hams in the study. Methyl ketones arise from the autoxidation (Ramirez and Cava,

2007) or beta oxidation (Dirinck and Opstaele, 1998) of fatty acids and are considered to be among the meat flavor precursors that contribute to the fatty aromas associated with cooked meat (Shahidi, 1998).

The mushroom/metallic-like aroma of 1-octen-3-ol was detected during olfactometry and had higher concentrations in 3SHORT and 5SHORT in comparison to other hams (Figures 9, Tables 6 and 7). This compound has been identified as a major contributor to dry-cured ham flavor due to its low odor threshold (Dirinck et al., 1997;

Sabio et al., 1998; Carrapiso et al., 2002). 1-octen-3-ol has often been described as an important component of meat volatiles particularly with the fatty characteristics of meat flavors and as a product of the autoxidation of linoleic acid or other polyunsaturated fatty acids (Mottram, 1991). Its association with fatty flavors in meat suggests its

possible contribution to the “pork complex” flavor and aroma attributes in American

dry-cured ham, which also characterized the least preferred short aged ham, 3SHORT.

The citrus-like aroma of limonene was not perceived during descriptive analysis,

but it was detected by olfactometry. Limonene in dry-cured ham has been related to

diet, where the compound originates from the feed and accumulates in the body of the

animal (Ruiz et al.,1999; Sánchez-Peña et al., 2005). The high concentration of this compound in Bayonne hams has been attributed to the black pepper treatment on the surface of the dry-cured ham during processing (Sabio et al., 1998). According to

Richard (1991), limonene is an important component of black pepper essential oil where mono- and sesquiterpenes/hydrocarbons make up 90% of its content. Similarly, the limonene compound characterized 3LONG, and this ham was manufactured with black and red peppers and no nitrates or nitrites in the process. This ham also received lower consumer acceptability scores when compared with the other dry-cured ham products.

Although not detected by the panelists during descriptive analysis, the sulfurous notes of the aroma-impact compounds methanethiol and carbon disulfide were important contributors to the aroma of American dry-cured ham, especially in 3SHORT and 3LONG which were characterized by these compounds (Figure 9, Table 7). The concentrations of methanethiol and carbon disulfide in the headspace of American dry- cured ham were consistently higher than their odor threshold values, revealing their important contribution to the overall aroma of this product. These compounds contribute directly to the meaty characteristics and provide an overall sulfurous note as

a part of meat aroma. Sulfur compounds in meat have been attributed to the side-chains of sulfur containing amino acids in protein or further reactions with other compounds in meat (Mottram, 1991; Fan et al., 2002).

Nearly all of the aroma impact compounds in this study have been previously reported as contributors to the distinct flavor of dry-cured ham products. 2- furanmethanol, methional, and the phenolic compounds were essentially the aroma impact compounds that positively impacted the consumer acceptability of American dry-cured ham in this study (Figure 9). On the other hand, the most differentiating compounds for the less preferred dry-cured ham products were reported to have been generated through lipid oxidation or amino acid degradation. Volatile composition, sensory characteristics, and low consumer acceptability of 3SHORT were not well explained since it was smoked and contained essentially the same cure ingredients as many other hams evaluated in the study. Variation in the aroma impact compounds and sensory relationships among the individual dry-cured hams may be due to varying factors such as processing techniques and type of raw material which can be easily affected by age, rearing condition, breed, or feeding regime that can determine the final quality of the product.

CHAPTER V

SUMMARY AND CONCLUSIONS

The sensory acceptability of American dry-cured ham could be described and differentiated by a trained sensory panel using twenty-five descriptive attributes.

Complementary instrumental analysis identified sixteen aroma impact compounds that may be responsible for the distinct American dry-cured ham flavor with essentially similar compounds in both short and long aged hams which can be differentiated through the relative concentrations of their aroma impact compounds. External preference and flavor mapping revealed the relationships between sensory descriptors and aroma impact compounds and could be used to comprehend the nature of dry-cured ham flavor as it is perceived by a consumer panel. The majority of the clusters preferred the “sweet”, “smoky” and “savory” flavors and aromas in American dry-cured hams which may be attributed to phenolic compounds including 4-methyl-2- methoxyphenol, 4-ethyl-2-methoxyphenol, 2-methoxyphenol, 2,6-dimethoxyphenol and

2-furanmethanol. These hams also had lower scores for aroma descriptors such as

“earthy”, “rancid”, “fermented” and “pork complex” and flavors such as “bitter”,

“salty”, “cured”, “aftertaste”, “rancid”, and “pork complex”. Dry-cured ham products with “cured” and “fermented” (3-methylbutanal, benzeneacetaldehyde), “rancid”

(hexanal, benzaldehyde), and “pork complex” (2-heptanone, 2-butanone, 1-octen-3-ol)

flavors and aromas received lower acceptability scores when compared to other hams.

Results from this research could be useful in monitoring the authenticity of dry-cured ham products of different origins, in developing a distinct “American dry-cured ham flavor” through aroma recombination, and in further improving product quality based on consumer preference and liking.

REFERENCES

Acree, T.E. 1984. A procedure for the sensory analysis of gas chromatographic effluents. Food Chemistry. 14(4): 273-286.

Acree, T.E. 1993. Gas chromatography-olfactometry. In IFT Basic Symposium Series 8 (Flavor Measurement). C.T. Ho and C.H. Manley, Eds., Marcel Dekker, New York, NY:77-94.

Acree, T. and Arn, H. 2005. Flavornet. Gas Chromatography-Olfactometry (GC-O) of natural products.

Andrés, A.I., Cava, R. and Ruiz, J. 2002. Monitoring volatile compounds during dry- cured ham ripening by solid-phase microextraction coupled to a new direct-extraction device. Journal of Chromatography. A 963(1-2):83-88.

Armero, E., Flores, M., Toldrá, F., Barbosa, J.A., Olivet, J., Pla, M. and Baselga, M. 1999. Effects of pig sire type and sex on carcass traits, meat quality and sensory quality of dry-cured ham. Journal of the Science of Food and Agriculture. 79(9):1147-1154.

Ares, G., Giménez, A. and Gámboro, A. 2006. Preference mapping of texture of dulce de leche. Journal of Sensory Studies. 21(6):553-571.

Baloga, D.W., Reineccius, G.A., Miller, J.W. 1990. Characterization of ham flavor using an atomic emission detector. Journal of Agricultural and Food Chemistry. 38(11):2021-2026.

Barbieri, G., Bolzoni, L., Parolari, G., Virgili, R., Buttini, R., Careri, M. and Mangia, A. 1992. Flavor compounds of dry-cured ham. Journal of Agricultural and Food Chemistry. 40(12):2389-2394.

Belitz, H.D. and Grosch, W. 1987. Food Chemistry. Springer-Verlag: New York, NY.

Berdagué, J.L., Denoyer, C., Le Quere, J.L. and Semon, E. 1991. Volatile components of dry-cured ham. Journal of Agricultural and Food Chemistry. 39(7):1257-1261.

Berdagué, J.L., Bonnaud, N., Rousset, S. and Touraille, C. 1993. Influence of pig crossbreed on the composition, volatile compound content and flavor of dry cured ham. Meat Science. 34(1):119-129.

Blank, I., Devaud, S., Fay, L.B., Cerny, C., Steiner, M. and Zurbriggen, B. 2001. Odor- active compounds of dry-cured meat: Italian-type salami and Parma ham. In ACS Symposium Series 794 (Aroma Active Compounds in Foods): Gary Takeoka, Matthias Guntert, Karl-Heinz Engel, Eds., Oxford University Press, Washington D.C.:9-20.

Bolzoni, L., Barbieri, G. and Virgili, R. 1996. Changes in volatile compounds of Parma ham during maturation. Meat Science. 43(3/4):301-310.

Buscailhon, S., Berdagué, J.L., Bousset, J., Cornet, M., Gandemer, G., Touraille, C. and Monin, G. 1994. Relations between compositional traits and sensory qualities of French dry-cured ham. Meat Science. 37(2):229-243.

Buttery, R.G., Ling, L.C. and Stern, D.J. 1997. Studies on popcorn aroma and flavor volatiles. Journal of Agricultural and Food Chemistry. 45(3):837-843.

Cadwallader, K.R. and Macleod, A.J. 1998. Instrumental methods for analyzing the flavor of muscle foods. In Flavor of Meat, Meat Products and Seafoods, 2nd ed. Fereidoon Shahidi, Ed., Blackie Academic and Professional, UK:355-368.

Careri, M., Mangia, A., Barbieri, G., Bolzoni, L., Virgili, R. and Parolari, G. 1993. Sensory property relationships to chemical data of Italian-type dry-cured ham. Journal of Food Science. 58(5):968-972.

Carrapiso, A.I. and Carrapiso, M.R. 2005. Precursors for dry-cured ham flavor. In ACS Symposium Series 905(Process and Reaction Flavors): Deepthi Weersinghe, Mathias Sucan, Eds., Oxford University Press, Washington D.C.:70-82.

Carrapiso, A.I. and García, C. 2004. Iberian ham headspace: Odorants of intermuscular fat and differences with lean. Journal of the Science of Food and Agriculture. 84(15):2047-2051.

Carrapiso, A.I., Jurado, A., Timón, M.L. and García, C. 2002. Odor-active compounds of Iberian hams with different aroma characteristics. Journal of Agricultural and Food Chemistry. 50(22):6453-6458.

Carrapiso, A.I., Ventanas, J. and García, C. 2002. Characterization of the most odor- active compounds of Iberian ham headspace. Journal of Agricultural and Food Chemistry. 50(7):1996-2000.

Casapia, E.L., Coggins, P.C., Schilling, M.W., Yoon, Y., and White, C.H. 2006. The relationship between consumer acceptability and descriptive sensory attributes in cheddar cheese. Journal of Sensory Studies. 21(1):112-127.

Cassens, R.G. 1995. Use of sodium nitrite in cured meats today. Food Technology. 49(7):72-80.

Cava, R., Ventanas, J., Tejeda, J.F., Ruiz, J. and Antequera, T. 2000. Effect of free- range rearing and tocopherol and copper supplementation on fatty acid profiles and susceptibility to lipid oxidation of fresh meat from Iberian pigs. Food Chemistry. 88(1):51-59.

Civille, G.V. and Lyon, B.G. 1996. Aroma and flavor lexicon for sensory evaluation: terms, definitions, references, and examples. ASTM Data Series Publication DS 66, ASTM, West Conshohocken, PA.

Daun, H. 1972. Sensory properties of phenolic compounds isolated from curing smoke as influenced by its generation parameters. Lebensmittel-Wissenschaft und– Technologie. 5(3):102-105.

Dirinck, P., Van Opstaele, F. and Vandendriessche, F. 1997. Flavor differences between northern and southern European cured hams. Food Chemistry. 59(4):511-521.

Dirinck, P., Van Opstaele, F. 1998. Volatile composition of southern European dry- cured hams. In Developments in Food Science 40 (Food Flavors: Formation, Analysis, and Packaging Influences). E.T. Contis, C.T. Ho, C.J. Mussinan, T.H. Parliament, F. Shahidi, A.M. Spanier, Eds., Elsevier Science, Amsterdam, The Netherlands:233-243.

Dunteman, G.H. 1989. Principal Component Analysis. Newbury Park: SAGE Publications.

Eakes, B.D., Blumer, T.N. and Monroe, R.J. 1975a. Effect of nitrate and nitrite on color and flavor of country-style hams. Journal of Food Science. 40(5):973-976.

Eakes, B.D. and Blumer, T.N. 1975b. Effect of various levels of potassium nitrate and sodium nitrite on color and flavor of cured loins and country-style hams. Journal of Food Science. 40(5):977-980.

Ebeler, S.E. 1999. Linking flavor chemistry to sensory analysis of wine. In Flavor Chemistry: Thirty Years of Progress. Roy Teranishi, Emily L. Wick, Irwin Hornstein, Eds., Kluwer Academic/Plenum Publishers, New York, NY:409-421.

El-Sayed, A.M. 2003. The Pherobase: Database of Insect Pheromones and Semiochemicals. < http://www.pherobase.com/> 83

Fan, X., Sommers, C.H., Thayer, D.W. and Lehotay, S.J. 2002. Volatile sulfur compounds in irradiated precooked turkey breast analyzed with pulsed flame photometric detection. Journal of Agricultural and Food Chemistry. 50(15):4257-4261.

Fang, W.N.G., Langlois, B.E. and Moody, W.G. 1997. Fate of selected pathogens in vacuum-package dry-cured (country-style) ham slices stored at 2 and 25oC. Journal of Food Protection. 60(12):1541-1547.

Flores, M., Aristoy, M.C., Toldrá, F. 1996. HPLC purification and characterization of soluble alanyl aminopeptidase from porcine skeletal muscle. Journal of Agricultural and Food Chemistry. 44(6):2578-2186.

Flores, M., Grimm, C.C., Toldrá, F. and Spanier, A.M. 1997. Correlations of sensory and volatile compounds of Spanish "Serrano" dry-cured ham as a function of two processing times. Journal of Agricultural and Food Chemistry. 45(6):2178-2186.

Flores, M., 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(3):169-178.

Flores, M., Spanier, A.M. and Toldrá, F. 1998. Flavor analysis of dry-cured ham. In Flavor of Meat, Meat Products and Seafoods. 2nd ed. Fereidoon Shahidi, Ed., Blackie Academic and Professional, UK:320-339.

Flores, M., Soler, C., Aristoy, M.C. and Toldrá, F. 2006. Effect of brine thawing/salting for time reduction in Spanish dry-cured ham manufacturing on proteolysis and lipolysis during salting and post-salting periods. European Food Research and Technology. 222(5-6):509-515.

Fu, S.G., Yoon, Y., and Bazemore, R. 2002. Aroma-active components in fermented bamboo shoots. Journal of Agricultural Food Chemistry. 50(3):549-554.

García-Esteban, M., Ansorena, D., Astiasaran, I., Martin, D. and Ruiz, J. 2004. Comparison of simultaneous distillation extraction (SDE) and solid-phase microextraction (SPME) for the analysis of volatile compounds in dry-cured ham. Journal of the Science of Food and Agriculture. 84(11):1364-1370.

Gianelli, M.P., Flores, M. and Toldrá, F. 2002. Optimization of solid phase microextraction (SPME) for the analysis of volatile compounds in dry-cured ham. Journal of the Science of Food and Agriculture. 82(14):1703-1709.

Green, B.G., Dalton, P., Cowart, B. Shaffer, G., Rankin, K. and Higgins, J. 1996. Evaluating the labeled magnitude scale for measuring sensations of taste and smell. Chemical Senses. 21(3):323-334. 84

Grimm, C.C., Champagne, E.T. and Ohtsubo, K.I. 2002. Analysis of volatile compounds in the headspace of rice using SPME/GC/MS. In Food Science and Technology. Volume 115 (Flavor, Fragrance, and Odor Analysis). Ray Marsili, Ed., Marcel Dekker Inc., New York, NY:229-248.

Grob, R.L. and Kaiser, M.A. 2004. Qualitative and quantitative analysis by gas chromatography. In Modern Practice of Gas Chromatography. Robert Grob, Eugene Barry, Eds., John Wiley and Sons Inc., Hoboken, NJ:403-460.

Grosch, W. 1993. Detection of potent odorants in foods by aroma extract dilution analysis. Trends in Food Science & Technology. 4(3):68-73.

Guinard, J.X. 1998. Data collection and analysis methods for consumer testing. In Proceedings of the Third International Food Science and Technology Conference. John R. Whitaker, Norman F. Haard, Charles F. Shoemaker, R. Paul Singh, Eds., Food and Nutrition Press, Davis, CA:504-516.

Guinard, J.X.; Uotani, B. and Schlich, P. 2001. Internal and external mapping of preferences for commercial lager beers: comparison of hedonic ratings by consumers blind versus with knowledge of brand and price. Food Quality and Preference. 12(4): 243-255.

Guinard, J.X. 2002. Internal and external preference mapping: understanding market segmentation and identifying drivers of liking. In ACS Symposium Series 825 (Chemistry of Taste). Peter Given, Dulce Paredes, Eds., Oxford University Press, Washington, D.C.:227-242.

Hinrichsen, L. and Pedersen, S.B. 1995. Relationship among flavor, volatile compounds, chemical changes, and microflora in Italian-type dry-cured ham during processing. Journal of Agricultural and Food Chemistry. 43(11):2932-2940.

Hootman, R.C. 1992. ASTM Manual Series MNL 13 (Manual on descriptive analysis testing). American Society Testing and Material, West Conshohocken, PA.

Huan, Y., Zhou, G., Zhao, G., Xu, X. and Peng, Z. 2005. Changes in flavor compounds of dry-cured Chinese Jinhua ham during processing. Meat Science. 71(2):291-299.

Jay, J.M. 1970. History of microorganisms in food. In Modern Food Microbiology. D. Van Nostrand Co., New York, NY.

Kato, S., Kurata, T., Ishiguro, S. and Fujimaki, M. 1973. Additional volatile compounds produced by pyrolysis of sulfur-containing amino acids. Agricultural and Biological Chemistry. 37:1759-1761.

Keane, P. 1992. The flavor profile. In ASTM Manual Series MNL 13 (Manual on descriptive analysis testing). R.C. Hootman, Ed., American Society Testing and Material, West Conshohocken, PA.

Kemp, J.D., Langlois, B.E., Solomon, M.B. and Fox, J.D. 1979. Quality of boneless dry-cured ham produced with or without nitrate, netting or potassium sorbate. Journal of Food Science. 44(3):914-915.

Krause, A.J., Lopetcharat, K. and Drake, M.A. 2007. Identification of the characteristics that drive consumer liking of butter. Journal of Dairy Science. 90(5):2091-2102.

Leffingwell, J.C., Leffingwell, D. 1991. GRAS Flavor Chemicals-detection Thresholds. Perfumer and Flavorist. 16(1):1-19.

Le Guen, S., Prost, C. and Demaimay, M. 2000. Critical comparison of three olfactometric methods for the identification of the most potent odorants in cooked mussels (Mytilus edulis). Journal of Agricultural and Food Chemistry. 48(4):1307- 1314.

Le Guen, S., Prost, C. and Demaimay, M. 2001. Evaluation of the representativeness of the odor of cooked mussel extracts and the relationship between sensory descriptors and potent odorants. Journal of Agricultural and Food Chemistry. 49(3):1321-1327.

Lei, Q. Chemistry of odor impact volatiles from soy protein concentrates. Diss. University of Kentucky, 2001. Ann Arbor: UMI, 2001. 3018897.

Lillard, D. and Ayres, A. 1969. Flavor compounds in country cured ham. Food Technology. 23(2):251-254.

Lopez, M.O., De la Hoz, L., Cambero, M.I., Gallardo, E., Reglero, G. and Ordonez, J.A. 1992. Volatile compounds of dry hams from Iberian pigs. Meat Science. 31(3):267- 277.

Lowery, D. 1986. Country Ham: The flavor of tradition. Southern Living. 10:102-109.

MacLeod, G. and Ames, J.M. 1987. Effect of water on the production of cooked beef aroma compounds. Journal of Food Science. 52(1): 42-45.

Maga, J.A. 1988. Smoke In Food Processing, CRC Press, Incorporated, Boca Raton, FL.

Marriott, N.G., Tracy, J.B., Kelly, R.F. and Graham, P.P. 1983. Accelerated processing of boneless hams to dry-cured state. Journal of Food Protection. 46(8): 717.

Marriott, N.G., Graham, P.P., Shaffer, C.K. and Phelps, S.K. 1987. Accelerated production of dry-cured hams. Meat Science. 19(1):53-64.

Marriott, N.G., Graham, P.P. and Claus, J.R. 1992. Accelerated dry curing of pork legs (hams): A review. Journal of Muscle Foods. 3(2):159-168.

Marriott, N.G. and Ockerman, H.W. 2004. The Ultimate Guide to Country Ham, An American Delicacy. Brightside Press, Radford, VA.

Marriott, N.G. and Schilling, M.W. 2004. Dry Cured Pork Research Review: National Country Ham Association, Inc. National Country Ham Association Annual Meeting. April 2-4, 2004. Morehead City, NC:1-62.

Marsili, R.T. 2007. Comparing sensory and analytical chemistry flavor analysis. In Sensory-Directed Flavor Analysis, Ray Marsili, Ed., CRC Press, Boca Raton, FL:1-22.

Masucci, J.A. and Caldwell, G.W. 2004. Techniques for gas chromatography/mass spectrometry. In Modern Practice of Gas Chromatography. Robert Grob, Eugene Barry, Eds., John Wiley and Sons Inc., Hoboken, NJ:339-402.

McCain, G.R., Blumer, T.N., Craig, H.B. and Steel, R.G. 1968. Free amino acids in ham muscle during successive aging periods and their relation to flavor. Journal of Food Science. 33(2):142-146.

McDaniel, M.R., Miranda-Lopez, R., Watson, B.T., Michaels, N.J. and Libbey, L.M. 1990. Pinot noir aroma: a sensory/gas chromatographic approach. In Flavor and Off- flavors. Proceedings of the 6th International Flavor Conference: Rethymnon, Crete, Greece, July 5-7; Charalambous, G., Ed., Elsevier Science Publishers B.V., Amsterdam, The Netherlands:23.

McEwan, J.A. 1996. Preference mapping for product optimization. In Multivariate Analysis of Data in Sensory Science. Tormod Naes, Einar Risvik, Eds., Elsevier, Amsterdam, the Netherlands:71-102.

McMaster, M. and McMaster, C. 1998. GC/MS. A Practical User’s Guide. John Wiley and Sons Inc., New York, NY.

McGorrin, R.T. 2007. Character-impact flavor compounds. In Sensory-Directed Flavor Analysis. Ray Marsili, Ed., CRC Press, Boca Raton, FL:223-268.

Meilgaard, M., Civille, G.V., and Carr, B.T. 1999. Sensory Evaluation Techniques. 3rd ed. CRC Press, Boca Raton, FL.

Mistry, B.S., Reineccius, G. and Olson, L.K. 1997. Gas chromatography-olfactometry for the determination of key odorants in foods. In Food Science and Technology. 79(Techniques for Analyzing Food Aroma):265-292. Mottram, D.S. 1991. Meat. In Volatile Compounds in Foods and Beverages, Hank Maarse, Ed., Marcel Dekker Inc., New York, NY:107-202.

Mottram, D.S. 1991. Meat. In Volatile Compounds in Foods and Beverages, Henk Maarse, Ed., Marcel Dekker Inc., New York, NY: 107-178.

Mottram, D.S. 1998. The chemistry of meat flavor. In Flavor of Meat, Meat Products and Seafoods, 2nd ed., Fereidoon Shahidi, Ed., Blackie Academic and Professional, UK: 5-26.

Mottram, D.S. 1999. Meat flavour. In Understanding Natural Flavours. John R. Piggott, Alistair Paterson, Eds., Aspen Publishers Inc., Gaithersburg, MD:141-161.

Muñoz, A.M. and Civille, G.V. 1992. Spectrum descriptive analysis. In ASTM Manual Series MNL 13 (Manual on descriptive analysis testing). R.C. Hootman, Ed., American Society Testing and Material, West Conshohocken, PA.

Murray, J.M. and Delahunty, C.M. 2000. Mapping consumer preference for the sensory and packaging attributes for Cheddar cheese. Food Quality and Preference. 11(5):419- 435.

Ockerman, H.W., Blumer, T.N. and Craig, H.B. 1964. Volatile chemical compounds in dry-cured hams. Journal of Food Science. 29(2):123-129.

Ott, A., Fay, L. B. and Chaintreau, A. 1997. Determination and origin of the aroma impact compounds of yogurt flavor. Journal of Agricultural and Food Chemistry 45(3):850-858.

Pawliszyn, J. 2001. Solid Phase Microextraction. In Headspace Analysis of Food and Flavors: Theory and Practice: Russell Rousseff, Keith Cadwallader, Eds., Kluwer Academic/ Plenum Publishers, New York, NY:73-87.

Pérez-Juan, M., Flores, M. and Toldrá, F. 2006. Generation of volatile flavour compounds as affected by the chemical composition of different dry-cured ham sections. European Food Research and Technology. 222(5-6):658-666.

Piotrowski, E.G., Zaika, L.L. and Wasserman, A.E. 1970. Studies on aroma of cured ham. Journal of Food Science. 35(3):321-325.

Poligné, I., Collignan, A., Trystram, G. 2002. Effects of salting, drying, cooking, and smoking operations on volatile compound formation and color patterns in pork. Journal of Food Science 67(8):2976-2986.

Pollien, P., Ott, A., Montigon, F., Baumgartner, M., Munoz-Box, R. and Chaintreau, A. 1997. Hyphenated Headspace-gas chromatography-sniffing technique: screening of impact odorants and quantitative aromagram comparisons. Journal of Agricultural and Food Chemistry 45(7):2630-2637.

Portocarrero, S.M., Newman, M. and Mikel, W.B. 2002. Reduction of Listeria monocytogenes, Salmonella spp., and Escherichia coli O157:H7 during processing of country-cured hams. Journal of Food Science. 67(5):1892-1898.

Qian, M., Nelson, C. and Bloomer, S. 2002. Evaluation of fat-derived aroma compounds in blue cheese by dynamic headspace GC/olfactometry-MS. Journal of the American Oil Chemists' Society. 79(7):663-667.

Qian, M.C., Burbank, H.M. and Wang, Y. 2007. Preseparation techniques in aroma analysis. In Sensory-Directed Flavor Analysis, Ray Marsili, Ed., CRC Press, Boca Raton, FL:111-154.

Ramarathnam, N., Rubin, L.J. and Diosady, L.L. 1991. Studies on meat flavor 1: Qualitative and quantitative differences in uncured and cured pork. Journal of Agricultural and Food Chemistry. 39(2):344-350.

Ramirez, R. and Cava, R. 2007. Volatile profiles of dry-cured meat products from three different Iberian X Duroc Genotypes. Journal of Agricultural and Food Chemistry. 55(5):1923-1931.

Reineccius, G. and Heath, H.B. 1994. Source book of flavors. 2nd ed. Chapman and Hall, New York, NY.

Richard, H.M.J. 1991. Spices and Condiments I. In Volatile Compounds in Foods and Beverages. Henk Maarse, Ed., Marcel Dekker Inc., New York, NY:411-441.

Rozum, J.S. 1998. Smoke flavourings in processed meats. In Flavor of Meat, Meat Products and Seafoods, 2nd ed. Fereidoon Shahidi, Ed., Blackie Academic and Professional, UK:342-354.

Rousseff, R., Bazemore, R., Goodner, K. and Naim, M. 2001. GC-olfactometry with solid phase microextraction of aroma volatiles from heated and unheated orange juice. In Headspace Analysis of Food and Flavors: Theory and Practice. Russell Rousseff, Keith Cadwallader, Eds., Kluwer Academic/Plenum Publishers, New York, NY:101- 111. 89

Rowe, D.J. 2002. High impact aroma chemicals. In Advances in Flavours and Fragrances: From the Sensation to the Synthesis, Special Publication - Royal Society of Chemistry 277. Karl A.D. Swift, Ed., Royal Society of Chemistry, Cambridge, UK:202- 226.

Ruiz, J., Cava, R., Ventanas, J. and Jensen, M.T. 1998. Headspace solid phase microextraction for the analysis of volatiles in a meat product: Dry-Cured Iberian Ham. Journal of Agricultural and Food Chemistry. 46(11):4688-4694.

Ruiz, J., Ventanas, J., Cava, R., Timón, M.L. and García, C. 1998. Sensory characteristics of Iberian ham: influence of processing time and slice location. Food Research International. 31(1):53-58.

Ruiz, J., Ventanas, J., Cava, R., Andrés, A. and García, C. 1999. Volatile compounds of dry-cured Iberian ham as affected by the length of curing process. Meat Science. 52(1):19-27.

Ruiz, J., Ventanas, J. and Cava, R. 2001. New device for direct extraction of volatiles in solid samples using SPME. Journal of Agricultural and Food Chemistry. 49(11):5115- 5121.

Sabio, E., Vidal-Aragon, M.C., Bernalte, M.J. and Gata, J.L. 1998. Volatile compounds present in six types of dry-cured ham from south European countries. Food Chemistry. 61(4):493-503.

Sánchez-Peña, C.M., Luna, G., Garcia-Gonzalez, D.L. and Aparicio, R. 2005. Characterization of French and Spanish dry-cured hams: influence of the volatiles from the muscles and the subcutaneous fat quantified by SPME-GC. Meat Science. 69(4):635-645.

Schilling, M.W., and Coggins, P.C. 2007. Utilization of agglomerative hierarchical clustering in the analysis of hedonic scaled consumer acceptability data. Journal of Sensory Studies. 22:477-491.

Schlich, P. 1995. Preference mapping: Relating consumer preferences to sensory or instrumental measurements. In Bioflavour’95: Analysis/Precursor Studies/ Biotechnology. Patrick Etiévant, Peter Schreier, Eds., INRA Editions, Versailles, France.

Shahidi, F. 1998. Flavour of muscle foods – An overview. In Flavor of Meat, Meat Products and Seafoods. 2nd ed. Fereidoon Shahidi, Ed., Blackie Academic and Professional, UK:1-4.

Shahidi, F. and Naczk, M. 2003. Contribution of phenolic compounds to the flavor and color characteristics of foods. In Phenolics in Food and Nutraceuticals. Fereidoon Shahidi, Marian Naczk, Eds., CRC Press, Boca Raton, FL:443-481.

Spanier, A.M., Flores, M., McMillin, K.W. and Bidner, T.D. 1997. The effect of postmortem aging on meat flavor quality in Brangus beef. Correlation of treatments, sensory, instrumental and chemical descriptors. Food Chemistry. 59(4):531-538.

Stalder, K.J., Saxton, A.M., Miller, R.K. and Goodwin, R.N. 2003. A comparison of hydrogen ion concentration and pH genetic predictions and fixed effect estimations when assessing pork quality. Journal of Animal Science. 81(3):611-616.

Statistical Analysis Software. 2002. SAS Companion for the Microsoft Windows Environment (Version 9.1). SAS Institute Inc., Cary, NC.

Steffen, A. and Pawliszyn, J. 1996. Analysis of flavor volatiles using headspace solid- phase microextraction. Journal of Agricultural and Food Chemistry. 44(8):2187-2193.

Stevens, S.S. 1971. Issues in psychophysical measurement. Psychological Review. 78(5):426-450.

Stone, H. and Sidel, J.L. 1985. Sensory Evaluation Practices. Academic Press, Orlando, FL.

Thompson, J.L., Drake, M.A., Lopetcharat, K. and Yates, M.D. 2004. Preference mapping of commercial chocolate milks. Journal of Food Science. 69(9):S406-S413.

Timón, M.L., Ventanas, J., Martin, L., Tejeda, J.F. and García, C. 1998. Volatile compounds in supercritical carbon dioxide extracts of Iberian ham. Journal of Agricultural and Food Chemistry. 46(12):5143-5150.

Toldrá, F., Flores, M. and Sanz, Y. 1997. Dry-cured ham flavor: enzymic generation and process influence. Food Chemistry. 59(4):523-530.

Toldrá, F., Flores, M. 1998. The role of muscle proteases and lipases in flavor development during the processing of dry-cured ham. Critical Reviews in Food Science and Nutrition. 38(4):331-352.

Toldrá, F., Aristoy, M.C. and Flores, M. 2000. Contribution of muscle aminopeptidases to flavor development in dry-cured ham. Food Research International. 33(3-4):181-185.

Toldrá, F. 2002. Dry-Cured Meat Products. Food and Nutrition Press, Inc. Trumbull, Connecticut.

Toldrá, F. 2004. Dry-cured ham. In Handbook of Food and Beverage Fermentation Technology: Y.H. Hui, Lisbeth Meunier-Goddik, Ase Sloveg Hansen, Jyette Josephsen, Wai-Kit Nip, Peggy Stanfield, Fidel Toldrá, Eds., CRC Press, New York, NY:369-384.

Toth, L. and Potthast, K. 1984. Chemical aspects of the smoking of meat and meat products. Advances in Food Research. 29:87-158

United States Department of Agriculture. 1999. “Country ham”, ‘Country style Ham”, “Dry Cured Ham”, “Country Pork Shoulder”, “Country Style Pork Shoulder”, and “Dry Cured Pork Shoulder”. Meat and poultry inspection regulations. Title 9. Sec. 319.106. Washington, D.C.:312.

Varlet, V., Knockaert, C., Prost, C. and Serot, T. 2006. Comparison of odor-active volatile compounds of fresh and smoked salmon. Journal of Agricultural and Food Chemistry. 54(9):3391-3401.

Ventanas, J., Cordoba, J.J., Antequera, T., Garcia, C., Lopez-Bote, C. and Asensio, M. A. 1992. Hydrolysis and Maillard reactions during ripening of Iberian ham. Journal of Food Science. 57(4):813-815.

Voltz, J. and Harvell, E.J. 1999. Ham What Am. In The Country Ham Book. The University of North Carolina Press, Chapel Hill, NC:17-21

Wittkowski, R., Ruther, J., Drinda, H. and Rafiei-Taghanaki, F. 1992. Formation of smoke flavor compounds by thermal lignin degradation. In ACS Symposium Series 490 (Flavor precursors: thermal and enzymatic conversions). Roy Teranishi, Gary Takeoka, Matthias Gntert, Eds., Maple Press, York, PA:232-243.

XLSTAT. 2007. Tutorial XLSTAT Version 2007.6. Available from www.xlstatl.com.

Young, N.D., Drake, M.A., Lopetcharat, K. and McDaniel, M.R. 2004. Preference mapping of cheddar cheese with varying maturity levels. Journal of Dairy Science. 87(1):11-19.

Zhang, J., Wang, L., Liu, Y., Zhu, J. and Zhou, G. 2006. Changes in the volatile flavour components of Jinhua ham during the traditional ageing process. International Journal of Food Science and Technology. 41(9):1033-1039.

Zotos, A., Hole, M. and Smith, G. 1995. The effect of frozen storage of mackerel (Scomber scombrus) on its quality when hot-smoked. Journal of the Science of Food and Agriculture. 67(1):43-48.

APPENDIX

Sensory Evaluation of American Dry-cured Ham

Date: Time of Day:______

Panelist Code: Gender: F M

Sample Code:______

0: None 15: High intensity

Directions: Please evaluate the American dry-cured ham sample provided. After chewing the sample please expectorate it in the cup and rinse with the water provided. Please press the green button if you are ready for the next sample.

AROMA

Rancid: Aroma associated with extremely oxidized fat or oil

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Molasses: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Fermented: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Caramelized: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Pork complex: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Figure A1. Score Sheet for Descriptive Sensory Analysis.

Smoky: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Earthy: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Savory: ______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Comments:

APPEARANCE

Color homogeneity: Presence of color homogeneity in the different muscles of a transversal cut

Not Very homogenous homogenous

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Salt crystals: ______No salt Salt crystals crystals |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Comments:

FLAVOR:

Cured: Intensity of the typical flavor from cured meat products

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Figure A1 continued. Score Sheet for Descriptive Sensory Analysis.

Rancid: Flavor associated with extremely oxidized fat or oil

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Salty: The taste on the tongue associated with sodium ions

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

After-taste: Intensity and time extension of the flavor after sample is swallowed

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Pork complex:______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Sweet:______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Savory:______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Bitter:______

|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Astringency: The chemical feeling factor on the tongue described as puckering/dry and associated with strong tea

Not Very astringent astringent |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Figure A1 continued. Score Sheet for Descriptive Sensory Analysis.

Mouth Drying:______

Not Very mouth drying mouth drying |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Salt burn:______No Extreme salt burn salt burn |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Comments:

ORAL TEXTURE

Hardness: Effort required to bite through lean and to convert the sample to a swallowable state Not Very hard hard |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Dryness: Amount of juices which is present in the mouth in the first chews

Juicy Very (not dry) dry |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Fibrousness: Extent to which fibers/strands are perceived on chewing Not Very fibrous fibrous |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Juiciness: Impression of lubricated food during chewing

Not Very juicy juicy |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Figure A1 continued. Score Sheet for Descriptive Sensory Analysis.

Chewiness:______Not Very chewy chewy |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Mushy:______

Not Very mushy mushy |----|----|----|----|----|----|----|----|----|----|----|----|----|----|----| 0 5 10 15

Comments:

Figure A1 continued. Score Sheet for Descriptive Sensory Analysis.

Samples: American Dry-cured Hams Date:______

Please taste each American dry-cured ham sample provided. After chewing if you do not wish to swallow the sample, you may expectorate it in the cup and rinse with the water provided. Rate each sample in each of the five categories listed. Each column will need one check mark.

295 799 380 754 342 885 372 571 AROMA Like extremely Like very much Like moderately Like slightly Neither like nor dislike Dislike slightly Dislike moderately Dislike very much Dislike extremely

295 799 380 754 342 885 372 571 APPEARANCE Like extremely Like very much Like moderately Like slightly Neither like nor dislike Dislike slightly Dislike moderately Dislike very much Dislike extremely

Figure A2. Score Sheet for Consumer Acceptability Tests.

295 799 380 754 342 885 372 571 TEXTURE Like extremely Like very much Like moderately Like slightly Neither like nor dislike Dislike slightly Dislike moderately Dislike very much Dislike extremely

295 799 380 754 342 885 372 571 OVERALL FLAVOR

Like extremely Like very much Like moderately

Like slightly

Neither like nor dislike Dislike slightly Dislike moderately

Dislike very much Dislike extremely

295 799 380 754 342 885 372 571 OVERALL ACCEPTABILITY

Like extremely

Like very much Like moderately

Like slightly

Neither like nor dislike Dislike slightly Dislike moderately

Dislike very much

Dislike extremely Comments:______

Thank you for your participation. We welcome you to return for each of the panels.

Figure A2 continued. Score Sheet for Consumer Acceptability Tests.

100

Dendrogram 138

115

Dissimilarity 101

L3 L3 L8 L7 L1 L1 L4 L5 L5 0 L2 L6 L9 L44 L44 L65 L21 L20 L11 L11 L50 L61 L31 L10 L19 L19 L51 L53 L64 L42 L39 L60 L30 L26 L14 L18 L22 L24 L25 L55 L17 L23 L37 L49 L68 L47 L71 L12 L33 L16 L15 L35 L46 L57 L56 L67 L52 L13 L28 L45 L59 L40 L34 L69 L27 L36 L54 L41 L66 L62 L70 L29 L32 L38 L43 L48 L58 L63

Figure A3. Dendrogram showing how panelists were grouped together into clusters based on dissimilarity of consumer acceptability scores for American dry-cured ham.