EFFECT OF FATTY ACID COMPOSITION ON
THE FLAVOUR OF KOREAN AND AUSTRALIAN BEEF
Elke. M. Stephens B. Ag. Sc (Hons)
A Thesis prepared in partial fulfilment of the requirements for the Degree of
Master in Agricultural Science.
Study undertaken within the Department of Animal Sciences Waite Agricultural Research Institute
and Roseworthy Livestock Systems Alliance, Adelaide University.
Submitted for Examination December 2001
TABLE OF CONTENTS
ABSTRACT…………………………………………………………………………………...i DECLARATION……………………………………………………………………………..ii ACKNOWLEDGEMENTS…….…………………………………………………………...iii CHAPTER 1 - INTRODUCTION...........................................................................................1
1.1 INTRODUCTION ..........................................................................................................2 1.2 CONSUMER PERCEPTIONS OF AUSTRALIAN BEEF IN KOREA .......................3 1.3 HANWOO CHARACTERISTICS.................................................................................5 1.4 TREATMENT OF AUSTRALIAN BEEF IN KOREA.................................................8 1.5 PROJECT AIMS.............................................................................................................8
1.5.1 Specific Aims...........................................................................................................9 CHAPTER 2 - REVIEW OF LITERATURE......................................................................10
2.1 MEAT QUALITY ........................................................................................................11 2.2 ODOUR & FLAVOUR PERCEPTION.......................................................................12
2.2.1 Physiology of Taste (Gustation) ...........................................................................12 2.2.2 Physiology of Olfaction ........................................................................................14 2.2.3 Sensory Physiology ...............................................................................................15
2.3 BEEF FLAVOUR.........................................................................................................17 2.4 DESCRIPTION OF MEAT FLAVOUR ......................................................................28 2.5 ROLE OF LIPIDS, FATS AND FATTY ACIDS - INFLUENCE ON PALATABILITY..................................................................................................................31
2.5.1 Genetic Differences in Flavour – Fatty Acid Composition ..................................36 2.5.2 Nutrition Effects on Flavour – Fatty Acid Composition.......................................39
2.6 PROCESSING AND COOKING INFLUENCES ON FLAVOUR.............................42 2.6.1 Effect of pH on Flavour ........................................................................................42 2.6.2 Effect of Freezing and Thawing on Flavour.........................................................44 2.6.3 Effect of Cooking on Flavour................................................................................45
2.7 SUMMARY..................................................................................................................47 CHAPTER 3 - MATERIALS AND METHODS .................................................................49
3.1 SUMMARY OF MATERIALS AND METHODS......................................................50 3.2 EXPERIMENTAL ANIMALS.....................................................................................50
3.2.1 Selection of Animals for Preliminary Trial...........................................................50 3.2.2 Selection of Animals for Main study .....................................................................52 3.2.3 Selection of Animals for Odour Assessment using a Chemical Sensor ................56
3.3 LABORATORY MEASUREMENT OF FAT TRAITS..............................................56 3.3.1 Muscle Fat Content...............................................................................................56 3.3.2 Melting point of fat ...............................................................................................57 3.3.3 Fatty acid composition..........................................................................................57
3.4 CHEMICAL SENSOR ANALYSIS ............................................................................58 3.5 OBJECTIVE MEASUREMENTS OF TENDERNESS...............................................59 3.6 TASTE PANEL EVALUATION .................................................................................60
3.6.1 Preliminary Study .................................................................................................62 3.6.2 Main Study ............................................................................................................65
3.7 STATISTICAL ANALYSIS ........................................................................................67 CHAPTER 4 - EFFECT OF AGING AND REPEATED FREEZING AND THAWING ON THE EATING QUALITY OF BEEF STRIPLOINS ...................................................68
4.1 INTRODUCTION ........................................................................................................69 4.2 MATERIALS AND METHODS..................................................................................71
4.2.1 Carcase characteristics ........................................................................................71 4.2.2 Treatment allocation.............................................................................................71 4.2.3 Other Measurements.............................................................................................73 4.2.4 Taste Panel Evaluation.........................................................................................74 4.2.5 Statistical Analysis................................................................................................74
4.3 RESULTS .....................................................................................................................76 4.4 DISCUSSION...............................................................................................................82 4.5 CONCLUSIONS ..........................................................................................................87
CHAPTER 5 - CHARACTERISATION OF THE FLAVOUR OF BEEF FROM THE NATIVE KOREAN BREED, THE HANWOO, IN RELATION TO THE FLAVOUR OF BEEF FROM AUSTRALIAN BREEDS........................................................................88
5.1 INTRODUCTION ........................................................................................................89 5.2 MATERIALS AND METHODS..................................................................................90
5.2.1 Selection of Animals for study ..............................................................................90 5.2.2 Other Measurements.............................................................................................91 5.2.3 Statistical Analysis................................................................................................91
5.3 RESULTS .....................................................................................................................93 5.4 DISCUSSION.............................................................................................................110 5.5 CONCLUSIONS ........................................................................................................115
CHAPTER 6 - RELATIONSHIP BETWEEN FLAVOUR AND FATTY ACID COMPOSITION ...................................................................................................................116
6.1 INTRODUCTION ......................................................................................................117 6.2 MATERIALS AND METHODS................................................................................118
6.2.1 Selection of Animals for study ............................................................................118 6.2.2 Fat Measurements...............................................................................................118 6.2.3 Taste Panel Evaluation.......................................................................................118 6.2.4 Statistical Analysis..............................................................................................119
6.3 RESULTS ...................................................................................................................119 6.4 DISCUSSION.............................................................................................................131 6.5 CONCLUSIONS ........................................................................................................137
CHAPTER 7 - CHARACTERISATION OF THE FLAVOUR OF KOREAN AND AUSTRALIAN BEEF USING A CHEMICAL SENSOR.................................................138
7.1 INTRODUCTION ......................................................................................................139 7.2 MATERIALS AND METHODS................................................................................139
7.2.1 Selection of Animals for study ............................................................................139 7.2.2 Fat Measurements...............................................................................................140 7.2.3 Taste Panel Evaluation.......................................................................................140 7.2.4 Chemical Sensor .................................................................................................140 7.2.5 Statistical Analysis..............................................................................................141
7.3 RESULTS ...................................................................................................................141 7.4 DISCUSSION.............................................................................................................153 7.5 CONCLUSIONS ........................................................................................................154
CHAPTER 8 - DEVELOPMENT OF AN EQUATION TO PREDICT FLAVOUR.....155 8.1 INTRODUCTION ......................................................................................................156 8.2 MATERIALS AND METHODS................................................................................156
8.2.1 Selection of Animals for study ............................................................................156 8.2.2 Fat measurements ...............................................................................................156 8.2.3 Taste Panel Evaluation.......................................................................................157 8.2.4 Chemical Sensor .................................................................................................157 8.2.5 Statistical Analysis..............................................................................................157
8.3 RESULTS ...................................................................................................................159 8.4 DISCUSSION.............................................................................................................172 8.5 CONCLUSIONS ........................................................................................................174
CHAPTER 9 - GENERAL DISCUSSION AND CONCLUSIONS .................................175 APPENDICES………………….…………………………………………………………..183 BIBLIOGRAPHY…….………………………………..…………………………………..195
TABLE OF FIGURES
Figure 1.1 - Important Factors from Korean Organisations Perspective – For Customers (MRC Channel Research, CSIRO, 1995)..........................................................................................................................................4 Figure 1.2 - Korean Native Cattle (Hanwoo) in a Korean Feedlot .........................................................................6 Figure 1.3 - Marbling Score ....................................................................................................................................6 Figure 2.1 - Diagrammatic representation of a section through human nasal and buccal cavities ......................14 Figure 2.2 - Compounds Possessing meaty aromas (MacLeod, 1986)..................................................................20 Figure 2.3 - Flavour Wheel (Kuentzel and Bahri, 1991) .......................................................................................29 Figure 3.1 – Striploin treatment allocation for the main study .............................................................................55 Figure 3.2 - Taste Panellist’s conducting tastings for the trial .............................................................................63 Figure 3.3 - Taste Panellist rating samples...........................................................................................................63 Figure 4.1 - Australian Quarter beef destined for the Korean Market ..................................................................69 Figure 4.2 - Australian quarter beef being prepared for freezing .........................................................................70 Figure 4.3 - Aging and Thawing treatments applied to striploins. ........................................................................73 Figure 4.4 - Striploin treatment allocation............................................................................................................74 Figure 4.5 - Correlation between sensory tenderness scores & objective Warner Bratzler shear values.............78 Figure 5.1 - Breed LSMEANS for Initial Juiciness................................................................................................96 Figure 5.2 - Breed LSMEANS for Sustained Juiciness ..........................................................................................96 Figure 5.3 - Breed LSMEANS for Beef Flavour ....................................................................................................97 Figure 5.4 - Breed LSMEANS for Corn Flavour ...................................................................................................98 Figure 5.5 - Breed LSMEANS for Beef Fat Flavour, Oily Flavour and Buttery Flavour......................................99 Figure 5.6 - Breed least square means for Intra-muscular Fat percentage ........................................................102 Figure 5.7 - Breed LSMEANS for Flavour Acceptability ....................................................................................103 Figure 5.8 - Relationship between breed-sex class least square means for bIMF% and cFlavour Acceptability 104 Figure 5.9 - Breed LSMEANS for PRIN1 (Calculated using Model 1 and 2)......................................................107 Figure 5.10 - Breed LSMEANS for PRIN3 (Calculated using Model 1 and 2)....................................................108 Figure 6.1 - Breed Least squares means (%) for Stearic acid and Oleic acid.....................................................122 Figure 6.2 - Breed Least squares means (%) for Myristic acid and Myristoleic acid .........................................123 Figure 6.3 - Breed Least squares means (%) for Palmitic and Palmitoleic acid ................................................124 Figure 6.4 - Breed Least squares means for monounsaturated fatty acids (MUFA’s%) & Melting Point (°C) ..125 Figure 6.5 - Breed Least Squares means for trans-vaccenic and vaccenic acid..................................................126 Figure 7.1 - Ionic abundances for Australian animals shown as a % of Hanwoo...............................................145 Figure 7.2 - Breed Comparison of Ion abundances (I40* & I44*) .....................................................................151 Figure 7.3 - Breed Comparison of Ion abundances (I41ns, I42*, I43* & I45*)...................................................152 Figure 7.4 - Breed Comparison of Ion abundances (I47**, I48** & I60***) ....................................................152 Figure 8.1 - Average Linkage Cluster Analysis for Breed – Fatty Acids.............................................................161 Figure 8.2 - Average Linkage Cluster Analysis for Breed – Flavour ..................................................................162 Figure 8.3 - Average Linkage Cluster Analysis for Project by Country – Fatty Acids........................................163 Figure 8.4 - Average Linkage Cluster Analysis for Breed – Flavours.................................................................164 Figure 8.5 - Average Linkage Cluster Analysis for Breed – Principal Component 1-10 ....................................166 Figure 8.6 - Average Linkage Cluster Analysis for Project – Ions I35 to I180 ...................................................167
TABLE OF TABLES Table 2.1 - Chemical Classes Reported in Cooked Beef (Mottram, 1991) ............................................................18 Table 2.2 - Some reactions generating the meaty aromas of Figure 2.2 (MacLeod, 1986)...................................19 Table 2.3 - Compounds of Cooked Beef Aroma Possessing Relatively High Flavour Dilution Factors ...............27 Table 2.4 - Common descriptors of meat flavour characteristics..........................................................................30 Table 2.5 - Names and Numeric Symbols of some common Fatty Acids in Bovine Adipose and Muscle Tissues .33 Table 2.6 - Correlations between long-chain fatty acids with sensory characteristics of m. longissimus dorsi (LD) of crossbred feedlot steers (Camfield et al., 1997)........................................................................................34 Table 2.7 - Correlations between fatty acids and flavours (Melton et al., 1982a) ................................................34 Table 2.8 - Correlations between individual fatty acids and flavour score in neutral and polar lipid fractions (Melton et al., 1982b) and in SC & IM fat (LD) samples (Westerling & Hedrick, 1979)......................................34 Table 2.9 - MUFA % differences between Japanese Wagyu (produced in Japan) and American Wagyu and Angus beef (fed for 524 days in America). Boylston et al. (1995). ........................................................................38 Table 2.10 - Sensory attributes and pH of meat cooked at different pH values.....................................................43 Table 3.1 - Means, standard deviations and ranges (minimum and maximum) for carcase measurements for 10 Angus and Angus cross pasture fed steers .............................................................................................................51 Table 3.2 - Carcass Characteristics of ‘96 drop SXB Heifers ...............................................................................54 Table 3.3 - Carcass Characteristics of ‘95 drop SXB Steers .................................................................................55 Table 3.4 - Carcass Characteristics of ‘95 drop DGM Steers...............................................................................55 Table 3.5 - Carcass Characteristics of ‘97 drop Hanwoo Steers ..........................................................................55 Table 3.6 - HP 4440 Chemical Sensor Headspace Autosampler Parameters .......................................................59 Table 3.7 - Meat Tasting Score Sheet for Preliminary Trial .................................................................................64 Table 3.8 - Meat Tasting Score Sheet for Main Trial ............................................................................................66 Table 4.1 - Main effects and interactions tested in the initial model - GLM (SAS, 1996). ....................................75 Table 4.2 - Tests of Significance for all attributes tested by the taste panel. .........................................................77 Table 4.3 - Least squares means and standard errors for each treatment as reported by taste panel ratings and tenderness objective measurement.........................................................................................................................78 Table 4.4 - General Linear Models Procedure – Residual Correlations between palatability attributes .............79 Table 4.5 - Least squares means and standard errors for percentage moisture loss for each treatment ..............80 Table 4.6 - Tests of Significance for meat colour and pH attributes. ....................................................................81 Table 4.7 - Least squares means and standard errors for Meat Colour (CIE L*, a*, b*), Fat Colour (CIE L*, a*, b*) and pH for each aging and thawing treatment. ...............................................................................................81 Table 4.8 - Means and standard deviations for fatty acid composition for the two extreme treatments F1 (frozen, one thaw) and A3 (aged, 3 thaws). ........................................................................................................................82 Table 5.1 - Main effects and interactions tested in the basic model using the GLM procedure (SAS, 1996). .......92 Table 5.2 - Analysis of Variance Table for the Basic Modela ................................................................................93 Table 5.3 - Analysis of Variance for the different models (1-4) fitted to the flavoursa ..........................................94 Table 5.4 - Least Square Means for IMFb, Tendernessc, pHc and individual Flavoursc......................................100 Table 5.5 - Least Squares Means for Flavours which were significant for breed by sex class for Model 2 (IMF% fitted as covariate). ..............................................................................................................................................101 Table 5.6 - Estimate of the Slope for IMF% ........................................................................................................102 Table 5.7 - Principal Component Eigenvectors and variation accounted for .....................................................106 Table 5.8 - Analysis of Variance Table for Breed Sex Class for Principal Components .....................................106 Table 5.9 - Residual Correlationsa between Flavours ........................................................................................109 Table 6.1 - Analysis of Variance Table - the effect of sex and breed on fatty acids.............................................120 Table 6.2 - L east Squares Means for Breed Sex Classb ......................................................................................127 Table 6.3 - Correlations between fatty acids .......................................................................................................128 Table 6.4 - Correlations between fatty acids and residual flavours ....................................................................130
Table 7.1 - Ions significant for country, project or breed....................................................................................142 Table 7.2 - LSMEANS for Ions significant for country x project .........................................................................144 Table 7.3 - LSMEANS for Ions significant for breed x country x project ............................................................145 Table 7.4 - Variation accounted for by each Principal Component ....................................................................148 Table 7.5 - Principal Components which were significant for breed ..................................................................148 Table 7.6 - Correlation between fatty acids and flavours with the Principal Components that were significant for breed ....................................................................................................................................................................150 Table 8.1 - Prediction of flavours using fatty acids .............................................................................................168 Table 8.2 - Prediction of flavours using principal components from the chemical sensor data..........................169 Table 8.3 - Prediction of flavours using principal components in addition to fatty acid data.............................171
TABLE OF APPENDICES APPENDIX 1 - Preliminary Trial 184 Figure 1 - Ambient chiller temperature and mean chilling rate of carcasses throughout 21 hour period 184 Figure 2 - Ambient chiller and freezer temperatures during initial freezing and aging of samples (4 days) 184 Table 1 – Treatment allocation for Preliminary Trial 185 APPENDIX 2 - LSMEANS for Ions significant for country 186 APPENDIX 3 - Correlations of Ions with fatty acids and flavours 187 APPENDIX 4 - Eigenvectors from the Principal Components Analysis showing the amount of variation explained by each ion making up Principal components 1 to 10 192
i
ABSTRACT
A preliminary trial to determine the effect of repeated freezing and thawing on beef striploins, showed that the effect of thawing on frozen non-aged beef significantly improved tenderness, flavour and acceptability, indicating that thawing had a similar effect to aging. In the subsequent study, 207 beef striploins were collected from the Southern Crossbreeding Project (SXB: 70 heifers grainfed for 80 days, 70 steers grainfed for 180 days), Davies Gene Mapping Project (DGM: 30 steers grainfed for 180 days) and also 37 Hanwoo striploins imported into Australia from Korea. SXB animals consisted of Hereford cross calves sired by Belgian Blue, Limousin, South Devon, Hereford, Angus, Wagyu and Jersey bulls. DGM animals consisted of purebred Limousins and Jerseys and Limousin by Jersey crosses. Sensory analysis of beef striploins involved semi-trained taste panel assessments, using nine-point category scales for initial and sustained juiciness, beef flavour, beef fat flavour, oily flavour, buttery flavour, chicken-skin flavour, corn flavour, grassy flavour and overall acceptability. Flavour acceptability was positively enhanced by increased levels of intramuscular fat (IMF%). Significant differences in breed were apparent for juiciness, beef flavour, buttery flavour and flavour acceptability, after adjusting data to a constant level of intramuscular fat, suggesting that some variation in flavour may be genetic. The Korean Hanwoo displayed a numerically higher intensity of chicken score and lower intensity of beef flavour. Australian cattle breeds differed in fatty acid composition between each other and also to that of the Korean Hanwoo. The latter had 57% mono-unsaturated fatty acids, which was significantly higher (P<0.001) than the Australian breeds (47%). Since IMF% was confounded with breed, breed differences were not significant when adjusted for IMF%. Jersey animals most closely resembled the Hanwoo in fatty acid profile, whilst animals containing Limousin differed markedly from the Hanwoo. A chemical sensor was able to establish significant differences between Korean Hanwoo and Australian animals and predominantly mirrored differences in fatty acid composition and to some extent flavour. Development of prediction equations from individual fatty acids was disappointing (R2< 15%). However, when fatty acid data, IMF% and chemical sensor data were combined to form prediction equations, moderate R2 values were obtained of (24% to 43%).
ii
DECLARATION
I declare that this work contains no material which has been accepted for the award
of any other degree or diploma in any University or other tertiary institution, and that
to the best of my knowledge and belief, this thesis contains no material previously published
or written by another person, except where due reference has been made in the text.
I give consent to this copy of my thesis,
when deposited in the University Library,
being available for loan and photocopying.
Elke. M. Stephens
iii
ACKNOWLEDGEMENTS
I wish to thank a number of people for their assistance and support throughout the duration of my study on this project. Firstly, I must thank Dr. Brian Siebert for assistance with the fatty acid analysis in a preliminary trial and also for analysing all of the fatty acids from the Southern Crossbreeding Project, Davies Gene Mapping Project and the Korean Hanwoo animals. This data set was used in the majority of analyses throughout this thesis. Thanks must also go to Dr. Peter Speck for enabling me to use the Meat Laboratory at Rutherglen Research Institute, Ag. Victoria. Mr. Greg Ferrier, from Rutherglen Research Institute was invaluable in helping to measure tenderness, meat colour and pH in the preliminary trial. Thankyou to the management and staff at T&R abattoir, Murray Bridge, in particular Neal Teasdale, who allowed me to take samples from the boning room for my preliminary trial. A number of people gave up their time voluntarily to participate as tasters throughout the trial. I would like to thank the staff and student members from the Waite Institute in 1998, who were my tasters for the preliminary trial, in particular staff and students from the Department of Animal Science. A big thank-you must also go to the 25 Roseworthy students from the Meat Production class in 1999, who made up the taste panel for the main trial. Their enthusiasm and interest in what I was doing made it a pleasure to work with them. Whilst conducting the taste panels out at Roseworthy for the main trial, a number of people helped me throughout this time, including Ian Molloy (SARDI), Helen Daley (Wool CRC) and a number of the laboratory staff from the Roseworthy teaching wing. I would also like to thank the Australian Wine Research Institute at the Waite, for the use of the ‘chemical sensor’. In particular to Dr. Michelle Wirthensohn and Dr. Graham Jones for their time in helping me set up the machine for use. I must thank ELDERS Limited, for their financial support on this project. Mr. Dennis Wignall, Mr. Nick Chrichton and Mr. Tim Smith provided direction and support throughout the project and enabled beef samples to be collected in Korea and imported into Australia. Without this industry support, this project would not have been possible. In the latter stages of writing, I would like to thank Primary Industries and Resources staff Bruce Hancock, Dale Manson and Andrew Curtis, who were supportive of me finishing my Masters and allowing me time off from my PIRSA SA Lamb project to complete it. I would like to thank Dr. Wayne Pitchford for motivational assistance and moral support throughout the period of study. His assistance with Statistics was invaluable. I must also thank my parents, Peter Hocking and Michelle Fenton who were always there to provide moral support. Also, to Michelle, Megan, Veronica and Jane, thank you for your continual friendship and support.
2
r beef in Korea, during
e period 1997 to 2001, is expected to increase due to lower beef prices, resulting from
hen designing this project, this system was in the process of deregulation and the
port quotas were set to rise during the period up to the year 2001, after which time they
m Australia are treated
oorly and consequently the image of Australian beef in Korea is falling due to quality
1.1 INTRODUCTION
In South Korea, the demand for beef is greater than for any other meat, with the majority of
meat imports into the country consisting of beef. Consumer demand fo
th
liberalisation of the beef market (Meat Research Corporation, 1993).
The Livestock Products Marketing Organisation (LPMO) is a Korean government agency
responsible for the regulation of the quantity of beef imports into the country, via yearly
quotas. W
im
ceased.
Total beef imports into Korea rose steadily from 53,000 tonnes in 1989 to 180,000 tonnes in
1996. This figure was expected to reach 225,000 tonnes by the year 2000 (Leeds and
Lugsdin, 1997). Of these imports, 44% is made up of grain-fed cuts whilst the remainder is
supplied as grass fed quarter beef (AMLC, 1996). The United States supplies 80% of the
grain-fed beef imported into Korea, whilst Australia produces 71% of the grass fed beef
(AMLC, 1996). Unfortunately, frozen quarters of grass fed beef fro
p
problems and poor retail display (Meat Research Corporation, 1993).
For all but one year in the 1990’s, Korea was Australia’s third largest export beef market
(Lugsdin, 2000). In 1998, Australia provided around 37% of the total imports compared to
55% provided by the US. New Zealand, Canada, Ireland and England supply the remainder
3
the quota system
to be deregulated, it is important that Australian beef is well recognised by the major beef
.2 CONSUMER PERCEPTIONS OF AUSTRALIAN BEEF IN
e, less than 2% of responses (Baghurst, 1997). This is in contrast to the
eneral opinion that tenderness is the single most important factor determining acceptability
p to obesity. In relation to both the benefits and
(Lugsdin, 2000). In order for Australia to maintain market share, the quality of beef being
presented in Korea needs to improve dramatically. By the year 2001, when
is
organisations in Korea and by the consumers as a high quality tasty product.
1
KOREA
Price is the dominant factor influencing the retail purchase of imported beef in Korea,
however taste is the dominant reason for the purchase of, and preference for domestic
Hanwoo beef (CSIRO, 1995). A survey of Koreans, commissioned by the Australian Meat
Research Corporation, in Seoul, Pusan and Taegu highlighted several important factors in
determining the selection of beef. For the major organisations involved in the beef industry in
Korea, price (33%) and taste (29%) were of major importance, whilst the important factors
from the Korean customers’ point of view were taste (36%), price (28%) and safety/health
(9%), as shown in Figure 1.1. For both the organisations and customers, tenderness was only
of minor importanc
g
(Lawrie, 1985c).
From a survey conducted by Baghurst, 1997, which related the attitudes of Korean adults
from Seoul and Pusan to protein-based foods, it was shown that the major benefits of beef
were perceived to be its role as a source of protein and that it was tasty. The major problems
included expense, fat content and relationshi
4
(MRC Channel Research, CSIRO, 1995)
problems associated with beef and lamb, sensory issues were of major importance, indicating
the potential for research efforts in this area.
Figure 1.1 - Important Factors from Korean Organisations Perspective – For Customers
4.0%
0.6%
0.6%
0.6%
0.6%
1.7%
2.8%
27.7%
Undecided
No Response
Supplier Brand
Fat Colour
Tenderness
Meat Colour
Price
2.3%
2.8%
4.0%
4.5%
9.1%
35.7%
Meeting Specifications
Freshness
Country of Origin
Dish Indication
Safety / Health
Taste
0.6%
0.6%
0.6%
0.6%
0.6%
Yield
Convenience of cuts
Packaging
Grainfed
Leanness
Marbling
0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%
Korean consumers regard beef imported from Australia as low quality product when
compared to beef produced in Korea and the US. Their ultimate preference is for the Korean
ative breed, Hanwoo (Ryu et al., 1994; CSIRO, 1995). However, domestic production is
likely to remain fairly stable (Meat Research Corporation, 1993), with domestic
n
5
er than quantity. In 1996, Hanwoo
eef in Korea retailed for around $45 to $55 per kg (AUD), US beef $30 per kg (AUD),
and 3). The carcass yield grade is calculated from a prediction
equation based on eye muscle area at the 13th
Despite this, carcasses may be graded up or down depending on their visual musculature.
Quality grade is determined primarily by marbling score (Figure 1.3), however, meat colour,
fat colour, texture and maturity are also taken into consideration.
The Korean Livestock Marketing Corporation (KLMC) is a subsidiary of the National
Livestock Cooperatives Federation (NLCF) and in 1996 had 14 Hanwoo meat marketing
shops and 5 imported beef marketing shops across Korea (Canadian Embassy, Seoul, 1996).
The Hanwoo beef stores in Seoul were established to promote graded domestic beef in Korea
(Meat Research Corporation, 1993).
1.3 HANWOO CHARACTERISTICS
Surveys by the CSIRO in 1995 indicated that there are three tiers of perception of quality of
domestic beef, with the top level being “special cows fed on beer and carefully tended by their
owners”. The next two levels include cows bred for beef consumption and dairy cows.
Despite the grading classification given, most beef produced in Korea is regarded as
competitiveness likely to be due to increased quality rath
b
whilst Australian beef only cost $15 to $20 per kg (AUD), (Crichton, ELDERS Limited, pers.
comm, 1996) which gives an indication of its poor image.
In 1991, a domestic grading system was introduced into Korea. Carcasses are graded on yield
(A, B and C) and quality (1, 2
rib site, back fat thickness and carcass weight.
“Hanwoo” (CSIRO, 1995).
6
edlot Figure 1.2 - Korean Native Cattle (Hanwoo) in a Korean Fe
Figure 1.3 - Marbling Score
Korean cattle are graded by yield and quality, with yield ranging from A
to C (where B is the standard yield between 74.5 and 77%. A and C are
Korean native cattle (Hanwoo or bos taurus coreanae) are a hybrid of Bos taurus x Bos Zebu
(Rhee and Kim, 2001). Most Korean cattle are similar in conformation to a dairy beef cross
animal with a small frame and D muscling (Crichton, ELDERS Limited, pers comm, 1996).
Figure 1.2 shows Hanwoo cattle in a Korean feedlot. Top-grade Hanwoo cattle have
extremely white fat colour due to being fed predominantly on rice straw, rice meal and grain
rations (Crichton, ELDERS Limited, pers comm, 1996).
above and below the standard yield, respectively). Quality grade (1, 2
or 3) is determined predominantly by marbling, however if either meat
colour or fat colour is unsuitable (dark meat, yellow fat) the carcase will
be downgraded. In order to achieve quality grade 1, marbling score
(Figure 1.3) must be greater than 4 (AMLC, 1995).
7
orted beef. However, when Hanwoo grade 3 was compared to
boratory panel, a
vour/texture profile panel and a consumer panel. The Japanese cooking method (similar to
Ryu et al. (1994) compared the quality of imported beef (ungraded high quality chilled USA
striploins) with Korean native cattle beef of quality grades B1, B2, B3 in addition to out of
grade native cattle. Korean grade 1 beef had higher intramuscular fat content and was more
tender than other Korean grades and imported beef. Based on chemical, physical and sensory
analyses of samples, it was concluded that US beef quality was between Korean grades 2 and
3. The composition of imported beef contained higher proportions of unsaturated fatty acid,
although this did not cause any significant differences for aroma, flavour or juiciness. Korean
sensory panel results showed that 78% preferred Hanwoo grade 1 to 3, whilst 75% preferred
Hanwoo grade 1 to imp
imported beef, no difference was detected (Ryu et al., 1994).
There have been numerous studies that have reported palatability differences in beef from
different countries of origin, particularly in relation to the Japanese Wagyu. These studies
indicated that Japanese consumers perceived that the Japanese Wagyu breed had a
characteristic taste and tenderness (Boylston et al., 1995), and that Wagyu beef was superior
in palatability when compared to other beef produced in North America (Busboom et al.,
1993). In the latter study, striploins from different cattle breeds were cooked using both
North American and Japanese cooking methods and evaluated by a trained la
fla
Korean style cooking) involved slicing meat thinly and boiling it in water, whereas the North
American cooking method involved roasting and broiling (grilling) steaks. The results
indicated that Wagyu beef produced in Japan was superior in palatability when compared to
Angus, Longhorn and Choice grade American beef. It did show that the American bred
Wagyu, fed for extended periods also produced very palatable beef for the Japanese market
(Busboom et al., 1993).
8
composition, post slaughter
nd processing factors are likely to cause differences in flavour. There has been a vast
of post slaughter factors in relation to tenderness (Bouton
nd Harris, 1972c; Bouton et al., 1973a; Bouton et al., 1978b; Bouton et al., 1978c; Bouton et
e Korean market is shipped to Korea in the form of frozen
11.1% grain-fed quarters; AMLC, 1996). Once
and in flavour (CSIRO, 1995).
In t of tenderness, odour and fatty acid composition
were performed on samples of beef of both Korean and Australian origin. In addition,
subjective sensory analysis was undertaken. This involved semi-trained taste panel
1.4 TREATMENT OF AUSTRALIAN BEEF IN KOREA
In addition to palatability differences, due to breed and fatty acid
a
amount of research done in the area
a
al., 1980; Hood, 1980; Dutson et al., 1980; Aalhus et al., 1994; Daly, 1997; Dransfield, 1997;
Farouk and Swan, 1997a; Morton et al., 1997; O’Mahony et al., 1997), however this area
needs to be further quantified for the effect of processing factors on flavour.
The majority of beef destined for th
quarters of beef (49.7% grass-fed quarters;
carcasses arrive in Korea, they are packaged by the NLCF prior to retail. This process
involves defrosting of frozen quarters (often by submersion in tanks of hot water) followed by
boning out according to Korean specifications and finally refreezing before distribution to
supermarket chains (Meat Research Corporation, 1993; CSIRO, 1995). Additionally, the
lack of chilling facilities (tubs of ice) during retail causes primal cuts to partially thaw while
on display (Ford, 1997). As a result of this treatment, Australian beef has been described as
dark in colour, tough, and very bl
1.5 PROJECT AIMS
he present study, objective measurements
9
from
req ing system and which most closely resembled the fatty acid
assi
diff s both internationally and within the Australian market.
he specific aims of the study were as follows :
- review current literature on flavour and studies carried out on the effect of fatty acid
composition and marbling on flavour
- mimic the type of processes involved in the export of Australian beef to Korea, in order to
determine the effe traits and fatty acid
composition of beef striploins.
- characterise Korean beef (Hanwoo) in terms of fatty acid composition and meat quality
traits (in particular flavour)
- evaluate Australian cattle genotypes in terms of fatty acid composition and flavour, in
order to determine which breeds most closely resemble the Hanwoo in terms of flavour.
- determine the factors that influence flavour (genotype, genotype by diet interaction, level
of fatness).
- separate the effects of fat content and fatty acid composition on flavour.
- predict flavour using data obtained from both fatty acid and chemical sensor analysis.
assessments, in Australia, in an attempt to quantify the degree of difference in flavour of beef
each country.
This study aimed at identifying Australian cattle that were best suited to meet the specific
uirements of the Korean grad
and flavour profiles of native Korean Hanwoo beef. The outcomes from this project will
st, not only the development of a quality product for Korea, but more importantly, equip
Australia with the tools necessary to be able to design meat to suit the individual tastes of
erent market
1.5.1 Specific Aims
T
ct of repeated freezing and thawing on palatability
11
.1 MEAT QUALITY
odour of
e meat during cooking. Once the product is consumed, tenderness, juiciness and taste
g eating quality.
he odours liberated from cooking differ between the species of the animal, the feeding
2
Meat quality is highly variable due to the nature of muscle structure and function. Muscle
tissue consists of approximately 75% water, 20% protein and highly variable amounts of
soluble organic compounds. Beef acceptance is based on many factors including safety,
nutrition and eating quality. Eating quality is determined by various sensory experiences.
Initially the colour and visual appearance of the meat is important, followed by the
th
(flavour and aroma) attributes become important in determinin
Tenderness is normally considered the most important determinant of meat quality (Lawrie,
1985c). A study by Dransfield et al. (1984), showed that when meat was tough, it dominated
the overall judgement, however, when tender, flavour played an important role. With large-
scale research efforts into tenderness currently dominating meat quality research in Australia,
(CRC for the Cattle and Beef Industry – Meat Quality, 2000) flavour is likely to become
increasingly important as tenderness becomes more uniform.
T
regime, the slaughter condition, the way in which the meat is stored and processed, and the
method of cooking (Mottram, 1991). Young and Braggins (1996) reported that significant
effects on sheep odour are also produced by the animal’s growth rate, age, diet and possibly
its sex. It is also well known that flavour and aroma intensities, indicative of grassy
undesirable characteristic, increase with increasing maturity (eg Berry et al., 1980).
12
PTION
rtant attributes
ion to the four basic taste sensations (sweet, sour,
lty and bitter), feeling factors such as astringency, bite, burning, cooling, numbing and
in t
2.2
Perception of the primary taste sensations (bitter, salty, sweet and sour) occurs via taste
are
how osa. In humans,
ste papillae comprise of fungiform, foliate and vallate papillae. Fungiform papillae cover
2.2 ODOUR & FLAVOUR PERCE
Caul (1957) defined flavour as “the sensations perceived by the tongue, mouth, throat and
nose when an object is eaten”, and more recently Meilgaard et al. (1991), described flavour as
“the impressions perceived via the chemical senses from a product in the mouth”. Aroma has
been described as the “sensations perceived by the nose when an object is sniffed” Caul
(1957). In Korea, the word “mat” is used to define the sensory properties of foods and
encompasses flavour, aroma and texture, which are considered the most impo
when judging foods (CSIRO, 1995).
The way in which humans perceive flavour is quite complex. It involves the detection of
volatiles in the nasal passage via the olfactory system and therefore includes the aromatics,
tastes and chemical feeling factors. In addit
sa
coating play a role in the overall flavour experience, as do odours and feeling factors detected
he nasal passage and after-tastes and after-feels that occur (Meilgaard et al., 1991).
.1 Physiology of Taste (Gustation)
receptor cells that are located within specialised structures known as taste buds. Taste buds
predominantly found on papillae embedded in the epithelial surface of the tongue,
ever they are also spread over the oral, oesophageal and tracheal muc
ta
the entire surface of the tongue, vallate papillae are located across the back of the tongue and
foliate papillae are found along the back edges of the tongue (Altner, 1978a).
13
The chorda tympani (cranial nerve 7) innervates the anterior two thirds of the tongue, via
fibres from the fungiform papillae, and fires in the presence of the basic taste sensations.
l nerve 9) innervates the posterior third of the tongue
a). The concentration and the
ration of the taste stimulus will affect the intensity of the taste sensation, and during long
e
stimulus (Altner, 1978a). Additionally, and secretion (saliva) acts to dilute the
stimulus substance at the level of the taste buds and can alter the intensity of the sensation.
The pattern of excitation typical to a single nerve fibre in response to a range of substances is
called a “taste profile” (Altner, 1978a).
Within taste buds, the three main types of cells are sensory cells, supporting cells and basal
cells. The receptor cells, having no neural axons of their own, synapse on the terminals of
sensory afferent fibres in order to initiate a sensory response after binding with specific taste
molecules (Altner, 1978a). Approximately 50 nerve fibres enter and branch within each taste
bud. They are comprised of the following:
•
• The glosopharyngeal nerve (crania
via the vallate and foliate papillae and responds to taste, thermal, touch and pain
sensations.
• The trigeminal nerve (which doesn’t perceive taste but responds to thermal, touch and
pain sensations); and the superior laryngeal nerve which innervates the epiglottis and
oesophagus and is involved in taste, thermal, touch and pain sensations.
The specificity of the receptor molecules in the membranes of sensory cells allows
discrimination between different taste substances (Altner, 1978
du
xposure to a stimulus, the intensity of a sensation normally decreases due to adaptation to the
serous gl
14
outh to the nasal cavity (see Figure 2.1). The
asal cavity, which is divided into two spaces by the nasal septum, is lined with mucous
Figure 2.1 - Diagrammatic representation of a section through human nasal and buccal cavities
2.2.2 Physiology of Olfaction
The olfactory system is involved in both the sense of smell and the sense of taste. Human
perception of flavour involves odourant molecules being detected by the olfactory system via
the rectronasal passage, which connects the m
n
membrane. The olfactory mucosa in humans is located in the dorsoposterior region of the
nasal cavity on the roof, septum and superior turbinates and is continuous with the respiratory
mucosa. The respiratory region lacks olfactory cells and consists of ciliated epithelium,
which contain mucous producing goblet cells (Altner, 1978b).
The olfactory mucosa contains olfactory (receptor) cells, supporting cells and basal cells. The
supporting cells, which occupy most of the epithelium biomass, are thought to be involved in
secretory, supportive, nutritive or electrically insulative properties essential for the
functioning of receptor cells. In vertebrates, the olfactory receptor cells respond to a broad
15
on the olfactory
pithelium on which various protein molecules exist (Meilgaard et al., 1991). Odourous
t on the
references of likes and dislikes accumulated through experience, but are often the result of
sceptible
range of odourants. Odour molecules must diffuse through a layer of mucous, which covers
the olfactory epithelium, before they reach the membranes of the cilia, located on the
olfactory receptor cells (Altner, 1978b). The cilia of the receptor cells is thought to be the
main site of odorant-receptor cell interaction.
The most popular theory of odour detection is that there are certain regions
e
molecules are thought to have different conformations and attach to particular sites. The brain
then receives a complex pattern of signals and must translate these signals in order to
recognise the odour (Meilgaard et al., 1991).
2.2.3 Sensory Physiology
Bartoshuk (1980) stated that variations in human taste are not solely dependen
p
complex chemical and genetic influences. In humans, it has been shown that genetic variation
exists for taste perception. Additionally humans vary in their ability to detect and recognise
certain taste substances. The detection threshold for a particular substance can be defined as
“the lowest concentration at which a taste can just be detected”, whilst the recognition
threshold is “the lowest concentration at which the quality of a taste stimulus can be
recognised” (Bartoshuk, 1978) and thus perceived.
Within the basic taste sensations, psychophysicists have found more human variation in sweet
and bitter sensations than in sour or salty taste sensations (Bartoshuk, 1980). This can be
attributed to salty and sour substances having quite simple chemical structures, whereas sweet
and bitter substances are usually large and complex organic molecules which are su
16
nother factor which causes differences in taste perception between individuals is the
ropylthiouracil (PROP) as intensely bitter and have more papillae and taste buds on
e anterior tongue than nontasters (Dabrila et al., 1995). The authors hypothesised that
influence taste perception and dietary intake of fat, with
igher PROP tasting (supertasters) associated with lower fat intake. This theory is supported
to change (Bartoshuk, 1980). Additionally specific substances have been known to alter taste
perception of subsequent substances. One such example is the detergents used in some
toothpastes (sodium laurel sulfate) which makes sugar in orange juice taste less sweet, while
the acid in the juice tastes bitter and sour. The mechanism by which this occurs is that the
taste-receptor membrane contains phospholipids and the detergent is thought to affect these in
much the same way as the detergent cuts through grease (Bartoshuk, 1980).
A
composition of saliva, due to the sodium content which varies not only between individuals,
but also within the same individual due to exercise, dehydration, disease and chewing
(Bartoshuk, 1980). Since saliva contains water, amino acids, proteins, sugars, organic acids
and salts, humans cannot perceive the absolute concentration of a taste substance, however,
the difference in the concentration of substances can be detected (Meilgaard, 1991).
Dabrila et al. (1995) proposed the notion of “supertasters” and “nontasters”. Supertasters
taste 6-n-p
th
differences in tongue sensitivity may
h
by Fisher et al. (1961) who indicated tasters had more food dislikes than non-tasters.
Another bitter substance, phenylthiocarbamide (PTC) has been used in family studies to
demonstrate that the inability to taste PTC is a simple Mendelian recessive trait. Individuals
who carry two recessive genes are insensitive to PTC (nontasters), while those who carry one
or two dominant genes are quite sensitive and are called tasters (Bartoshuk, 1980).
17
sians (6-10%) than Caucasians (26-30%), indicating that
enetic as well as environmental factors determine cross cultural differences in flavour
es on
eef samples, from the same animals, varied inconsistently between sensory research
laborator r small,
but important regional biases in sensory assessments from different regions. Prescott (1997)
stated that it is for this reason that an understanding of the differences between cultures in
sensory food preferences is considered of prime importance when tailoring foods to specific
markets.
2.3 BEEF FLAVO
“Meat flavour deve result a complex interaction of
recursors derived from both the lean and fat components of m producing a host of volatile
It has also been suggested by Drenowski and Rock (1995) that the frequency of the non-
tasting gene is lower among A
g
perception. Despite this, Prescott (1997) reported that there do not seem to be differences in
the way that different cultures discriminate taste, or in the way that they perceive flavours and
odours. If this is true, then descriptive panels in Australia would be quite acceptable for
quantifying flavour differences between Australian and Korean product.
However, there does seem to be evidence for the existence of cultural differences in food
preference and acceptance. Dransfield et al. (1984) reported that flavour intensity scor
b
ies in different countries. The authors indicated that there was evidence fo
UR
lops during cooking and is the of
p eat
compounds that contribute to meat flavour” (Mottram, 1991). There have been numerous
reviews on meat flavour which have examined some of the reactions that occur between
flavour precursors and also how different classes of volatiles may be formed (MacLeod, 1986;
Mottram, 1991; Farmer, 1992; MacLeod, 1994; Mottram, 1994). An extensive review on the
18
), reported that no single compound or class
f compounds is responsible for the flavour of meat, however flavour precursors in lean beef
w
the volatile fraction beef cooked in boiling wa hydrogen sulfide,
ammonia, ac iacetyl. Additio rmic, acetic,
propionic, butyric and isobutyric acids and dimethyl sulfide tablished.
Table sses Reported in Cooked Beef (Mottram, 1991) OUND BE
chemistry of meat flavours is given by Mottram (1991), in which the various classes of
compounds in meat are reported.
In the 1950’s and 1960’s research into meat flavour was dominated by the search for meat
flavour precursors. Hornstein and Crowe (1960
o
ere identified to be low molecular weight compounds. Yueh and Strong (1960) found that
from lean ter contained
eta d dldehyde, acetone, an nally, the presence of fo
was tentatively es
2.1 - Chemical ClaCOMP EF Hydrocarbons 193
Alcohols and Phenols 82 Aldehydes 65 Ketones 76
Carboxylic Acids 24 Esters 59
Lactones 38 Furones and Pyrans 47
Pyrroles and Pyridines 39 Pyrazines 51
Other Nitrogen Compounds 28 Oxazoles and Oxazolines 13
Non-heterocyclic Sulfur compounds 72 Thiophenes 35
Thiazoles and Thiazolines 29 Other Heterocyclic Sulfur Compounds 13
Miscellaneous Compounds 16 TOTAL 880
Mottram (1991) indicated that there have been approximately 880 volatile compounds of
different chemical classes reported in cooked beef (Table 2.1) with the most important taste
compounds comprising amino acids, peptides and hypoxanthine, lactic, inosinic,
orthophosphoric, succinic and other acids, sugars and sodium salts of glutamic and aspartic
acids. Despite the large num
have been reported to possess m
basic structure (MacLeod, 1986 & 1994). Figur
identified to date as possessing m
responsible for the produc
19
ber of volatiles identified in meat, only a small number of these
eaty aromas, however these compounds possess the same
e 2.2 gives an outline of the compounds
eaty aromas, and Table 2.2 indicates some of the reactions
tion of these volatiles.
meaty aromas of Figure 2.2 (MacLeod, 1986) REACTION COMPOUND
Thermal degradation of cysteine 43, 44, 46, 48, 51, 52 Thermal degradation of cystine 9, 43, 44, 46, 48, 49, 52, 59
Cysteine / pyruvaldehyde 52 Cysteine / butanedione 66
Cysteine / rhamnose 11,14,30, 46, 43, 48 Cysteine / methionine / 2-furaldehyde 31
Cysteine / methionine / 5-methyl-2-furaldehyde 30 Cysteine / glucose 32, 48, 68
Cysteine / cystine-ribose 48, 55 Cyclotene / H2S / NH3 8, 9, 45
Acetaldehyde / H2S 3, 46, 49 Acetaldehyde / H2S / NH3 68
Acetaldehyde / CH3SH / H2S 1 Propanol / CH3SH / H2S 2
Propanol / acetaldehyde / CH3SH / H2S 69, 70 3-Methylbutanal / H2S / NH3 47, 71
Acetaldehyde / pyruvaldehyde / H2S / NH3 53 Acetaldehyde / acetoin / NH3 67
Crotonal / H2S 44 Butanedione / H2S 6
Pentane-2,3-dione / H2S 7 Butanedione / acetaldehyde / NH3 67
Butanedione / acetaldehyde / NH3 / H2S 64, 66 Pentane-2,3-dione / acetaldehyde / NH3 / H2S 55
HMFone / H2S 13,19, 20, 22, 23, 33, 34, 35, 36, 37, 38, 41, 42 HDFone / H2S 14, 39
HDFone / cysteine 7, 39, 40, 43, 46, 48, 52 Thermal degradation of thiamin 7, 12, 14, 19, 21, 24, 54
Thiamin / CH2SH 15, 16, 25, 26
Table 2.2 - Some reactions generating the
23
During heating, interactions between non-volatile precursors in muscle tissue produce
numerous substances including volatiles which contribute to meat flavour (Shahidi, 1994a).
Primary reactions involved in meat flavour development include lipid oxidation and
degradation; thermal degradation and subsequent reactions of proteins, peptides, amino acids,
sugars and ribonucleotides; and thermal degradation of thiamine (McLeod, 1994). Secondary
reactions involve the products of these reactions in turn becoming reactants and combining in
various ways to produce various mixtures of volatiles (McLeod, 1994). In addition to
volatiles, amino acids, peptides and nucleotides may also contribute to the basic taste
sensations, thus complicating the overall taste experience further (Shahidi, 1994a).
The two most important chemical pathways in the formation of meat volatiles are:
1) Strecker degradation of α-amino acids and the subsequent formation of alkylpyrazines
2) Maillard reactions
Maillard reactions occur between compounds containing free amino groups and carbonyl
compounds and are responsible for the production of numerous meat volatiles (McLeod,
1994).
One of the most important reactions involved in meat flavour development is the Maillard
reaction between amino compounds and reducing sugars, which occurs more readily under
conditions of dry heat. Primary reactions that occur on heating include pyrolysis of amino
acids and peptides, carbohydrate degradation, interaction of sugars with amino acids or
peptides, degradation of ribonucleotides and thiamin degradation of lipid. In addition,
volatile compounds, which contribute to meat flavour are also produced by secondary
reactions that occur between products of the initial reactions (Mottram, 1991). Consequently,
24
application of different cooking methods will lead to different reactions taking place, which
ultimately liberate different mixes of volatiles contributing to subtle differences in flavour.
Mottram (1991) reports that hydrocarbons make up the largest class of volatiles in cooked
meat, despite the fact that their contribution to flavour is minimal. Pyrroles, purrolines and
pyrrolidines occur in the volatiles of many heated foods, and arise from the pyrolysis of
amino acids such as proline. Pyrroles may be important in defining the aroma of roast beef
(Mottram, 1991). Thiazoles and thiazolines also contribute to both the meaty and roast
characteristics. 2,4-dimethyl-5-ethylthiazole has a nutty, roast, meaty, liver-like flavour and
is found in roast beef, grilled pork and fried chicken (Mottram, 1991).
The benzopyrroles, indole and 3-methylindole (skatole) are responsible for the undesirable
faecal odours sometimes detected in meat – particularly in sheep. This odour possibly arises
from the animals grazing on certain plants which inhibit the animal from excreting indole and
skatole, derived from the metabolism of tryptophan, thus accumulating in the meat causing a
faecal taint (Mottram, 1991).
Free fatty acids in meat are derived from triglycerides and phospholipids by enzyme
hydrolysis or by thermally induced hydrolysis or oxidation during cooking. Sofos and
Raharjo (1989) indicated that rancidity and warmed over flavours in restructured meat
products arises due to disruption of tissue membranes by heating or mechanical treatment and
subsequent release of phospholipids which are prone to oxidation and thus rancidity.
The contribution of lipids to meat flavour is quite a contentious issue and continues to be
studied. Early work on meat flavour suggested that it was the fat component of meat which
25
gives rise to the characteristic odour of meats from different species (Hornstein and Crowe,
1960; Batzer, 1960). When lean portions of ground beef and pork were heated, they gave
identical meaty odours, while fat from each species (when heated) gave the characteristic beef
and pork aromas (Hornstein and Crowe, 1960). It was also noticed that odours obtained from
heating the fat fractions were more varied than those produced by the lean meat fractions,
indicating that the different flavours of meat of the same species were likely to be due to the
nature of the fat (Hornstein and Crowe, 1960). Additionally, Batzer et al. (1960) isolated
substances from raw beef muscle which, when boiled in water, gave a characteristic beef
broth odour and flavour; and when heated with fat, gave an odour similar to broiled steak.
Evidence for the role of a particular compound in fat, thought to be responsible for producing
species-specific aromas on heating, was reported for lamb and pork but not beef (Wasserman
and Talley, 1968). Pork is characterised by a high content of unsaturated -enoic, -dienoic and
-trienoic acids. Sheepmeat aromas contain large quantities of methyl-branched fatty acids
such as 4 methyloctanoic, 4 methylnonanoic and tetradecanoic acids, which produce
characteristic mutton odours (Mottram, 1991; Young and Braggins, 1996). Reid et al. (1993)
reported that 4-methylocatonic and 4-ethyloctanoic acids were the most significant
contributors to sheep meat odour. It has been reported that lipid is considered less important
for beef flavour (Wasserman and Talley, 1968; MacLeod, 1986), despite the fact that beef
contains esters of long chain fatty acids, which give it a fatty character (Mottram, 1991).
Mottram and Edwards (1983) showed that the selective removal of triglycerides from beef
caused no significant difference in cooked aroma, whereas the removal of triglycerides and
phospholipids generated marked sensory differences. It was concluded from this study that it
is the phospholipids (essential membrane lipid) that are important for the development of
26
desirable flavours in beef during cooking, whilst the triglycerides are not essential. However,
recent research has indicated that the high levels of monounsaturated triglycerides are
associated with more desirable flavours in beef, particularly in relation to the Japanese Wagyu
breed (May et al., 1993). This will be covered in the next section of this review.
In lean muscle tissue, intramuscular lipids (triglycerides and phospholipids) are a source of
many volatile components and they dominate gas chromatograms of cooked meat aromas.
Despite this, the contribution of these compounds to the flavour may be quite small due to
having a high odour threshold value. Mottram (1994) indicated that only those compounds
with low odour threshold values are likely to contribute to meat flavour, such as aldehydes,
unsaturated alcohols and unsaturated ketones. It was also reported that saturated and
unsaturated aldehydes with 6-10 carbons are major volatile components of all cooked meats
and therefore play an important role in meat aroma (Motrram, 1994).
Volatiles may be arranged in order of their flavour significance according to their flavour
dilution factor which is proportional to its’ aroma value. The aroma value is described as “the
ratio of concentration of flavour compound to its odour threshold” (MacCleod, 1994). Table
2.3 represents the compounds of cooked beef aromas which possess relatively high flavour
dilution factors. Only 2-methylfuran-3-thiol and bis(2-methyl-3-furyl)disulphide compounds
were described as meaty. However, due to their extremely high flavour dilution factor (Table
2.3), their odour potency is quite strong.
Gasser and Grosch (1990) also reported a similar study on chicken and reported that the major
differences between beef and chicken aromas were that sulphur compounds predominated in
beef, whilst volatiles from oxidation of unsaturated lipids prevailed in chicken.
27
Table 2.3 - Compounds of Cooked Beef Aroma Possessing Relatively High Flavour Dilution Factors (MacCleod, 1994 - adapted from Gasser and Grosch, 1988)
FLAVOUR DILUTION FACTOR
AROMA COMPONENT ODOUR QUALITY
512 2-Methylfuran-3-thiol Meaty, sweet, sufphurous Unknown Roasted Methional Cooked Potato Non-2(E)-enal Tallowy, fatty Deca-2(E),4(E)-dienal Fatty, fried potato β-Ionone Violets Bis(2-methyl-3-furyl) disulphide Meaty256 2-Acetyl-1-pyrroline Roasted, sweet Oct-1-en-3-one Mushroom Phenylacetaldehyde Honey, sweet 128 2-Acetylthiazole Roasted Nona-2(E),4(E)-dienal Fatty64 Octan-2-one Fruity, musty Oct-2(E)-enal Fruity, fatty, tallowy Decan-2-one Musty, fruity Unknown Sulphurous, onion Dodecan-2-one Musty, fruity 32 Hept-2(E)-enal Fatty, tallowy Octa-1,5(Z)-dien-3-one Geranium, metallic Unknown Musty, fatty Benzothiazole Pyridine, metallic 16 Hexanal Green Hex-2(E)-enal Green Heptan-2-one Fruity, musty Heptanal Green, fatty, oily Dimethyl trisulphide Cabbage, sulphurous Benzythiol Sulphurous Nona-2(E),6(Z)-dienal Cucumber Undecan-2-one Tallowy, fruity Tridecan-2-one Rancid, fruity, tallowy 8 Oct-1-en-3-ol Mushroom Nonan-2-one Fruity, musty Nonanal Tallowy, green Unknown Tallowy, cardboard 5-Methylthiophene-2-carboxaldehyde Mouldy, sulphurous Unknown Sulphurous 3-Acetyl-2,5-dimethylthiophene Sulphurous A deca-2,4-dienal (not E,E) Fatty4 2-Methyl-3-(methylthio)furan Sulphurous 2-Acetylthiophene Sulphurous, sweet
28
2.4 DESCRIPTION OF MEAT FLAVOUR
Figure 2.3 shows the “Flavour Wheel” which is a pictorial illustration of some of the basic
flavour relations used in the food industry – particularly in the development of synthetic food
flavourings. The meaty, animalic flavour note is one of the most complex to be described,
since roast beef flavour differs, for example from that of barbequed or simply boiled meat
(Kuentzel and Bahri, 1991). Although there are chemical examples of meaty animalic flavour
notes, a perfect reconstitution of meat flavour with all its’ nuances seems to be practically
impossible (Kuentzel and Bahri, 1991).
There have been a large number of descriptive terms developed by flavour profile panels to
describe the various flavours and odours of meat. Miller et al., (1996) described beef flavours
and aromas in terms of the aromatics (cooked beef/brothy, serum/bloody, cowy/grainy,
cardboard, painty, fishy and liver/organy); mouth-feels (metallic and astringent); and the basic
tastes - sweet, sour, bitter and salty. Descriptive terms for desirable meat characteristics have
included sweet and browned, while grassy and astringent terms were associated with
undesirable flavours (Berry et al., 1980). Table 2.4 documents other frequently used terms in
the description of meat flavour.
30
Table 2.4 - Common descriptors of meat flavour characteristics
ESIRABLE UNDESIRABLE D
Berry et al., 1980 (Beef) - Sweet, browned gent
- Grassy, astrin
Berry et al., 1980 (Beef) - Brown, caremelised, bouillion-like, beefy Musty – nutty, mouldy like
stringent mouthfeel rassy – animal, chemical, medicinal
- AGMetallic – serum, blood salts, liver-like
Melton, 1982a (Beef) - Beef fat flavour (flavour of freshly ooked beef fat)
- Intense dairy – milky flavour, soured or other off flavours, higher intensity of dairy flavour and aftertaste and presence of
c
soured dairy flavour, putrid flavours and dirty socks aromas
ers et al. 1987 (Beef)
Mouth-filling blend, brown flavours
- Bloody-serumy, metallic, sour
Bow-
Camfield et al.,1997 (Beef) - Cooked beef/broth, cooked beef fat, cowy/grainy
- Serum/bloody, Cardboard, Liver, soured/grainy
Young & Scales, 1993 (Sheep) - Fat odour – clean, mild, sweet, vanilla, burnt, roasted, caramel, spicy, chocolate - Lean flavour/odour – Beefy, big flavour, fatty
- Fat odour - fishy, stale, rancid, sweaty, sheepy, medicinal,rubbery, sulphury, tallow, musty - Lean flavour/odour – metallic, bland, bitter, foreign, sheepy, kidney/offal, grassy, bloody
Young & Braggins, 1996 (Sheep) - Strong, Beefy, Meaty and sweet flavours
- Bland, flat, low lamb flavour, stale, musty
Reid et al., 1993 (Sheep)
- Stale, cardboard, rancid
31
.5 ROLE OF LIPIDS, FATS AND FATTY ACIDS - INFLUENCE ON
a medium that regulates the
between water, fat and vapour phases (Gurr and Harwood,
ed-over
eat (Sofos and Raharjo, 1989). This is predominantly due to the susceptibility of
e flavour profile making the perceived flavour sensation less
2
PALATABILITY
Lipids influence flavour through their effect on flavour perception (mouthfeel, taste and
aroma), flavour stability and flavour generation (de Roos, 1997). Fat has been described as
being both a source of taste and aroma compounds and also
distribution of these compounds
1996). Flavour compounds such as free fatty acids, aldehydes, ketones, lactones and other
volatiles arise from the degradation of lipids by lipolysis, oxidation and thermal degradation.
The methylene group adjacent to a double bond, is a site for oxidative attack, so the more
unsaturated an acid, the greater its susceptibility to oxidation (Mottram, 1991). Lipid
oxidation of the fat component of meat is often responsible for rancidity and warm
flavours in m
muscle foods to oxidative changes being related to “the nature, proportion and degree of
unsaturation of fatty acids in their lipids” (Sofos and Raharjo, 1989).
Fatty acids with chain lengths greater than ten carbon atoms have little odour because they are
not sufficiently volatile compared to those with shorter chain lengths (Gurr and Harwood,
1996). Additionally, in low fat foods, aroma compounds are released early and their
disappearance occurs early in th
pleasurable (Gurr and Harwood, 1996; de Roos, 1997).
In the past there have been many studies on the influence of fat and marbling on palatability
of beef (Goll et al., 1965; Wasserman and Talley, 1968; Smith et al., 1983; Crouse et al.,
1989; Sofos and Raharjo, 1989; Berry and Leddy, 1990; Seirer et al., 1992; Wheeler et al.,
32
rior marbling ability.
) reported that steaks from carcasses with
at least 5mm of fat were superior to steaks fr this amount of fat,
how ity.
Berry et al., (1980) reported th s were also classed as having
more desirable flavours, characterised by sweet and browned flavours.
1994; Rymill et al., 1997). This is quite a contentious issue amongst scientists within the
meat industry, with varying views on the importance of marbling.
Despite the fact that many studies have shown a lack of evidence for the benefit of marbling
(Goll et al., 1965; Kregel et al., 1986; Crouse et al., 1989; Wheeler et al., 1994; Rymill et al.,
1997), the USDA grading system is based on marbling score and has been reported as being
highly correlated with flavour (Smith et al., 1983; Berry et al., 1980). Currently, with the
trend for high levels of marbling for beef destined for the Japanese market, Australian
producers are placing increasing selection pressure on animals with supe
Goll et al. (1965), reported that marbling had no effect on sensory scores for tenderness,
juiciness or flavour and that there was no relationship between intensity scores of juiciness
and flavour with marbling or tenderness measurements. Rymill et al. (1997) also reported
that intramuscular fat percentage did not produce any perceivable differences in tenderness
and juiciness of beef steaks. The authors from both of these studies indicated that other factors
such as carcass maturity (Goll et al., 1965) and degree of doneness (Rymill et al., 1997) had
more influence on the eating quality of beef than differences in marbling.
Contrasting studies have shown that fat levels were highly related to meat flavour (Beilken et
al., 1990; Dolezal et al., 1982). Dolezal et al. (1982
om carcasses with less than
ever steaks with greater than 7mm of fat did not further improve cooked beef palatabil
at higher mean marbling score
33
It is well documented that the type of fat (f omposition) contributes significantly to
meat flavour differences between animals. ws the names and numeric symbols
of the common fatty in bovine scle tissues. Table 2.6, Table 2.7
and Table 2.8 show rrelations b individual fatty acids flavour from four
independent studies. o been rep increased fatness is associated with a
decrease in the level on (Perry e 8). Recent studies have indicated that
decreased proportion ed fatty acids (and hence increased levels of unsaturated fatty
acids – in particular urated fatty ot only beneficial from a nutritional
ciated with more desirable flavour characteristics (Schroeder et
, 1980; Busboom et al., 1993; May et al., 1993)
Numeric Symbo atty Acid
atty acid c
Table 2.5 sho
acids found adipose and mu
various co etween
It has als orted that
of saturati t al., 199
s of saturat
monounsat acids) are n
point of view, but are also asso
al.
Table 2.5 - Names and Numeric Symbols of some common Fatty Acids in Bovine Adipose and Muscle Tissues
l F
14:0 Myristic
14:1(n-5) Myristoleic
16:0 Palmitic
16:1(n-7) Palmitoleic
18:0 Stearic
18:1(n-7t) trans-Vaccenic
18:1(n-9t) Elaidic
18:1(n-7c) Vaccenic
18:1(n-9c) Oleic
18:2(n-6) Lin lo eic
18:3(n-3) α-Linolenic
18:3(n-6) Linγ- olenic
34
Table 2.6 - Correlations between long-chain fatty acids with sensory characteristics of m. longissimus dorsi (LD) of crossbred feedlot steers (n=108) (Camfield et al., 1997).
Fatty Acid Beefy Cowy Cardboard Painty 14:0 -0.35*
14:1 (n-5) -0.30* -0.32* 0.30* 16:0
16:1 (n-7) 1 28* 8:0 0.28* -0.
18:1 (n-9) .33* -0 -0.30* 18:2 (n-6) * ** 0.38* 0.41 18:3 (n-3) -0.31*
Table 2.7 - Correlations between fatty acids and flavours (n=95) (Melton et al., 1982a) Fatty Acid Cooked Beef
Fat Flavour Flavour F
FlaMilky Oily
Flavour r
our Liver ishy
vour Sou
Flav14:1 0.27* * -0.39*-0.36* * 15:0 -0.38** 0.36** 0.52*** 0.36** 16:1 0.41** 0.32* -0.32* -0.41** -0.44*** 17:1 0.26* -0.37** -0.41** -0.36** 18:0 -0.51*** -0.33*** 0.57*** 0.64*** 0.53** 18:1 0.30* -0.38** -0.46*** -0.33* 18:3 -0.39** 0.41** 0.53*** 0.57*** 19:1 0 0.26* 20:4 -0.34** 0.30* 0.43*** 0.36** 20:1 -0.30* 0.27*
Carbohydrate 0.40** -0.32*
Table 2.8 - Correlations between individual fatty acids and flavour score in neutral and polar lipid fractions(n=60) (Melton et al., 1982b) and in SC & IM fat (LD) samples
(n=54) (Westerling & Hedrick, 1979). Melton et al. (1982b) Westerling and Hedrick (1979)
Fatty Acid Neutral Polar Lipid Subcutaneous fat Intramuscular fat Lipid (SC) (IM)
14:1 -0.33** 16:0 -0.52** -0.52** 18:0 -0.30** -0.24** -0.56** -0.60** 18:1 0.29** 0.69** 0.67** 18:2 -0.63** 18:3 -0.51*** - 0.41*** 20:1 -0.43** SFA -0.65** -0.66** UFA 0.65** 0.66**
Carbohydrate 0.36***
35
f these
cids were associated with more desirable flavour (Melton et al., 1982). These findings were
leic acid) and flavour acceptability (P<0.05).
fatt <0.05) to beefy flavour (r=0.28), associated with
sirable acceptability (Table 2.6). High levels of this acid are normally associated with less
he duration of
problems with understanding the role of lipids on flavour has been that much of
e research work has not differentiated clearly between subcutaneous fats and intramuscular
fats (marbling). Intramuscular fat contains phospholipid, which has a higher content of
It is generally agreed that undesirable flavours in beef are associated with samples containing
very high percentages of linolenic acid (18:3) relative to zero, and low percentages of oleic
(18:1) acids (Melton et al., 1982; Larick and Turner, 1989). Melton et al. (1982) showed that
higher concentrations of 18:1, in the neutral lipid, and water soluble carbohydrates were
positively correlated with flavour score and thus had a more desirable flavour. Negative
correlations were found between flavour score and the percentages of 14:1, 18:0, and 18:3 of
the neutral lipids and with 18:0 and 18:3 of the polar lipid. Lower concentrations o
a
also in agreement with Dryden and Marchello (1970) who reported that high concentrations of
14:1, 16:0, 18:0, 18:2 (myristoleic, palmitic, stearic and linoleic acids respectively) were
scored as less desirable, whilst higher concentrations of 18:1 (oleic acid) were associated with
improved sensory scores. Dryden and Marchello (1970) reported a correlation of 0.66
between 18:1 (o
The findings of Camfield et al. (1997) are unusual, in that stearic acid (18:0) was the only
y acid significantly correlated (P
de
desirable flavours. In agreement with other studies, is that by increasing t
concentrate feeding, flavour intensity and overall palatability improved. Additionally, oleic
acid concentration increased in animals significantly (P<0.05) due to increasing time of
animals on the concentrate diet, suggesting a positive effect of higher 18:1 on flavour.
One of the
th
36
unsaturated fatty acids, compared to subcutaneous fat (Moody, 1983). Fats influence flavour
in two ways :
1) oxidation of the unsaturated fatty acids to produce carbonyl compounds and ‘off-flavours’
2) subcutaneous fat acts as a depot for fat-soluble compounds which volatilise upon heating
to release flavours (Moody, 1983).
2.5.1 Genetic Differences in Flavour – Fatty Acid Composition
It has been shown that fatty acid composition varies between different breeds (Sturdivant et
al., 1992; May et al., 1993; Boylston et al., 1995; Siebert et al., 1996; Zembayashi and
ishimura, 1996; Yang et al., 1999) and that this has an effect on palatability factors of beef,
choice
eaks (P<0.05).
N
particularly in the Japanese Wagyu breed (Xie et al., 1996b; Sturdivant et al., 1992; May et
al., 1993).
The superior palatability of the Japanese Wagyu when compared to Angus, Longhorn and US
choice beef in America has been reported by Busboom et al. (1993). They demonstrated that
highly palatable beef can be produced in America if the Wagyu breed is utilised (American
crossbred Wagyu) in a controlled extended feeding regime. Trained flavour profile panel
evaluation indicated that both Japanese Wagyu and American Wagyu beef produced a more
intense appropriate fatty aromatic (P<0.05) and a sweeter taste than Angus, Longhorn and US
choice beef. A consumer panel rated steaks from Japanese Wagyu and Angus Wagyu as
being more tender, juicy, desirable in flavour and more palatable than Angus and US
st
Xie et al. (1996b) evaluated the differences in carcase traits and fatty acid composition in
muscle and adipose tissue from Wagyu crossbred steers and Angus steers in order to
37
determine the effect of breed and specific sire on those traits. It was reported that Wagyu
cross steers had increased marbling, maturity and USDA quality scores, larger eye muscle
area and lower fat thickness (P<0.05) than Angus steers. Additionally Wagyu cross steers had
creased levels of 14:0, 14:1, 16:0, 16:1 and decreased levels of 18:0 and 18:1 than Angus
resian) in a recent study by Perry et al. (1998). The
saturation
vels (and thus flavour desirability).
in
sired steers (P<0.05). There were no significant differences reported for percentage of total
monounsaturated fatty acid (MUFA) between the breeds. Angus steers had a higher fat
thickness than Wagyu steers, which was reported to have contributed to the elevated level of
18:1 (oleic acid) in Angus animals. Fat thickness was positively correlated with 18:1 and
negatively correlated with total saturated fatty acids (r=0.25 and –0.24 respectively).
Fatty acid composition and melting point of fat were reported to be affected by sire breed
(Hereford, Brahman, Simmental, and F
authors suggested that breed differences in fatty acid could result in differences in the
proportions of fatty acids derived from particular rumen bacteria. It is known that ruminants,
unlike monogastrics, hydrogenate dietary unsaturated fatty acids in their fore-gut, so that
mainly saturated fatty acids are available for absorption from the gut (Tume, 1995).
It has been noted that bulls whose progeny excel in carcass characteristics, are generally not
the same as those with more desirable fatty acid compositions (Xie et al., 1996b) which could
make selection of so-called superior animals difficult. Additionally, steers whose growth
rates were restricted to 0.75kg per day, produced beef with better flavour than those fattened
on a higher plane of nutrition (Zembayashi and Nishimura, 1996). This can be attributed to
the fact that slower growth rate animals will be at a later stage of maturity and hence
increased level of fatness at slaughter which is associated with increases in monoun
le
38
ironmental influences could not be ruled
ut. The level of MUFA’s achieved in Wagyu crosses (1/2 and 7/8) produced in America
w n
a
coA desaturase within adipose tissue. Stearoyl-CoA desaturase catalyses the conversion of all
saturated fatty acids to n-9 MUFA’s and thus a single enzyme could be responsible for the
levated MUFA’s observed in the Wagyu adipose tissue (Sturdivant et al., 1992).
UFA % differences between Japanese Wagyu (produced in Japan) and American Wagyu and Angus beef (fed for 524 days in America). Boylston et al. (1995).
Japanese Wagyu American Wagyu American Angus
In addition to the genetic differences reported for fatty acid composition of cattle, there also
seems to be a difference in fatty acid profiles of animals produced under different
environmental locations. For example, there are no studies that have achieved the same levels
of monounsaturated fatty acids as those produced in Japan. A study by Sturdivant et al.
(1992) showed that purebred Wagyu cattle from Japan displayed extreme proportions of
MUFA’s (68%) in their adipose tissues for which env
o
as 58%. The authors of this study put forward the theory that the elevated MUFA i
dipose tissue of Wagyu cattle could have been the result of an elevated activity of stearoyl-
e
Boylston et al. (1995) also compared Japanese Wagyu produced in Japan, with 75-83%
purebred American Wagyu, Angus, Longhorn and US Choice beef fed for 524 days in
America. Japanese Wagyu had increased levels of unsaturated fatty acids when compared to
other breeds with the exception of the Angus Wagyu. Japanese Wagyu was lower in 16:1 and
higher in 18:1 than beef from other sources. The MUFA% within the intramuscular lipids of
beef from different sources is reported in Table 2.9.
Table 2.9 - M
Triacylglycerol 64 58 55
Phospholipid 73 61 74
39
ility characteristics, with the main differences in flavour profile being
fatty flavour in steaks from grain fed steers; and a grassy flavour in steaks from forage fed
have been
ssociated with higher levels of PUFA’s such as linoleic and linolenic acids (see Table 2.7).
flavours in
ass fed beef and may have also led to a more rapid development of oxidative rancidity.
Larick and Turner (1989) also suggested that the source of increased 18:2 and 18:3 in the
2.5.2 Nutrition Effects on Flavour – Fatty Acid Composition
There have been numerous studies that have reported the effect of diet on beef flavour
(Schroeder et al., 1980; Chastain et al., 1981; Melton et al., 1982a,b; Larick and Turner 1989;
Miller et al., 1996; Camfield et al., 1997). Grain finishing of beef cattle is reported to
improve overall palatab
a
steers (Schroeder et al., 1980). This report is in agreement with Melton et al. (1982a&b) who
suggested that descriptions of intense beef fat flavour were normally associated with highly
desirable beef.
Other studies have reported distinctive undesirable character notes within the flavour profile
of grass fed beef (Melton et al., 1982a; Larick and Turner, 1989). Grass fed beef has been
described as having a more intense dairy or milky flavour in addition to the presence of
soured and off flavours (Melton et al., 1982a). These undesirable flavours
a
Melton et al. (1982a), showed that the undesirable flavours (milky-oily, sour and fishy)
decreased intensity with corn feeding for 140 days compared to meat from animals fed for
lesser periods of time.
Larick and Turner (1989) showed that grain and grass feeding produced different fatty acid
profiles in the lean tissue of beef cattle. It was suggested that there were higher (P<0.05)
concentrations of polyunsaturated acids (PUFA) in the polar lipid fraction of grass fed
animals, and that these may have contributed to the higher intensity of undesirable
gr
40
with improvements in palatability and
duction in shear force (Harrison et al., 1978; Westerling and Hedrick, 1979; Schroeder et
addition to differences in flavour due to fatty acid composition, Melton et al. (1982b)
were considered to be of acceptable desirability.
tissues from animals fed on forage may be due to fatty acids which escaped hydrogenation in
the rumen prior to absorption and are instead, deposited in the adipose tissue.
On the other hand, an increase in monounsaturated fatty acids (MUFA), due to longer periods
of time on grain, has been reported to be associated
re
al., 1980; Melton et al., 1982a). Xie et al. (1996) reported that crossbred Wagyu cattle fed a
high concentrate diet for 90 days produced acceptable carcasses and that after this period may
have already reached their genetic potential to deposit marbling. An additional 80 days on
feed didn’t improve USDA quality grade or palatability, however an increase in the level of
MUFA’s in the intramuscular fat was demonstrated with a longer period on feed.
In
indicated that grain fed animals with the most desirable flavour scores had a high free sugar
content, whereas undesirable grass fed animals had lower carbohydrate concentrations. Grass
fed beef had less desirable flavours due to the presence of soured dairy flavour and other off
flavours, coupled with a less intense beef fat flavour.
Cattle at pasture have tendencies to accumulate β-carotene in the fat causing an increase in
yellowness and an associated grassy smell (Seirer et al., 1992). In contrast, Chastain et al.
(1981), reported that sensory panel scores showed no significant differences for flavour,
tenderness and juiciness between grass and grain fed animals. It was concluded that panellists
were able to detect subtle differences between beef from grass and grain feeding regimes,
however all samples
41
here have been numerous studies in the area of flavour research on sheep meats. The
any
avour profile panels tend to be aimed at identifying the factors causing muttony or rancid
ponsible for the sheep-meat specific odour. Thus grain fed
mb is preferred for markets requiring a mild sheep flavour, whereas for markets desiring a
In addition to comparisons between grass and grain-fed beef, other studies have been carried
out to determine whether differences in flavours were produced when animals were fed on
different sources of grain. The Japanese believe that feeding cattle barley-based diets results
in beef with more desirable flavour, than when fed on corn-based diets (cited by Miller et al.,
1996). Miller et al. (1996), conducted a study to test this claim, and fed Angus cross-bred
steers for a period of 100 days on 3 different diets (corn, corn/barley or barley). The authors
concluded that grain source did not affect flavour of beef since the descriptive meat
palatability attributes did not differ in steers across the three diets, nor were there any
differences in total lipid or fatty acid composition. Additionally, the flavours associated with
corn, corn/barley, or barley grain sources were not detected by a highly trained descriptive
sensory taste panel.
T
flavour / odour of cooked sheep meat is relatively strong compared to beef and thus m
fl
flavours in lamb (Young and Braggins, 1996). The sensory evaluation of lamb, with
simultaneous chemical analysis of all the volatiles using an ‘electronic nose’ (pattern
recognition device) was carried out to determine those volatiles responsible for the flavour of
sheep meat (Young and Braggins, 1996). Overall, flavour and odour intensity was much
lower in corn/wheat fed animals compared to slow grown pasture fed animals slaughtered at
90 days. The pasture fed lambs had higher concentrations of certain branched chain fatty
acids, which were thought to be res
la
stronger sheepmeat odour, lambs fed on pasture would be suitable.
42
future demand for meat depends primarily upon its safety,
holesomeness and health properties, but that it should also be presented attractively to invite
orm and enjoyable experience to the consumer when it
eating quality of meat are the control of the
te of cooling of the carcass, together with the control of the rate of change of pH. These
factor also
has an impact on meat
2.6.1 ffect
Madr and M uses
distinct changes in sensory at
various pH level ed in concentration as pH decreased
ere furanthiols, mercaptoketones, aliphatic sufides, dithianones, some thiophenes, furans
2.6 PROCESSING AND COOKING INFLUENCES ON FLAVOUR
Lister (1996) describes how the
w
purchase and guarantee consistent, unif
is eaten. He describes how tenderness is the first casualty of the improved processing and fast
chilling of meat to improve meat safety, and now believes that the constant striving for
producing lean meat, may ultimately prove detrimental to the eating quality, since lipid is an
important source of meat flavours.
Chrystall and Daly (1996) reviewed processing factors influencing meat quality. The authors
suggest that the two main tools for tailoring the
ra
s are extremely important for control of tenderness, however the ultimate pH level
flavour.
E of pH on Flavour
uga ottram (1995) showed that adding acid to meat prior to cooking ca
tributes after heating. Table 2.10 shows odour descriptions at
s. The headspace volatiles which increas
w
and aldehydes. Those which decreased with decreasing pH included pyrazines, thiazoles and
thiophenes; whereas hydrocarbons, ketones, alcohols, bicyclic compounds and dithiolanones
were unchanged (Madruga and Mottram, 1995).
43
Table 2.10 - Sensory attributes and pH of meat cooked at different pH values
(adapted from Madruga and Mottram (1995).
PH Odour description
5.6 Boiled meat, burnt, weak meaty flavour, fatty.
5.0 Cooked beef, gravy, fatty, mince meat, oniony, juicy meat, hint of liver.
4.5 Cooked beef, fatty, oniony, musty, fresh roast meat, mince meat.
4.0 Meaty, fatty, sour, kidney, livery, stale, sweaty, musty, overcooked
meat, off liver pate.
Dry, firm and dark (DFD) meat arises due to glycogen depletion in the muscle occurring
before slaughter (due to stress) which leads to meat having a high ultimate pH. The eating
quality of DFD (high pH) meat is considered to be less flavoursome and be less acceptable
than meat of normal pH (Dransfield, 1980).
In sheep meat, increases in pH have been reported to be associated with higher frequencies of
riptions in sheep meat (Young
es, from cooked lamb fat, were measured
mic headspace apparatus (‘electronic nose’). However, no compounds showed a
undesirable odours dominate.
bland / flat / low and stale / musty unacceptable flavour desc
and Braggins, 1996). The effect of pH was shown to be significant, with decreasing odour and
flavour as pH increased. These results are in agreement with early studies in beef by Bouton
et al. (1958), who reported that an increase in pH, due to freezing, was related to a lower
odour of beef. Similar trends have been shown in beef by Dransfield (1981).
In the study by Young and Braggins (1996), volatil
in a dyna
statistically significant increase in concentration with increasing pH. The authors postulated
that the favourable odour notes present in low pH meat may mask the less desirable odour
notes and only when the masking odours are reduced, due to an increase in meat pH, do the
44
y, 1997; Dransfield, 1997; Dutson
t al., 1980; Farouk and Swan, 1997a; Hood, 1980; Morton et al., 1997; O’Mahony et al.,
, with some
uthors indicating that frozen storage decreases tenderness (Jakobsson, 1973; Wheeler, 1990),
ence is a result of different rates and
mperatures of freezing and more importantly, differences in the duration of freezer storage.
tion of thawing have varied. The few studies
2.6.2 Effect of Freezing and Thawing on Flavour
There has been a vast amount of research done in the area of post slaughter factors in relation
to tenderness (Bouton and Harris, 1972c; Bouton et al., 1973a; Bouton et al., 1978b; Bouton
et al., 1978c; Bouton et al., 1980; Aalhus et al., 1994; Dal
e
1997), however this area needs to be further quantified for the effect of processing factors on
flavour.
It has been reported that freezing is detrimental to muscle tissue, due to ice crystal formation;
dehydration and distortion of fibres; increased solute concentration; fat hydrolysis and lipid
oxidation (Farouk and Swan, 1997b; Varnam and Sutherland, 1995), all of which can affect
the palatability of beef. Despite this, conflicting results have been reported
a
while others report an increase in tenderness (Cohen, 1984; Crouse and Koohmaraie, 1990;
Winger and Fennema, 1976; Ferrier and Hopkins, 1997).
It is evident that in part, some of the conflicting evid
te
Additionally, the rate, temperature and dura
which have measured the effects of freezing and thawing on flavour, have reported no
difference in the flavour of beef (Savell et al., 1980; Jeremiah et al., 1993) and pork (Kemp et
al., 1976). Other studies have reported that freezing reduces flavour intensity in beef
(Jakobsson and Bengtsson, 1973; Wheeler, 1990) and lamb (Smith et al. 1968).
45
ally higher (non significant)
alues for a bloody aromatic flavour, whilst the chilled product resulted in a higher (non
between non-volatile
omponents of lean and fatty tissues to give a host of volatile compounds that contribute to
. However, other studies that have examined the effect of
ooking on flavour score have indicated that flavour development is not as intense at lower
Jeremiah et al. (1993) developed flavour profiles to actually describe the flavours. They found
that samples which had been frozen and thawed, had numeric
v
significant) incidence of a browned aromatic flavour.
2.6.3 Effect of Cooking on Flavour
Other studies have shown differences in flavour as a result of cooking methods (Bowers et al.,
1987; Berry and Leddy, 1990a,b; Cross et al., 1976; Cross et al., 1979). It is well known that
meat develops its characteristic flavour and aroma on heating, and that raw meat has little
aroma and only a blood-like taste (Mottram, 1991). Mottram (1991) stated that “during
cooking, a complex series of thermally induced reactions occur
c
the aroma component of the sensation we perceive as flavour”.
The effect of cooking on the structural characteristics of meat, predominantly tenderness, has
been studied extensively by Bouton and others (Beilken et al., 1986; Bouton and Harris,
1972a; Bouton and Harris, 1978; Bouton and Harris, 1981; Bouton et al., 1977; Bouton et al.,
1981; Bouton et al., 1975c; Bouton et al., 1976a, 1976b; Bouton et al. 1976c; Leander et al.,
1980). These studies have shown that in general, as end-point temperatures increase,
tenderness and juiciness decrease
c
end-point temperatures (Berry and Leddy, 1990a; Bowers et al. 1987).
A study to characterise the flavour of beef M. longissimus dorsi muscle at seven internal end-
point temperatures between 55°C and 85°C, reported that at lower end-point temperatures,
46
flavour was characterised by less desirable descriptors. Bloody-serumy, metallic and sour
flavours were noted at low end-point temperatures, whereas at high end-point temperatures
descriptors included mouth-filling-blend and browned flavours; indicating more desirable
avours associated with Maillard browning reactions (Bowers et al., 1987). It was concluded
it should be noted that while research-type uniform cooking methods are
dvantageous for objective measurements and assessment of differences between steaks, they
anese
fl
that beef flavour components and juiciness change most between 55°C and 65°C and then
again between 80°C and 85°C.
In addition to end-point temperatures having an effect on flavour, the duration of cooking is
also known to affect flavour development. Cooking for less time to a rare degree of doneness
(lower end-point temperature), has been reported to be insufficient to develop the typical
browned, caramelised flavours (Berry and Leddy, 1990a). These studies are in contrast to
other reports in which beef flavour acceptability was not affected by cooking rate and where
steaks cooked to lower internal temperatures of 60°C and 70°C were more flavourful than
steaks cooked to 90°C (Cross et al., 1976).
The above studies highlight the huge variations in effect of cooking on palatability.
Additionally,
a
may not be representative of the palatability characteristics observed under another set of
cooking conditions. For this reason, cooking protocols should include both a research type
cooking method for discriminatory power with trained taste panels, and a more representative
cooking method for consumer taste panels (particularly where there may be cultural
differences in cooking style). Busboom et al. (1993) compared an American style broiling
method and a Japanese style shabu-shabu preparation (3mm thin strips dipped in boiling
water). Sensory panel results, using a trained American taste panel, indicated that Jap
47
Wagyu beef was more palatable than other breeds when cooked as shabu-shabu beef, while
palatable when broiled as American style steaks (Busboom et
l., 1993).
ssessment rather than selecting those methods preferred by the bulk of the population.
extremely
omplex sensation. The variation in human perception of flavour makes it necessary for
latability.
is also evident, that the majority of studies on flavour have been conducted in other
ountries such as America and Europe and there seems to be little flavour research reported in
American breeds were highly
a
Dransfield et al. (1984) evaluated beef samples in research centres in Belgium, Denmark,
England, Grance, Germany, Ireland, Italy and the Netherlands. Laboratory panels assessed
meat by grilling the steaks and cooking cubes in casseroles according to local custom and
using different scaling methods. The authors speculated that texture (tenderness and juiciness)
is distinct and more universal than flavour assessment, however recommended that cooking
methods should be selected on the basis of those which give the most discriminating
a
2.7 SUMMARY
The avove review shows that the detection and interpretation of flavour is an
c
thorough training of panellists and use of appropriate experimental designs to ensure
meaningful results. In the case of beef, flavour has been shown to be affected by genetics,
nutrition and post-slaughter treatment factors. It is well established that fatty acid
composition is responsible for major flavour differences, as determined by sensory flavour
panels, and that higher levels of oleic (18:1) acid are associated with improved pa
It
c
48
ustralia. Additionally, apart from Korean studies, there has been no research that
haracterises or compares the meat quality of Hanwoo beef to other cattle breeds.
is therefore the aim of this research project to bridge this gap in knowledge for the benefit
f both the scientific community and the beef industry of Australia. This study will assist in
e understanding of the genetic, nutritional and post slaughter processing factors that
fluence meat quality, particularly flavour, so that future production can be aimed at “tailor
aking” Australian beef to suit Korean and other specific market requirements.
A
c
It
o
th
in
m
50
HODS
determining the relationship between fatty acid
omposition and flavour. For the main study, a subset of animals was further analysed using
3.2 EXPERIMENTAL ANIMALS
3.2.1 Selection of Animals for Preliminary Trial
m the same supplier and representative of those currently exported to
Korea from this abattoir, were selected at the hot weight scales on the basis of carcase weight
and dent cessing,
all carcases were uns . Carcases were followed through to boning, where striploins
were co val inim d ma m values
of the 10 carcases used in the experiment are wn in le 3.1 e gro animals
selected r for all carc traits thus w
replicates for aging and thawing treatments.
3.1 SUMMARY OF MATERIALS AND MET
This project was part of a larger collaborative program between Adelaide University, South
Australian Research and Development Institute (SARDI) and Elders Limited, entitled
“Genetics of beef quality traits – Fat Metabolism” awarded to Pitchford, Siebert and Bottema
(ARC, 1997-1999) and includes two separate studies, which utilised different experimental
animals. A preliminary study aimed at determining the effect of repeated freezing and
thawing on the palatability of beef striploins, whilst the main study aimed at characterising
the flavours of different breeds, in addition to
c
a Chemical Sensor and Gas Chromatography / Mass Spectroscopy instruments.
Ten grass-fed steers, fro
ition. Since the abattoir did not use electrical stimulation as part of it’s pro
timulated
llected from both sides. The mean u SD) and mes (± u anm x uim
sho Tab . Th up of
for the trial were simila ase and ere suitable for use as
51
cross pasture fed steers Mean St.Dev Min Max
Carcases were placed in the chiller approximately 45 minutes post slaughter. After 23 hours,
carcases were taken out of the chiller and quartered at the 10/11th rib site, then allowed to
bloom for a period of 30 minutes. They were assessed on AUSMEAT criteria for marbling
score, meat colour, fat colour, pH, loin temperature and rib fat depth, as shown in Table 3.1.
Throughout chilling, the loin temperature of each carcass and the ambient chiller temperature
were recorded using thermocouple data loggers (Appendix 1, Figure 1).
Table 3.1 - Means, standard deviations and ranges (minimum and maximum) for carcase measurements for 10 Angus and Angus
Dentition 6.0 1.5 2 8 Hot Standard Carcase Weight (kg) 278.0 8.8 263 289 H 3.8 9 20 10 2.8 4 15
ot P8 Fat depth (mm) 13.7 th /11th Rib Fat depth (mm) 9.8
24 hr striploin temperature °C 9.5 0.3 9.2 10.2 24 hr pH 5.8 0.1 5.8 6.0 AUSMEAT Meat Colour 2.7 0.5 2 3 AUSMEAT Fat Colour 1.7 0.6 1 3 AUSMEAT Marbling Score 1.0 0.9 0 2 IM Fat content of striploin (%) 5.4 3.2 2.7 13.1
Approximately 25 hours post slaughter, 20 striploins were collected from the boning room,
acuum-sealed in plastic bags and transported to the research facilities in Adelaide. One
v
striploin from each carcase was frozen immediately (30 hours post slaughter) in a -18°C
freezer (Frozen = F treatment), whilst the striploin from the other side of the carcase was
chilled at 4°C for 27 days (Aged = A treatment). Striploins allocated to the non-aging
treatment were removed from their boxes and laid out individually on racks so that all
surfaces were exposed to air movement for rapid freezing. Temperatures within the 4°C
chiller and -18°C freezer were monitored using data loggers for a period of 4 days to
determine temperature fluctuations (Appendix 1, Table 2).
52
in
ustralia. The animals were a subset of those of the Davies Gene Mapping and Southern
animals in addition to the
ciprocal F1 Jersey x Limousin cross animals. They consisted of purebred Limousins,
born in 1995 (Davies Gene Mapping Project & the Southern Crossbreeding Project) and the
3.2.2 Selection of Animals for Main study
Beef striploins were collected from 170 animals, representative of breeds produced
A
Crossbreeding Projects, located in the mid-North (Martindale) and the South East (Struan) of
South Australia, respectively. Preliminary results from these projects have been reported
(Rutley et al., 1995, Malau-Aduli et al., 1998a,b,c and Malau-Aduli et al., 2000).
The Davies Gene Mapping Project (DGM) comprised a backcross design based on Jersey and
Limousin breeds. This program utilised two breeds, which varied markedly in carcass
attributes; the Jersey (high marbling, low yielding) and the Limousin (low marbling, high
yielding). These breeds were used to produce purebred F1 and backcross progeny with the
aim being to study the mode of inheritance of important meat quality traits and to map major
genes controlling these traits. The present study utilised calves born in 1995 from the parental
generation in Phase 1, and comprised purebred Jersey and Limousin
re
purebred Jerseys and Limousin by Jersey crosses to give LL, JJ and LJ calves, respectively.
Animals from the Southern Crossbreeding Project consisted of seven sire breeds crossed with
Hereford dams. The sire breeds included Belgian Blue, Limousin, South Devon, Hereford,
Angus, Wagyu and Jersey which resulted in BH, LH, SH, HH, AH, WH and JH calves,
respectively. The Southern Crossbreeding program was designed to improve breed utilisation
for targeting specific markets, and the Davies Gene Mapping Project was designed to estimate
the genetic parameters for important meat quality traits across a wide range of genotypes. For
the purpose of this study, 10 animals from each breed were randomly selected from steers
53
hen sire effects were tested, they were generally not significant.
The
(which erent environments) meant that sex was
in f
sex eff ects. The animals from both programs,
wer
(heifers igh density grain ration (65% barley). Steers (DGM and
XB) were slaughtered at 25 months of age (mean carcass weight 326kg), whereas heifers
ere slaughtered at 15 months of age (mean carcass weight 218kg). Animals from these
in the main experiment were derived from:
1) 70 heifers selected from the Southern Crossbreeding Project (SXB) based at Struan
oorte, SA. (Table 3.3)
3) 30 steers selected from the Davies Gene Mapping Project (DGM) based as Mintaro,
heifers born in 1996. Although they were randomly selected, the animals were chosen to
evenly represent sires, to ensure that there were two to four sires represented in each group.
W
fact that heifers were from a different drop and managenement group from the steers,
although finished together, were raised in diff
act confounded with project, drop and management system (cohort group), therefore all
ects throughout this thesis are in fact cohort eff
e raised in three separate management / cohort groups and slaughtered after 80 days
) or 180 days (steers) on a h
S
w
programs were all electrically stimulated within 10 minutes of slaughter, with a current of
200mA and peak voltage of 45V for 40 seconds, using a low voltage rectal-nostril stimulator.
The efficacy of electrical stimulation was monitored on line using a stimulation unit monitor.
The carcases used
Research Centre, Naracoorte, SA. (Table 3.2)
2) 70 steers selected from the Southern Crossbreeding Project (SXB) based at Struan
Research Centre, Narac
Martindale, SA. (Table 3.4)
54
dditionally, 37 Korean Hanwoo (HAN) striploins were imported into Australia, with the
Unfortunately, nformation of
Hanwoo animals from Korea was absent, and e s ot b cluded in statistical
models
In total breed by se lasses: ma - AHF HF, HHF,
JHF, LHF, SHF, WHF; SXB Males - AHM, BHM, HHM, JHM, LHM, SHM, WHM; DGM
ales - LLM, LJM, JJM and Hanwoo animals – Hanwoo F and Hanwoo M.
The striploin (m. longissimus dorsi) was collected from each animal at s ghter and
subsequ amuscular fat tent, ing po of fa atty acid
composistion, objective tenderness and taste pa l flavo sessment (see Figure 3.1). A
further ised for Chemical sor an MS analysis. This
bset will be discussed in section 3.2.3.
A
help of Elders representatives in Korea (Table 3.5). These striploins were typical of the top
grade Hanwoo Beef produced in Korea under feedlot conditions. All striploins had quite high
levels of intramuscular fat. Table 3.2 –3.5 document the carcass attributes for Australian and
Korean animals.
documentation of the precise nutrition, management and sire i
therefor ire has n een in
throughout this study.
, there were 207 samples, 19 x c SXB Fe les , B
M
lau
ently sampled for intr con melt int t, f
ne ur as
subset of these animals was util Sen d GC
su
Table 3.2 - Carcass Characteristics of ‘96 drop SXB Heifers Mean St.dev Min Max
Dentition n/a n/a n/a n/a Hot Standard Carcase Weight (kg) 218 27 149 278 Hot P8 Fat depth (mm) 11 3.5 3 20 AUSMEAT Meat Colour 1B 1.0 1B 5 AUSMEAT Fat Colour 0 0.7 0 4 AUSMEAT Marbling Score 1 0.6 0 2 IM Fat content of striploin (%) 3.4 1.2 1.3 6.7 MUFA % 48 3.0 40 54
55
Table 3.3 - Carcass Characteristics of ‘95 drop SXB Steers Mean St.dev Min Max
Dentition 2 0.6 0 4 Hot Standard Carcase Weight (kg) Hot P8 Fat depth (mm) 16 5.6 5 32 AUSMEAT Meat Colour 1B 0 1B 2 AUSM 2 AUSMEAT 3
326 38 239 398
EAT Fat Colour 0 0.6 0 Marbling Score 1 0.6 0
IM Fat content of striploin (%) 4.5 1.3 1.3 10.9 MUFA % 47 5.6 33 56
Table 3.4 - Carcass Characteristics of ‘95 drop DGM Steers Mean St.dev Min Max
Dentition 2 0.4 0 2 Hot Standard Carcase Weight (kg) 319 69 199 437 Hot P8 Fat depth (mm) 14 4.4 5 24 AUSMEAT Meat Colour 1B 0 1B 2 AUSMEAT Fat Colour 0 0.9 0 4 AUSMEAT Marbling Score 1 0.6 1 3 IM Fat content of striploin (%) 5.4 2.7 1.3 13.2 MUFA % 47 3.9 39 56
Table 3.5 - Carcass Characteristics of ‘97 drop Hanwoo Steers Mean St.dev Min Max
IM Fat content of striploin (%) 10.5 3.0 5.3 16.5 MUFA % 54 5.2 44 65
Figure 3.1 – Striploin treatment allocation for the main study
Objective Tenderness
sample
12/13th rib subcutaneous fat used for fatty acid composition and Melting Point of fat
12/13th rib slice used for measurement of intramuscular fat content (IMF%)
1.5cm thick steaks used for taste Panel flavour assessment
noSteak slice used for chemical
Posterior end t used
sensor analysis
56
3.2.3 Selection of Animals for Odour Assessment using a Chemical Sensor
0 animals raised
nd slaughtered in Korea. The latter were representative of the Korean native breed, the
o Australian animals were a subset of those of the Southern
2) 10 Jersey and 10 Limousin steers selected from the Davies Gene Mapping Project
(DGM) based as Mintaro, Martindale, SA.
Beef Striploins were collected from 30 steers raised in Australia and from 3
a
Hanwo and were grain-fed. The
Crossbreeding and Davies Gene Mapping Projects (Rutley et al., 1995 and Malau-Aduli et
al.., 1998, respectively). The Australian steers (1995 drop) were raised in two separate groups
and slaughtered after 180 days (steers) on a grain ration. Steers were slaughtered at 25
months of age (carcass weight ~300kg). They were derived from :
1) 10 Hereford steers selected from the Southern Crossbreeding Project (SXB) based at
Struan Research Centre, Naracoorte, SA.
3.3 LABORATORY MEASUREMENT OF FAT TRAITS
3.3.1 Muscle Fat Content
Muscle samples from the 12th/13th rib site of the striploin (m. longissimus dorsi) were trimmed
of all visible fat and approximately 100g blended to a homogeneous paste in a food processor
(Braun, Model CAS) with a chopper attachment. A subsample (1.5-2.0g) was accurately
weighed and extracted with chloroform/methanol (2/1) according to the method of Christie
(1989), using a Folch wash. Dried extracts were considered the total fat content. Results
were expressed as a percentage of the wet weight.
57
each capillary tube was inspected momentarily and returned
its appropriate well. Melting point was considered the “slip-point”, the temperature at
on was determined by gas chromatography. The instrument used was a
ewlett Packard Model 5890A gas chromatograph. It was equipped with a 25mm x 0.32mm
(i.d) fused silica capillary column (BPX70, SGE Melbourne, Australia). Fatty acid methyl
3.3.2 Melting point of fat
Approximately 5g of subcutaneous fat was sliced from the samples (used for muscle fat
content) and placed in 10ml glass bottles. These were then heated in an oven at 100°C for at
least 30 min. Approximately 10mm of melted fat was drawn into a 1mm open ended
capillary tube. Once solidified, the upper meniscus of the fat was marked with a felt pen and
groups of samples placed in a refrigerator (2-5°C) overnight. Groups of samples (12-15) were
placed in the wells of a DNA thermal cycler and the wells filled with water. The instrument
was programmed to run from 25°C (5 min), then rise at 1°C per minute for 25 minutes. At
each degree rise in temperature,
to
which the solid fat “slipped” up from its marked position (modified method of AOCS, 1993).
3.3.3 Fatty acid composition
At the time samples were taken for melting point determination, a 20 µl sample of molten fat
was taken for fatty acid analysis of the triacylglycerol fraction of the fat. The fatty acids were
first methylated by the acid methylation procedure of Christie (1989) using 1% H2SO4 in
dried methanol. The methyl esters formed in this reaction were extracted twice with 5ml
petroleum ether (40-60°C) after 3ml of water was added. The petroleum ether extracts were
removed and dried at 40°C in a heating block under a stream of nitrogen gas. The dried fatty
acid methyl esters were reconstituted in 400µl of trimethyl pentane (iso-octane) in vials ready
for chromatographic analysis.
Fatty acid compositi
H
58
nt. The carrier gas was hydrogen
ead pressure 30kPa) and the column was programmed from an initial temperature of 150°C
r each
nimal, 10 replicates of 2.5g samples of meat were weighed and placed into individual 10ml
glass vials. Vials were then sealed er
septum, nd loaded into the HP 4440 Ch amples were transferred singly to an
oven and equilibrated at 100°C for 30 m ced the septum
and the ial was pressurized with heli s then vented back
through the needle and a sample loop. subsequently swept by a
stream of helium through a heated trans into the mass sensor. The
mass range selected for these samples wa mass 35 (Ion 35) and ionic mass 180
(Ion 180 ll samples had been a e means of the abundances
ester preparations were injected (0.1µl) into the instrume
(h
(0 min) to a final temperature of 170°C (0 min) at the rate of 2°C per minute. Chromatogram
output was computed with an integrator recorder (Hewlett Packard Model 3392 Series II).
Standard fatty acid methyl ester (FAME) mixtures were used to calibrate both gas
chromatographic systems using reference standards GLC-68-B, Nu-Check Prep.
Identification of sample fatty acids was made by comparing the relative retention times of
peaks from samples with those of standards. These were calculated as normalised area
percentages of fatty acid. The names and numeric symbols of the individual fatty acids are
shown in Table 2.5.
3.4 CHEMICAL SENSOR ANALYSIS
From the 60 selected striploins, meat from each animal, was run through an automated
headspace sampler coupled to a quadrupole mass sensor (HP 4440 Chemical sensor).
Individual steaks were thawed for 21 hours at 3°C, before cutting into small pieces. Fo
a
with a crimp cap and a PTFE-coated silicone rubb
a emical Sensor. S
inutes, following which a needle pier
v um. The headspace vapour wa
The loop contents were
fer line (125°C) directely
s between ionic
). Once a nalysed in this manner, th
59
of ions 35 to 180 were calculated from t ach of the 60 samples. Table
3.6 details the HP 4440 chemical sensor parameters required to set up the autosampler.
Table 3.6 - HP 4440 Chemical Sensor Headspace Autosampler Parameters
Zone Temperatures °
he 10 replicates from e
Oven = 100 C Loop = 110°C Tr. Line = 125°C HP Temperature Controller = 125°C Event Times HS Cycle time = 5.2 min Vial Equilibrium time = 30 min Pressurize time = 0.3 min Loop fill time = 0.15 min Loop equilibrium time = 0.02 min Injection time = 0.5 min Vial parameters First vial = 1 Last vial = 41 Shake = 2 (High) Pressure Carrier Pressure = 1.9ps Vial Pressure = 14psi
3.5 OBJECTIVE MEASUREMENTS OF TENDERNESS
Objective samples (8cm width portions) were thawed in sealed plastic bags for 24 hours at
4°C before testing. A small block of approximately 150g was prepared from the striploin by
removing all subcutaneous fat and connective tissue. Following preparation, the blocks were
placed on a tray at 4°C and allowed to bloom for 30 minutes prior to pH and colour
measurements. Measurements of pH were taken using a pH meter fitted with a specially
designed meat probe. Meat colour was measured using a Minolta CR-300 to obtain CIE L*
(lightness) a* (redness) and b* (yellowness) values (Hunter and Harold, 1987). Meat colour
was only measured in the preliminary trial and not in the main trial.
Prior to cooking, the 150g blocks were placed into the chiller for 30 minutes to ensure all
samples were at similar temperatures. Samples were cooked in individual plastic bags in a
60
e of doneness. They were then cooled in
old running water for 30 minutes, before returning to the chiller for overnight storage at 4°C.
sting machine (Model LRX) - Main Trial
tion of the degree of
ifference between samples.
70°C water bath for 45 minutes to a medium degre
c
The following day, samples were prepared for assessment. Several slices (0.66cm) were
taken from the lateral side and subsequently cut into 1.5cm width sections following the
direction of the meat fibre. Tenderness measurements were performed so that the 0.66cm
thick samples were sheared perpendicular to the direction of fibres, using a Warner-Bratzler
shear device fitted to :
1) an Instron Materials Testing machine (Model 4301) - Preliminary Trial
2) a Lloyd Instron Te
3.6 TASTE PANEL EVALUATION
There are numerous sensory methods used for determining flavour attributes such as ranking,
scaling and descriptive analysis. The method used depends on the outcome desired and the
number of samples to be tested. Ranking involves the panellist ranking samples in order of
intensity of an attribute. This method requires little training, however the panellists must be
familiar with the attribute in question (Meilgaard et al, 1991). Additionally ranking can only
be used on a limited number of samples and the results give no indica
d
Scaling techniques, on the other hand, use numbers or words to express the intensity of an
attribute and can discriminate better between samples than ranking. The most common types
of rating scales used are those with 7 to 9 categories. Such scales can be arranged as a line
scale or category scale. Panellists rate the intensity of an attribute by making a mark on a
horizontal line, normally 15cm with word or numerical anchors at each end. The left-hand
61
f a product Caul (1957). The profile
ethod firstly considers the overall impression (amplitude) of the aroma, followed by a list of
our and subtleties of the product, rather than
st indicating whether it has a strong or bland flavour.
side of the scale is equal to none or zero stimulus, whilst the right end represents a strong
level of stimulus (Meilgaard et al, 1991). One of the main problems with scaling methods is
that untrained panellists may avoid using the extreme ends of the scale.
Descriptive analysis techniques such as the SPECTRUM™ and Flavour Profile methods can
be used to gain very informative descriptions of products, however require panellist training
which can be quite time consuming and expensive. The profile method of flavour analysis
was developed by Arthur D. Little in the late 1940’s (Meilgaard et al., 1991) and was reported
as a technique which could focus on the whole flavour o
m
factors detectable in the aroma, their order of appearance and the degree of detectability
(intensity) as threshold, slight, moderate and strong. Once aroma has been characterised, the
same method is followed for flavour analysis. The profile method of analysis is based on
descriptive or associative terms developed by the flavour panel for a particular product.
Where possible the associative terms refer to a definite chemical or reference material (Caul,
1957). The author suggested that the flavour profile method gives more information on
flavour than other sensory tests such as ranking of samples, scaling methods and difference
tests because it describes the characteristic flav
ju
Keeping these techniques in mind it was decided that for the purpose of the main study, a 9-
point category scale would be developed which could not only give us information on the
individual flavours present but also indicate the magnitude and direction of difference in
flavours between samples. One of the advantages of utilising this system was the fact that it
could be used with a semi-trained flavour panel, without the need for time consuming and
62
eliminary Study
ollowing thawing for 21 hours at 3°C, 2cm thick steaks were grilled to an internal
ession was held before the first tasting,
amiliarised with the scoring system and the attributes to be rated.
0 samples was given a random 3 digit number code to
inate any chance of panellist bias. Additionally, the 60 samples were split into two steaks
session.
The first samp start of each
ssion by every taster. The remaining 6 samples consisted of the 6 aging by thawing
treatments applied to an individual animal. Panellists tasted the 6 samples in a random order,
costly training of a flavour profile panel. Additionally, a category rather than a line scale was
selected for ease of data entry, and is unlikely to differ from a continuous scale due to very
few values being close to the extremes (1 & 9).
3.6.1 Pr
F
temperature of 70°C on a double-sided hotplate (Silex® 610-80, Hamburg), then rated by 20
untrained panellists. They were selected from the University of Adelaide staff based on their
availability and interest in tasting beef. The age range was from 25 to 60, with approximately
equal numbers of males and females. A brief training s
when panellists were f
Samples were rated on a 9-point category scale for tenderness, juiciness, beef flavour, foreign
flavour and acceptability (where 1 = extremely tough, extremely dry, extremely bland, not
detectable and dislike extremely, respectively). The score-sheet for this trial is shown in
Table 3.7.
Sixty different samples were tested (2 aging by 3 thawing treatments applied to striploins
from 10 animals). Each of the 6
elim
so that 120 samples were tested throughout the study.
Six tasting sessions were conducted with the 20 panellists receiving 7 samples each
le (replicate sample) was a ‘warm-up’ sample and was tasted at the
se
63
ples
Figure 3.3 - Taste Panellist rating samples
Samples were rated on a 9-point category scale for
tenderness, juiciness, beef flavour, foreign flavour and
acceptability (where 1 = extremely tough, extremely
dry, extremely bland, not detectable and dislike
extremely, respectively).
which was pre-allocated, to eliminate any effect of tasting order. Panellists received sam
from a different animal every session, ensuring that the 6 treatments from each animal, were
tasted by 12 different panellists.
Panellists were allocated to one of 5 groups, so that the same four panellists remained in the
same group, for each session (Figure 3.2 and Figure 3.3). Tasting sessions were conducted
over a period of 6 days, with 5 different groups of panellists attending each session.
Figure 3.2 - Taste Panellist’s conducting tastings for the trial
Table 3.7 - Meat Tasting Score Sheet for Preliminary Trial - You will receive 7 cubes of meat to taste. Please taste samples in the order they appear on your scoresheet and tick the box which best describes each attribute. Samples may be spat out in the cup provided after evaluation if necessary. Water may be used to cleanse the palate between samples. Tenderness - the ease with which meat structure is disorganised duing mastication. - Rate sample on intensity, from tough to tender. Juiciness - the juice released by the meat on mastication. - Rate sample on intensity of the amount of juice present, from dry to juicy. Flavour - the detection of volatiles in the nasal passage, which includes the aromatics; basic tastes and mouthfeels. - Rate sample on intensity of beef flavour, from bland to flavoursome. Foreign Flavour - Foreign (normally undesirable) odours or flavours present which are not normally associated with beef flavour. ie : Rancid (off) flavours and odours caused by oxidisation of fats (Painty, Fishy odours). - Rate sample on the intensity of foreign flavour, from not detectable to extremely strong. Acceptability - The overall impression of the sample. - Rate the sample on its intensity of desirability. Sample No _________ TENDERNESS Extremely Very Moderately Slightly Neither Tough Slightly Moderately Very Extremely Tough Tough Tough Tough nor Tender Tender Tender Tender Tender
JUICINESS Extremely Very Moderately Slightly Neither Dry Slightly Moderately Very Extremely Dry Dry Dry Dry nor Juicy Juicy Juicy Juicy Juicy
BEEF FLAVOUR Extremely Very Moderately Slightly Neither Bland Slightly Moderately Very Extremely Bland Bland Bland Bland nor Flavoursome Flavoursome Flavoursome Flavoursome Flavoursome
FOREIGN FLAVOUR Not Just Very Mild Distinct Extremely Detectable Detectable Mild Mild to Distinct Distinct to Strong Strong Strong
ACCEPTABILITY Dislike Dislike Dislike Dislike Neither Like Like Like Like Like Extremely Very Much Moderately Slightly nor Dislike Slightly Moderately Very Much Extremely
64
65
3.6.2 Main Study
Following thawing for 21 hours at 3°C, 1.5cm-thick steaks were grilled to an internal
temperature of 70°C on a double-sided hotplate (Silex® 610-80, Hamburg). Steak cubes (2 x
2 x 1.5cm) were rated by a 25-member semi-trained taste panel. Panellists rated samples on a
9-point scale for initial juiciness and sustained juiciness (where 1 = extremely dry) and
flavour acceptability (where 1 =extremely unpleasant). Additionally, panellists rated the
intensity of beef flavour, beef fat flavour, oily flavour, buttery flavour, chicken flavour, corn
flavour, grassy flavour and rancid flavour (where 1 = not detectable and 9 = extremely
strong). The score-sheet for this study is shown in Table 3.8.
The 25 panellists were selected from a group of 35 Meat Production students based at the
Roseworthy campus of the University of Adelaide. Three training sessions were conducted as
part of their normal practical sessions. During this time, the aim was to familiarise the
panellists with the scoring system and flavour terminolgy, in addition to improving their
ability to recognise and identify the sensory attributes to be assessed. The first practical
involved panellists using a triangle test to pick the ‘odd sample out’. In order to try and
calibrate the panel, a second training session involved panellists being presented with
reference samples representative of the extremes in attributes likely to be encountered
throughout the study. Students had to taste these samples and score them using the scoring
system to be used in the trial. Following this, a group discussion of each sample was carried
out to describe the tenderness, juiciness, and in particular the flavour of the sample. Students
were also given 2 theory sessions on sensory analysis and had to complete a written
assignment. Students were screened on the basis of interest and ability to attend the tasting
sessions. There were approximately equal numbers of males and females, ranging between
18 and 25.
Table 3.8 - Meat Tasting Score Sheet for Main Trial SAMPLE
INITIAL Extremely Very Moderately Slightly Neither Dry Slightly Moderately Very Extremely JUICINESS Dry Dry Dry Dry nor Juicy Juicy Juicy Juicy Juicy
SUSTAINED Extremely Very Moderately Slightly Neither Dry Slightly Moderately Very Extremely JUICINESS Dry Dry Dry Dry nor Juicy Juicy Juicy Juicy Juicy
FLAVOUR Not Just Very Mild Distinct Extremely Detectable Detectable Mild Mild to Distinct Distinct to Strong Strong Strong
BEEF
BEEF FAT
OILY
BUTTERY
CHICKEN SKIN
CORN
GRASSY
RANCID / OFF
FLAVOUR Extremely Very Moderately Slightly Neither Pleasant Slightly Moderately Very Extremely ACCEPTABILITY Unpleasant Unpleasant Unpleasant Unpleasant Nor Unpleasant Pleasant Pleasant Pleasant Pleasant
66
67
Seven tasting sessions were conducted with the 25 panellists receiving five or six samples
each session. Samples were given a random 3 digit number code to eliminate any chance of
panellist bias. Due to the large number of samples to taste, a warm up sample was not given.
However, panellists tasted the six samples in a random order, which was pre-allocated, to
eliminate any effect of tasting order. Each steak (n=207) was tasted by five different
panellists in a different order. During each session, panellists received 4 samples from the
SXB programs (steers and heifers), 1 DGM steer sample and 1 Hanwoo sample. This ensured
that the samples from each breed and program were not only spread evenly throughout the
sessions, but also that a comparison between the Hanwoo and the Australian breeds could be
made at every session (panellists were unaware of which sample was the Hanwoo).
3.7 STATISTICAL ANALYSIS
Least squares means of flavour and fatty acid data were calculated for each breed and sex
class, using PROC GLM (SAS, 1996). The specific fixed effects and covariates fitted in the
models are given in detail in subsequent chapters. Intramuscular fat percentage was used as a
covariate in a number of models in order to remove the variation in intramuscular fat between
breed classes, so that flavour differences could be assessed more accurately. Residual
correlations between traits were computed by PROC CORR (SAS, 1996), after adjusting taste
panel data for main taste panel effects including session, group and taster. Multivariate
Analysis techniques including Principal Components analysis and Cluster Analysis (PROC
PRINCOMP and PROC CLUSTER, SAS, 1996) were used as another way of looking at the
taste panel data. Finally, PROC REG and PROC GLM (SAS, 1996) were used to develop
prediction equations for flavour, based on fatty acid and chemical sensor measurements. All
statistical tables from these analyses are presented within each relevant Chapter of the thesis.
68
CHAPTER 4
AND THAWING ON THE EATING QUALITY OF
BEEF STRIPLOINS
EFFECT OF AGING AND REPEATED FREEZING
69
4.1 INTRODUCTION
Korean consumers regard beef imported from Australia as low quality in comparison to beef
produced in the Korea or the USA (Ryu et al., 1994; CSIRO, 1995). Their ultimate
preference is for the Korean native cattle breed, Hanwoo. Baghurst (1997) examined the
attitudes of Korean adults from Seoul and Pusan to protein-based foods. The study showed
that sensory issues were of major importance, and the lack of data in this area demonstrated
the need for research.
Figure 4.1 - Australian Quarter beef destined for the Korean Market
70
The majority (61%) of Australian beef destined for Korea is shipped in the form of frozen
quarters of beef (AMLC, 1996). Once carcases arrive in Korea, they are packaged by the
National Livestock Cooperatives Federation (NLCF) prior to retail sale. This process
involves thawing of the frozen quarters (often by submersion in tanks of hot water) before
they are boned out to Korean specifications. Finally, cuts are re-frozen before distribution to
supermarket chains (Meat Research Corporation, 1993; CSIRO, 1995). Koreans have
described Australian beef as poor in meat colour, taste and smell (Meat Research Corporation,
1993). This reaction is possibly due to the thawing and refreezing treatment.
Figure 4.2 - Australian quarter beef being prepared for freezing
It has been reported that freezing is detrimental to muscle tissue, due
to ice crystal formation; dehydration and distortion of fibres;
increased solute concentration; fat hydrolysis and lipid oxidation
(Farouk and Swan, 1997; Varnam and Sutherland, 1995), all of
which can affect the palatability of beef. Despite this, conflicting
results have been reported, with some authors indicating that frozen
storage decreases tenderness (Jakobsson and Bengtsson, 1973;
Wheeler et al, 1990), while others report an increase in tenderness
(Cohen, 1984; Crouse and Koohmaraie, 1990; Winger and Fennema,
1976; Ferrier and Hopkins, 1997).
It is evident that, in part, some of the conflicting evidence is a result of different rates and
temperatures of freezing and differences in the duration of freezer storage. Additionally, the
rate, temperature and duration of thawing were varied. The few studies which have measured
the effects of freezing and thawing on flavour, have reported no difference in the flavour of
71
beef (Savell et al, 1980; Jeremiah et al, 1993) and pork (Kemp et al, 1976). Other studies
have reported that freezing reduces flavour intensity in beef (Jakobsson and Bengtsson, 1973;
Wheeler et al, 1990) and lamb (Smith et al, 1968).
The primary aim of this study was to evaluate the processes involved in the treatment of
Australian beef exported to Korea, in order to determine the effect of repeated freezing and
thawing (including water thawing) on beef tenderness and flavour. A secondary aim was to
determine whether fatty acid composition alters during storage.
4.2 MATERIALS AND METHODS
4.2.1 Carcase characteristics
Ten grass-fed steers, from the same property and representative of those currently exported to
Korea, were slaughtered at the Murray Bridge abattoir (T&R Pty. Ltd) and a subsample
selected at the hot weight scales on the basis of carcase weight and dentition. They were
followed through to boning, where striploins were collected from both sides of the animals. A
more detailed description of the carcase characteristics of the animals used in this study is
presented in Chapter 3 - section 3.2.1. The group of animals selected for the trial were
similar for all carcase traits and thus were suitable for use as replicates for aging and thawing
treatments.
4.2.2 Treatment allocation
After 27 days of aging, 10 aged (A) and 10 frozen (F) striploins were each cut into three 15cm
width portions, and randomly allocated to a freezing and thawing treatment group (1, 2 or 3).
The 60 portions were then individually weighed, packaged in polyethylene bags, sealed and
72
placed in a -18°C freezer. Treatments were randomised so that each thawing treatment
appeared in each position along the length of the striploin, in order to minimise any effect of
position of muscle on the treatment.
Therefore, 6 treatments A1, A2, A3, F1, F2 and F3 were applied to each of the 10
experimental animals. Thawing protocols consisted of modifications of recommended
procedures for Korean processing of Australian beef (MRC, 1993). Seven days post
treatment, aged and frozen samples from treatments 2 and 3 (individually wrapped in plastic
bags) were placed in a 20°C water bath for a period of 3.5 hours until thawing was complete.
Portions were then re-weighed, re-packaged and laid out on racks in the -18°C freezer and re-
frozen. After a further seven days, aged and frozen samples from treatment 3 were laid out in
the 4°C chiller and air thawed (still enclosed in sealed plastic bags) for a period of 24 hours
until thawed. Portions were then re-weighed, re-packaged and frozen as before. A summary
of the treatment protocol is shown in Figure 4.3 and a full outline of the treatment allocation
is presented is Appendix 1, table 1. Samples were kept in frozen storage for 2 weeks before
testing. During this time, each of the 60 portions were further divided into three 2cm width
steaks for sensory assessment, and one 8cm width portion for objective tenderness
measurement.
Moisture loss was measured at all stages of the trial – during aging (Aging loss), during
thawing treatment (Thawing loss), thawing before testing (Post-Tmt loss), and cooking
(Cooking loss). Total moisture loss (Total loss) was calculated as the sum of various losses.
73
Figure 4.3 - Aging and Thawing treatments applied to striploins.
Frozen, thawed at 4°Cfor testing
THAWINGTmt (1)
1 Thaw
Frozen, water thaw at 20°C,re-freeze, thawed at 4°C
for testing
THAWINGTmt (2)2 Thaws
Frozen, water thaw at 20°C,re-freeze, air thaw at 4°C,re-freeze, thawed at 4°C
for testing
THAWINGTmt (3)3 Thaws
AGING TMT - A & F(A) Aged for 30 days at 4°C, frozen at -18°C.(F) Frozen at -18°C, 25 hrs post slaughter.
4.2.3 Other Measurements
Melting point of fat and triacylglyceride fatty acid determination were measured on both F1
and A3 treatments, whilst muscle fat content determination was carried out on F1 samples
only. The methods for the measurements of these fat traits are documented in Chapter 3 –
sections 3.3.1, 3.3.2 and 3.3.3.
Chapter 3 – section 3.5 outlines the objective measurements of meat colour, pH and
tenderness (measured using a Warner-Bratzler shear device fitted to an Instron Materials
Testing machine (Model 4301)). In addition to measuring meat colour, subcutaneous fat
samples were retained for colour assessment. Figure 4.4 shows the division of the striploin
into 3 thawing treatments and the site at which Warner Bratzler, and sensory samples were
taken for objective (Obj) and subjective (Subj) measurements respectively.
74
Figure 4.4 - Striploin treatment allocation
Surface usedfor pH &
colourmeasurement
Side from which WBshear samples were cut
4.2.4 Taste Panel Evaluation
Following thawing for 21 hours at 3°C, 2cm thick steaks were grilled to an internal
temperature of 70°C on a double-sided hotplate (Silex® 610-80, Hamburg), and rated by 20
untrained panellists. Details of the taste panel evaluation are given in Chapter 3 – section
3.6.1.
4.2.5 Statistical Analysis
Analysis of variance was carried out using the GLM procedure (SAS, 1996) using Type I
sums of squares on sensory data containing animals 1-10 as replicates for aging and thawing
treatments. In the initial model, all main effects and their interactions were tested against the
total error mean square, in addition to being tested on the appropriate error term (Table 4.1).
Following this, non-significant main effects and interactions were removed from the model.
The final model consisted of aging, thawing, aging by thawing interaction, group, taster
nested within group, and session by taster nested within group, tested on the total error mean
sqaure.
75
Table 4.1 - Main effects and interactions tested in the initial model - GLM (SAS, 1996).
Source DF Aging 1 Thawing 2 Aging x Thawing 2
Aging x Thawing x Animal 54 Animal 9 Aging x Animal 9 Thawing x Animal 18 Aging x Thawing x Animal 18
Session x Taster(Group) 110 Session 4 Group 4 Taster(Group) 19 Session x Group 12 Session x Taster(Group) 71
Session x Taster(Group) 110 TOTAL 839
Least squares means and standard errors were calculated for taste panel evaluations of
tenderness, juiciness, beef flavour, foreign flavour and acceptability. Additionally, the
residual least squares means for the final model were tested for correlations between attributes
using the CORR procedure (SAS, 1996).
A test of repeatability of a taster was carried out on the ‘warm-up’ sample (replicate), which
was tasted by every taster at the beginning of each session. The MIXED procedure (SAS,
1996), which consisted of session and taster as main effects in the model, was used to test the
repeatability of the panellists (tasters) for all attributes. Taster was classed as a random effect.
Repeatability was calculated from the covariance parameter estimates for taster.
The same model structure was used to determine the effect of storage and treatment on
moisture loss from storage through to cooking, in addition to total moisture loss. The residual
least squares means for each model were tested for correlations with taste panel ratings of
juiciness.
76
A simple GLM model (SAS, 1996), containing animal and treatment (A3 or F1) was used to
determine the effect of treatment on the fatty acid composition and melting point properties of
the fat.
Warner-Bratzler shear results were tested using a model containing animal, aging, thawing,
the two-way interaction between aging and thawing and the three-way interaction between
animal, aging and thawing.
A model containing aging, thawing and the two-way interaction was also used to test meat
colour, fat colour (L*, a*, b* values) and pH. Least squares means were calculated.
4.3 RESULTS
The effect of taster on all palatability attributes was significant and needed to be adjusted for
in the final model (Table 4.2). The fact that all attributes were significant for taster nested
within group, suggests that tasters were quite varied in their ratings. This is not surprising
considering that the panel was untrained.
The significance of aging and thawing treatments on palatability attributes are shown also in
Table 4.2. When group, taster nested within group and session by group interactions were
significant, there was an indication that significant adjustments had been accounted for in the
model. Overall, the final model explained 57%, 24%, 29%, 46% and 43% of the amount of
variation in tenderness, juiciness, beef flavour, foreign flavour and acceptability respectively.
77
The least squares means and standard errors for palatability attributes are shown in Table 4.3.
Aging and thawing produced significant increases in tenderness. Additionally, the 2-way
interaction between aging and thawing was highly significant, since thawing produced larger
improvements in the tenderness of frozen beef than aged beef. Despite this, freezing and
thawing did not make meat as tender as aged meat.
The effect of aging and thawing on juiciness was moderately significant, with no significant
interaction seen. Repeated thawing decreased the juiciness of both aged and frozen samples.
Aged samples (A1) were significantly more flavoursome than frozen samples (F1), whilst
repeated freezing had no effect on beef flavour. There was little difference in panellists’
scores for foreign flavour. Aged samples were rated as having only a slightly higher
incidence of foreign flavour.
Table 4.2 - Tests of Significance for all attributes tested by the taste panel. Source DF Tenderness Juiciness Beef
Flavour Foreign Flavour
Acceptability
Aging 1 *** * *** * *** Thawing 2 *** ** NS NS *** Aging by Thawing
2 *** NS NS NS ***
Group 4 *** NS *** *** *** Taster within
Group 19 *** *** *** *** ***
Session by Group
25 *** *** *** NS ***
Repeatability %
- 22% 37% 38% 25% 46%
* P<0.05, ** P<0.01, *** P<0.001
78
Table 4.3 - Least squares means and standard errors for each treatment as reported by taste panel ratings and tenderness objective measurement.
Tenderness WB shear kg force
TendernessTaste Panel
Juiciness Beef Flavour Foreign Flavour
Acceptabilty
A1 3.51a 6.39c 5.59a 5.97ac 1.92ab 6.12a
A2 3.26b 6.63cd 5.40abc 5.94ac 1.98a 6.12a
A3 3.48ab 6.78d 5.18bcd 6.09c 2.06a 6.08ac
F1 6.18d 3.66a 5.50ab 5.47b 1.98a 4.69b
F2 5.33c 3.97a 4.96dc 5.70ab 1.66bc 4.80b
F3 5.24c 5.33b 5.09c 5.78abc 1.79ac 5.72c S.E 0.08 0.13 0.13 0.13 0.11 0.14
Note : Least squares means with different superscripts within columns are significantly different (P<0.05) 1=extremely tough, extremely dry, extremely bland, not detectable and dislike extremely and 9 = extremely
tender, extremely juicy, extremely flavourful, extremely strong and like extremely. For WB tenderness <5kg shear force = tender
Figure 4.5 - Correlation between sensory tenderness scores and objective Warner Bratzler shear values (WB Tend)
y = -0.85x + 9.16R2 = 0.91
4
5
7
Sens
ory
Tend
1
WB Tend (kg F)
2
3
6
8
9
0 2 4 6 8
The effect of aging and thawing on acceptability, tended to follow the same pattern as
tenderness. Aged samples were significantly more acceptable than frozen samples, whilst the
effect of thawing only increased the acceptance of frozen samples.
79
on tenderness, with the
xception of A2, which was only slightly different from A1 (P<0.05). In contrast, for the
froze
Acceptability was positively correlated wi n iness and beef flavour and
negatively correlated with foreign f r (Tab ). Ad onally, te ess, juiciness and
beef flavour were moderately correlated with other .29-0.41) e high correlation
(0.91) of Warner Bratzler shear force values with sensory tenderness scores (Figure 4.5)
indicated that the taste panel was able to effectively score tenderness.
Juiciness Beef Foreign Acceptability
For both sensory and objective tenderness, aged samples were significantly more tender than
frozen samples. Additionally, freeze / thawing significantly affected tenderness. The
interaction between freeze / thawing and aging was significant, indicating that the aged and
un-aged meat behaved differently with repeated freezing and thawing. Warner-bratzler
results indicate that for aged samples, freeze / thawing had little effect
e
n samples, thawing significantly increased tenderness.
th tender ess, juic
lavou le 4.4 diti ndern
each (0 . Th
Table 4.4 - General Linear Models Procedure – Residual Correlations between palatability attributes
Flavour Flavour Tenderness 0.41 0.29 -0.08 0.54
*** *** * *** Juiciness 0.40 -0.09 0.50
Beef Flavour *** * ***
-1.00 0.58
Foreign Flavour** ***
-0.27 ***
an the frozen sample
llowing storage. However, once thawing treatments were applied to frozen samples, there
was significantly more drip loss from frozen samples (Thawing loss). There was a significant
The effect of aging and thawing varied in their effect on drip loss depending on the period of
treatment (Table 4.5). Not surprisingly, aged beef lost more water th
fo
80
creased drip loss in both aged
nd frozen samples during this stage. The effect of aging on cooking loss caused significantly
m
Aged samples lost sign ly more ure than samp ted thawing
also increased drip loss from frozen and aged samples (Table 4.5). Juiciness was negatively
correlated with both cooking loss and total moisture loss (-0.18 and –0.16 respectively),
indicating th s moistu ss increas uiciness d sed .
Loss Loss Loss Loss
interaction between aging and thawing treatment on drip loss during this period following
treatment. After the final thaw of samples prior to testing (Post-treatment loss), there was
significant effects of aging, thawing and their interaction was seen, with frozen samples losing
a greater percentage of drip. Repeated freezing and thawing in
a
ore moisture to be lost from aged samples than from frozen samples.
ificant moist frozen les, whilst repea
at a re lo ed, j ecrea
Table 4.5 - Least squares means and standard errors for percentage moisture loss for each treatment
Aging Thawing Post-Tmt Cooking Total Loss
A1 1.7a 0.0a 1.3a 24.4a 26.6a
A2 1.7a 1.2a 2.0b 23.9a 27.5a
F1 0.2b 0.0a 3.9c 21.4b 24.7c
1.0a 4.2c 22.8c 26.9a
F3 0.2b 7.8c 1.9b 20.8b 28.5d S.E 0.12 0.41 0.14 0.32 0.33
A3 1.7a 4.8b 1.9b 24.2a 30.4b
F2 0.2b
Note : Least squares means with different superscripts within columns are significantly different (P<0.05)
Aged meat was redder (higher CIE a*) and yellower (higher CIE b*) than meat that had been
frozen (Table 4.6 and Table 4.7). The effect of repeated freezing and thawing significantly
increased CIE L* (lightness) values. There was a slightly significant interaction effect
between aging and thawing, such that the effect of thawing treatment only increased L*
values of frozen samples.
81
at from the frozen samples were
dder (higher CIE a*) values than fat from the aged samples. The effect of thawing
significantl ference in
b* values between aged a frozen f z n am si nifica tly higher
values. However, b* values were not altered aw tre nt ed les had
significan pH th froz samp In line with ect of repeated freezing
nd thawing also led to an increa
ur ou
Aged samples were lighter (higher L* values) for fat colour, whilst thawing treatment had the
effect of only increasing L* values for frozen samples. F
re
y reduced a* values of frozen samples only. There was a significant dif
nd samples, with ro e s ples having g n
by th ing atme . Ag samp
tly higher an en les. this, the eff
a sed pH of frozen samples.
Table 4.6 - Tests of Significance for meat colour and pH attributes. Meat Colo ColFat r
Source DF L* a* L* a* b* b* PH Ag 1 *** *** ing NS *** *** *** ***
Thawing 2 * *** *** ** NS NS NS *** Aging by
wing2 *** *
Tha * NS NS NS *
* P<0.05, ** P<0.01 <0.00
, Fat Colour (CIE L*, a*, b*) and pH for each aging and thawing treatment.
Meat Colour Fat Colour
, *** P 1
Table 4.7 - Least squares means and standard errors for Meat Colour (CIE L*, a*, b*)
L* a* b* L* a* b* pH A1 35.6a 17.6a 9.0a 70.2a 0.4ac 14.7a 5.57a
A2 33.7b 18.4a 9.8a 68.1bd -0.2a 13.8ab 5.66b
F1 33.6b 14.0b 7.0b 64.0c 5.7b 17.7c 5.44c
F3 35.9a 14.6b 7.6b 69.7ab 1.8cd 17.3c 5.56a
A3 36.1a 17.8a 9.8a 70.6a -0.8a 12.5b 5.62b
F2 34.7ab 14.4b 7.4b 67.2d 3.2d 17.4c 5.51d
S.E 0.58 0.67 0.32 0.69 0.51 0.54 0.02 Note : Least squares means with different superscripts within columns are significantly different (P<0.05)
Fatt is,
there were lmitic and
vaccenic fatty acids, however th nces did not affect the total MUF ing point
y acid composition was not affected by treatment (F1 compared with A3). Despite th
only small significant differences between animals for myristic, pa
ese differe A or melt
82
which were both ected by d animal. Tab s th ments of
fatty acids for A3 and
Table 4.8 - Means and standard deviations for fatty acid composition for the two extreme treatments F1 (frozen, one thaw) an , 3
name
unaff treatment an le 4.8 show e measure
treatments F1.
d A3 (aged thaws).
Fatty acid Common F1 A3 14:0 Myristic 4.5 ± 0.9 4.7 ± 0.7 14:1 Myristolic 2.1 ± 0.5 2.1 ± 0.8 16:0 Palmitic 31.3 ± 2.5 32.1 ± 1.9
16:1 (9c) Palmitoleic 16:1 (9t) 1.0 ± 0.2 0.9 ± 0.2
Stearic 14.4 ± 1.7 14.2 ± 1.8 Elaidic 1.9 ± 3.8 0.6 ± 1.2
18:1 (11t) Trans-vaccenic 0.8 ± 0.8 0.5 ± 1.0 18:1 (9c) Oleic 37.1 ± 4.4 37.7 ± 2.7
4.9 ± 1.4 5.0 ± 1.2
18:0 18:1 (9t)
18:1 (11c) Vaccenic 0.4 ± 0.4 0.4 ± 0.3 18:2 Linoleic 1.2 ± 0.2 1.3 ± 0.2 18:3 Linolenic 0.4 ± 0.1 0.4 ± 0.1
Sum mono-unsaturated fatty acids (MUFA) 44.6 ± 5.4 45.2 ± 3.8 Melting Point of fat 39.7 ± 1.3 40 ± 1.3
4.4 DISCUSSION
The repeatability between panellists, for palatability scores of the replicate sample (R) was
disappointingly low for all attributes (Table 4.2). This can be explained by a number of
factors including the fact that the replicate sample was tasted as the first sample in each
session, and thus may be considered as a ‘warm-up’ sample. Additionally, panellists may
have been utilising different parts of the scale to score palatability attributes, which would
have the effect of lowering overall repeatability. Despite this, the final model contained
significant adjustments for the effect of session and taster-within group to adjust for this.
Additionally, the objective measurements of tenderness (obtained from Warner-bratzler shear
results) were in agreement with the taste panel results (as shown in Table 4.3 and Figure 4.5).
83
s
nstable. It was found to decrease in activity by 45%, after 6 weeks storage at -70°C.
a (1976) and Cohen (1985).
son and
engtsson, 1973; Wheeler et al, 1990) may be a result of testing meat immediately following
As shown in Table 4.1, aged samples were significantly more tender than frozen samples,
whilst thawing treatments had a larger effect on frozen samples. Frozen samples that had
been thawed three times (F3), were much more tender (5.24 vs 6.18kg shear force) than those
that had been thawed only once (F1), suggesting that the meat aged whilst thawing. Crouse
and Koohmaraie (1990) suggested that freezing meat before the aging period may enhance
postmortem proteolysis. The calcium dependent proteases CDP-I and CDP-II are thought to
be responsible for postmortem aging in beef, whilst an endogenous inhibitor (calpastatin),
inhibits the activity of both proteases (Koohmaraie, 1990). The same author also reported that
CDP-I and CDP-II were stable under conditions of frozen storage, however, calpastatin wa
u
Whipple and Koohmaraie (1992) also reported a reduction of calpastatin activity after
freezing longissimus dorsi steaks at -30°C, with subsequent improvements in tenderness over
fresh samples. The results of the present study support the theory of freezing and thawing
tenderising meat and are in agreement with Winger and Fennem
Jeremiah (1980) found that freezing increased the tenderness of pork chops. However, Smith
et al (1968) reported that the tenderness of lamb loin chops and lamb roasts decreased after
freezing, but in rib chops, tenderness increased. Other studies reported no effect of freezing
on tenderness or texture profiles of beef (Jeremiah et al, 1993).
Studies that have shown a decrease in tenderness following frozen storage (Jakobs
B
frozen storage. Such results are comparable with the F1 samples in the present study, which
showed quite low tenderness ratings. Despite the fact that thawing occurred over relatively
84
voursome than frozen samples (Table 4.3). Other studies
ave reported a decrease in flavour with frozen storage (Jakobsson and Bengtsson, 1973; and
he flavour profiles between frozen and chilled shabu-shabu beef (Jeremiah et al, 1993)
ly thawed steaks. Frozen beef was less juicy than aged beef, as seen by
remiah (1980). Thawing treatment had a significant effect on total loss, with repeated
short durations, F2 and F3 samples were both exposed to thawing at 20°C for a period of 3
hours. This temperature seems to have been sufficient to enable proteolysis to occur.
Beef flavour was not affected by repeated freezing and thawing treatment, however, aged
samples were significantly more fla
h
Wheeler et al., 1990). Whilst initial freezing of samples (F1) seemed to be detrimental to beef
flavour, no change in flavour was seen for aged beef samples as a result of repeated freezing
and thawing. A3 samples were similar in beef flavour ratings to A1 samples. Therefore, the
results suggest that aging beef prior to freezing improves flavour initially and that flavour
does not decrease due to freezing.
T
showed that samples which had been frozen and thawed had numerically higher values for a
bloody aromatic flavour, whilst the chilled product resulted in a higher incidence of a
browned aromatic flavour. In the present study, aged beef had significantly more foreign
flavour than frozen beef. Panellist’s comments were extremely varied for foreign flavour, and
it seems that some panellists rated aged flavour as a foreign flavour.
Aging and thawing produced significant effects on juiciness, with lower juiciness ratings for
aged and repeated
Je
freezing and thawing leading to increased drip loss, which was reflected in the juiciness
scores. Juiciness was significantly negatively correlated with cooking loss and total loss (-
0.18 and -0.16 respectively) indicating that as drip loss increased, juiciness decreased. The
85
e dehydrated
bres, whereas water which is unable to be absorbed accumulates in the extracellular space
awed samples, this process seemed to aid tenderisation.
weakness of the correlations, suggest that this did not play a large role in determining
juiciness ratings.
Aged beef lost significantly more drip throughout the trial, but had increased juiciness ratings.
One possible explanation for this anomaly, is that the correlation between juiciness with beef
flavour (0.40) and tenderness (0.41) was quite high. This suggests that if samples were more
flavoursome (as in the case of aged samples) they were more palatable and possibly
stimulated salivation, causing panellists to rate the samples as being more juicy.
In a review of frozen meat and meat products (Varnam and Sutherland, 1995), it was
suggested that during thawing, melting of ice in the extracellular space of muscle cells leads
to a migration of water through the sarcolemma membrane towards the still frozen
intracellular space. Water that reaches the intracellular space is reabsorbed by th
fi
and is ultimately lost as drip. The authors also suggested that the rate of thawing should be
matched to the freezing rate, otherwise large drip losses result due to the inability of water to
be reabsorbed into the fibre from extracellular spaces. For samples which were thawed in the
water bath at 20°C, (F2, F3, A2 and A3), the thawing rate would have been faster than the
freezing rate, which may explain the increased drip loss from these treatments.
Additionally, in slow frozen meat, the majority of denaturation takes place during the initial
freezing process rather than throughout storage (Varnam and Sutherland, 1995). In the
present study, the denaturation of muscle fibres would have occurred each time samples were
refrozen, thus reducing the ability of muscle cells to reabsorb drip on thawing. Despite the
increased drip from th
86
Acceptability of steaks was significantly higher for aged beef and followed the same trends as
0.54, 0.50 and 0.58
spectively) which suggested that all these attributes contributed equally to panellists opinion
of lubrication in the mouth.
gh the
tensity of red and yellow hues was reduced in the beef fat. The aged beef fat appeared grey
in contrast to the pinky yellow colour of frozen fat samples. The effect of thawing treatment
for tenderness ratings. There was little improvement in aged beef with thawing treatment.
However, for frozen beef, thawing significantly improved acceptability suggesting once
again that the effect of thawing on frozen steaks was one of aging. This aging effect was
associated with an increase in tenderness, flavour and juiciness, which all increased
acceptability ratings.
Acceptability was correlated with tenderness, juiciness and beef flavour, (
re
of acceptability (Table 4.4). Tenderness was correlated with juiciness (0.41) and beef flavour
(0.29), whilst beef flavour was correlated with juiciness (0.40). The positive correlation of
beef flavour and juiciness may be related to the release of flavours in juicier samples, whereas
dry samples tended to become powdery and less flavoursome. Additionally, dry samples may
have needed to be chewed more due to a lack
Gill and Badoni (1997) reported patchy bleaching of muscle tissue after immersion of beef in
85°C water, however the results of the present study indicate that the effect of repeated
freezing and thawing was not detrimental to meat colour. The temperature of the water thaw
(second thawing treatment) in the present study was only 20°C which would explain the less
extreme effect of treatment on meat colour.
Aging treatment had a more dramatic effect on both meat and fat colour. Aged meat had a
higher intensity of red and yellow hues in the muscle, than frozen samples, althou
in
87
on beef fat colour was significant for frozen samples only, with treatment increasing lightness
nd reducing redness of the fat such that values were similar to aged beef fat values. This is
beef
arcases in some countries (ie Korea) is normally carried out at much higher temperatures and
o whether fatty acid compostion is altered during storage
dicate that even at the extreme treatment of aging and with repeated freezing and thawing,
no significant change is seen to fatty acid composition when compared to freshly frozen
oduct. This indicates that in subsequent trials presented in this study, valid comparisons can
be made between the fatty acid composition and meat that was tested following
frozen storage for a period of time.
a
in agreement with Gill and Badoni (1997) who reported that heating of stained fat produced
dull brown tones as opposed to the more attractive pink or red tones of untreated fat.
4.5 CONCLUSIONS
Whilst these results indicate that the practice of repeated freezing and thawing on beef
increases tenderness and acceptability of frozen beef due to an aging effect, the practice is not
as good as aging fresh meat. If carried out under more extreme commercial situations, the
benefits to palatability may well be negated. The practice of thawing Australian
c
for a longer duration than what was used in this experiment. It is likely that under such
extreme thawing conditions, discolouration of the meat would occur at higher temperatures
(as indicated by Gill and Badoni, 1997) making it unsuitable for retail display. Additionally,
there may be an increased risk of microbial contamination associated with this practice.
The conclusions gained in relation t
in
pr
flavour of
88
CHARACTERISATION OF THE FLAVOUR OF
BEEF FROM THE NATIVE KOREAN BREED, THE
HANWOO, IN RELATION TO THE FLAVOUR OF
BEEF FROM AUSTRALIAN BREEDS
CHAPTER 5
89
s stated earlier, Korean consumers regard beef imported from Australia as a low quality
he quality of imported beef (ungraded high quality chilled
riploins from the USA) with those of Korean native cattle beef of quality grades B1, B2, B3
The aim of the present study was to document the differences in palatability attributes, in
particular flavour, between various Australia cattle genotypes and that of Hanwoo beef
imported into Australia from Korea, to determine which breeds most closely resemble the
Hanwoo in terms of flavour. It also aimed at ining the relationship between various
flavours and the intramuscular fat content.
5.1 INTRODUCTION
A
product when compared to beef produced in Korea and the USA. Their ultimate preference is
for the Korean native breed, Hanwoo (Ryu et al., 1994; CSIRO, 1995). One of the main
reasons for this preference is due to the perceived superior flavour of the Hanwoo beef, when
compared to imported beef by Korean consumers.
Ryu et al. (1994) compared t
st
in addition to out of grade (D grade) native cattle. It was found that Korean grade 1 beef had
higher intramuscular fat content and was more tender than the other Korean grades and
imported beef. Despite this, there were no significant differences reported in terms of aroma,
flavour or juiciness properties.
n
determ
90
5.2 MATERIALS AND METHODS
5.2.1 Selection of Animals for study
Beef striploins were collected from 170 animals, that were representative of the various
breeds produced in Australia. The animals were a subset of those of the Southern
Crossbreeding (SXB) and Davies Gene Mapping Projects (DGM) (see Chapter 3 – section
3.2.2). In the SXB project, seven sire breeds were used over Hereford dams. The breeds
included Belgian Blue (BH), Limousin (LH), South Devon (SH), Hereford (HH), Angus
(AH), Wagyu (WH) and Jersey (JH). Animals from the DGM project consisted of purebred
Limousins (LL), purebred Jerseys (JJ) and Limousin by Jersey crosses (LJ).
The animals were raised in three separate groups and slaughtered after 80 days (heifers) or
180 days (steers) on a grain ration. Steers were approximately 25 months of age at slaughter
(mean carcass weight 326kg), whereas heifers were slaughtered at 15 months of age (mean
carcass weight 218kg). The fourth group selected for this trial, were grain-fed in Korea and
striploin cuts were imported into Australia after slaughter. Groups 1 to 3 were electrically
stimulated post-slaughter, whereas group 4 was unlikely to have been electrically stimulated.
The animals thus consisted of :
1) 70 heifers selected from the Southern Crossbreeding Project (SXB)
2) 70 steers selected from the Southern Crossbreeding Project (SXB)
3) 30 steers selected from the Davies Gene Mapping Project (DGM)
4) 37 Korean Hanwoo (HAN) striploins imported into Australia (3 heifers in first shipment,
34 steers in second shipment)
91
In total, there were 207 samples, 19 breed by sex classes: SXB Females - AHF, BHF, HHF,
JHF, LHF, SHF, WHF; SXB Males - AHM, BHM, HHM, JHM, LHM, SHM, WHM; DGM
Males - LLM, LJM, JJM and Hanwoo animals – Hanwoo F and Hanwoo M.
5.2.2 Other Measurements
A comprehensive description of the taste panel methodology is presented in Chapter 3 -
section 3.6.2. The methodology for thawing, preparation and cooking of samples was similar
to that used in the preliminary trial. However, in addition to panellists rating samples on a 9-
point scale for initial juiciness and sustained juiciness (where 1 = extremely dry) and flavour
acceptability (where 1 = extremely unpleasant), they also rated the intensity of beef flavour,
beef fat flavour, oily flavour, buttery flavour, chicken flavour, corn flavour, grassy flavour
and rancid flavour (where 1 = not detectable and 9 = extremely strong). Sensory tenderness
was not rated, since flavour description was the main aim. Additionally, objective tenderness
measurements were taken on samples and considered sufficient to explain tenderness.
The same methodology was used to prepare and cook samples for objective measurements, as
was used for the preliminary trial, with the only differences being that tenderness was
measured using aWarner-Bratzler shear device fitted to an Lloyd Instron Testing machine
(Model LRX). Additionally, whilst pH was measured, meat colour and fat colour were not
measured. Muscle fat content (intramuscular fat percentage) was carried out on all samples
as described previously (Chapter 3 - section 3.3.1).
5.2.3 Statistical Analysis
Analysis of variance was carried out using the GLM procedure (SAS, 1996). A basic model
was developed to adjust the data for session, group and taster. The effect of session, in
92
addition to group within session was tested. Additionally, since taster name was not recorded,
taster was nested within session by group (Table 5.1). Residual flavours were calculated and
used for all subsequent analysis. Following this, the effect of sex (cohort), breed and breed by
sex group (19 combinations) on the individual flavours were tested, with (Model 2) and
without (Model 1) intramuscular fat content (IMF%) as a covariate. Additionally, Warner-
Bratzler shear force (WBSF) was added to the model, with (Model 3) and without (Model 4)
IMF%, to determine the effect of tenderness on the flavour assessments, particularly the effect
on flavour acceptability. Residual least squares means and standard errors for sex, breed and
sex by breed class for each of the flavours, were calculated from all models. Additionally, pH
and tenderness residual least square means were calculated using Model 1 and reported for all
19 breed by sex classes. For flavour acceptability, residual least square means were
calculated from the full model after adjusting for both IMF% and tenderness (WBSF).
Table 5.1 - Main effects and interactions tested in the basic model using the GLM procedure (SAS, 1996).
Source DF Session 6 Group within Session 28 Taster within (Session x Group) 140
The models were :
Model 1 : Residual flavour = Sex Breed Sex x Breed
Model 2 (IMF) : Residual flavour = IMF Sex Breed Sex x Breed
Model 3 (IMF, WBSF) : Residual flavour = IMF WBSF Sex Breed Sex x Breed
Model 4 (IMF, WBSF) : Residual flavour = WBSF Sex Breed Sex x Breed
Sire was fitted as a random effect in the model using PROC MIXED. Only the Australian
data set was used for this, since sire was unknown for the Hanwoo samples. Principal
components were also formed from the taste panel data using the PRINCOMP procedure
93
(SAS, 1996) and an analysis of variance was performed using PROC GLM, using Model 1
and 2. The effect of sire within breed sex class was significant for initial and sustained
juiciness (P<0.001) only. For all other flavours, sire was non-significant when tested using
PROC GLM (SAS, 1996). When fitted as a random effect in a mixed model (PROC MIXED,
SAS, 1996), the sire variance component converged to zero. As a consequence of these
analyses, sire was dropped from the model.
5.3 RESULTS
When the basic model (Table 5.1) was fitted to the flavour data, the effect of session, group
and taster was significant for all palatability attributes (apart from the effect of session on
chicken flavour). This model was therefore fitted to the data for all subsequent analyses to
ensure that significant adjustments were made to account for effects of session, group and
taster (Table 5.2). Table 5.3 shows results from subsequent models fitted to the flavour data.
Table 5.2 - Analysis of Variance Table for the Basic Modela
Flavour
R2
% Variation Accounted for (Type I Sums of Squares)
Total SS
Session Groupb Tasterc Initial Juiciness 39 5*** 8*** 19*** 3653 Sustained Juiciness 39 4*** 4*** 24*** 3505 Beef Flavour 52 3*** 8*** 38*** 3200 Beef Fat Flavour 51 3*** 4*** 42*** 3723 Oily Flavour 52 2*** 6*** 41*** 3151 Buttery Flavour 49 3*** 5*** 38*** 2804 Chicken Flavour 42 0.5 4*** 36*** 1861 Corn Flavour 46 1*** 6*** 38*** 1311 Grassy Flavour 51 2*** 5*** 43*** 1694 Rancid Flavour 38 1*** 8*** 29*** 760 Flavour Acceptability 36 1*** 6*** 26*** 2410
aBasic Model = flavours adjusted for session, group and taster bGroup nested within session c Taster nested within session by group
94
Table 5.3 - Analysis of Variance for the different models (1-4) fitted to the flavoursa
% Variation Accounted for (Type I SS) TOTALFLAVOUR Model R2 IMF WBSF Sex Breed Breed x Sex SS Initial Juiciness 1 20 2 17*** 2 283 2 24 14*** 0 9* 1 283 3 23 11*** 0 0 0* 1 264 4 20 0 1 17*** 2 264 Sustained Juiciness 1 23 3** 16*** 4 227 2 27 14*** 0 9* 3 227 3 27 12*** 1 0 10** 4 214 4 23 0 3* 16*** 5 214 Beef Flavour 1 15 0 10* 4 152 2 15 1 0 10* 4 152 3 17 1 0 0 11* 4 141 4 16 0 0 11* 5 141 Beef Fat Flavour 1 15 2 9* 4 143 2 18 9*** 0 6 4 143 3 17 7*** 0 0 7 3 134 4 13 1 1 7 4 134 Oily Flavour 1 18 1 11** 7* 129 2 21 9*** 0 5 7* 129 3 18 6*** 0 0 5 7 117 4 15 1 0 7 6 117 Buttery Flavour 1 19 2* 10* 6 124 2 20 9 0 5 6* 124 3 18 6 0 0 5 6 112 4 16 0 1 8 6 112 Chicken Flavour 1 13 1 8 5 62 2 13 3* 0 6 5 62 3 13 1 0 0 6 6 54 4 13 0 0 7 6 54 Corn Flavour 1 15 0 9* 5 46 2 15 1 0 9* 5 46 3 16 1 1 0 8 6 42 4 15 1 0 9 6 42 Grassy Flavour 1 5 1 2 2 54 2 7 0 1 3 2 54 3 7 0 1 1 2 2 49 4 7 1 1 3 2 49 Rancid Flavour 1 9 0 6 3 34 2 9 0 0 6 3 34 3 10 0 1 0 6 3 33 4 10 1 0 6 3 33 Flavour 1 19 1 16*** 3 116Acceptability 2 21 4** 0 14*** 2 116 3 24 4** 5*** 0 13*** 2 115 4 22 2* 1 17*** 3 115
aflavours adjusted for session, group and taster
95
When the effects of sex, breed and the two-way interaction between breed and sex (Model 1)
were tested using PROC GLM (SAS, 1996), the effect of sex (cohort group/weight/age) on
flavour was only significant for sustained juiciness (P<0.01) and buttery flavour (P<0.05),
(Table 5.3). Steers had higher scores than heifers (4.7 compared to 4.58 for sustained juiciness
and 2.53 compared with 2.33 for buttery flavour). Breed was significant (P<0.05) for all
flavours apart from chicken, grassy and rancid flavours (Figure 5.1-5.5, and Figure 5.7).
Belgian Blue sired calves (BH) had the lowest value, 3.91 (Figure 5.1), recorded for initial
juiciness. Their scores were not significantly different from LL, JH, LH, LJ or SH, but
significantly different from WH, HH, JJ, Hanwoo and AH. AH animals had the highest
(P<0.05) initial juiciness score, 5.44, which was not significantly different from the Hanwoo,
or JJ. When initial juiciness was adjusted for IMF%, BH steers still had the lowest score
(4.19). Initial juiciness scores for Hanwoo animals dropped down from 5.37 to 4.83 as a result
of the adjustment for intramuscular fat. AH animals on the other hand, maintained the highest
score for initial juiciness (5.56), which was not significantly different to Hanwoo or JJ.
Similar trends were seen for sustained juiciness, with BH steers having the lowest score of
4.16, which was not significantly different from LL, LH, JH, SH or WH. AH animals had the
highest score of 5.51, which was significantly different to all other breeds (P<0.05) apart from
JJ, Hanwoo and LJ. After adjusting for intramuscular fat, the Hanwoo dropped considerably
in sustained juiciness score from 5.42 to 4.95. Despite this, the ranking’s at the top end of the
scores for sustained juiciness remained the same with AH steers having the highest score
(5.61), which was not significantly different from JJ (5.33), Hanwoo (4.95) and LJ (4.78).
BH maintained the lowest score for sustained juiciness after adjusting for IMF%, with a score
of 4.4, which was not different to other breeds, apart from JJ (P<0.05) and AH (P<0.01).
96
Figure 5.1 - Breed LSMEANS for Initial Juiciness
1
2
3
4
5
6
BH LL JH LH LJ SH WH
HH JJ
Han
woo A
H
model 1 model 2 (IMF) model 3 (IMF, WB)
Calculated from Models 1 (P<0.001), 2 (P<0.05), & 3 (P<0.05)
Figure 5.2 - Breed LSMEANS for Sustained Juiciness
1
2
3
4
5
6
BH LL LH JH SH WH
HH LJ
Han
woo JJ A
H
model 1 model 2 (IMF) model 3 (IMF, WB)
Calculated from Models 1 (P<0.001), 2 (P<0.05), & 3 (P<0.01)
LJ animals had the lowest beef flavour intensity of 4.04, which was not significantly different
to any of the breeds apart from WH, AH and JJ. The highest beef flavour score, of 5.14, was
recorded for the JJ steers, followed by AH (4.86) and WH (4.74). When the scores were
97
adjusted for intramuscular fat, significant breed differences were still seen for beef flavour,
with JJ still having the highest score of 5.09, which wasn’t significantly different to AH, WH,
JH or LH. Hanwoo animals dropped from 4.15 to 3.99 for beef flavour, making it the lowest
ranked breed for beef flavour. This was only significantly different to WH, AH (P<0.05) and
JJ (P<0.01). LJ remained fairly constant at 4.01. The rankings did not change for the top
scores for beef flavour, with JJ having the highest score of 5.09 (adjusted down for IMF%)
and AH and WH remaining fairly constant.
Figure 5.3 - Breed LSMEANS for Beef Flavour
1
2
3
4
5
6
LJ
Han
woo BH
HH LL SH LH JH WH AH JJ
model 1 model 2 (IMF) model 3 (IMF, WB)
Calculated from Models 1 (P<0.05), 2 (P<0.05), & 3 (P<0.05)
Although not significant, chicken flavour was highest for Hanwoo animals, with a score of
2.40 (±0.31) for the Hanwoo heifers and 2.09 (±0.09) for the Hanwoo steers (Table 5.4). HH
steers, WH heifers and SH steers had the lowest scores for chicken flavour of 1.48, 1.54 and
1.54 (±0.17) respectively.
98
AH animals had the highest score for corn flavour (1.85) and JJ animals had the lowest score
(1.34). This became non-significant when adjusted for intramuscular fat, although the ranking
of breeds did not change. Grassy flavour and rancid flavour were non-significant for breed,
both before and after adjusting for IMF%, suggesting that these flavours were either not
detectable in the samples tested, or alternatively, the panel may have had difficulty in scoring
these flavours as was indicated by the high frequency of scores of 1 (not detectable) recorded
for these flavours.
Figure 5.4 - Breed LSMEANS for Corn Flavour
1
2
JJ JH HH LL BH
Han
woo SH LJ W
H LH AH
model 1 model 2 (IMF)
Calculated from Model 1 (P<0.05) and 2 (P<0.05)
LL animals had the lowest beef fat flavour score of 2.94, which was significantly different to
Hanwoo, LH, HH, WH, AH and JJ animals. JJ and AH steers had the highest beef fat flavour
scores of 3.9 and 3.71 respectively. JJ animals were significantly different to JH, LJ, BH
(P<0.05) and LL (P<0.01).
99
Breed was significantly different (P<0.05) for oily flavour and buttery flavour (Figure 5.5). JJ
had the highest score for oily flavour, however this was not significantly different to any other
breed apart from LL (P<0.05). LL recorded the lowest score for oily flavour which was
significantly different to the Hanwoo (P<0.01), LH, HH, WH, AH and JJ (P<0.05).
AH animals had the highest score of 2.86 for buttery flavour, which was not significantly
different from JJ, WH, JH, Hanwoo or LJ (2.66, 2.60, 2.49, 2.48 and 2.44 respectively). SH
had the lowest score for buttery flavour (2.19) which was not significantly different from any
other breed apart from AH (P<0.01).
It should be noted however, after adjusting for intramuscular fat percentage, no breed
differences were apparent for beef fat flavour, oily flavour or buttery flavour, suggesting that
these flavour scores reflected differences in fat content (Table 5.5).
Figure 5.5 - Breed LSMEANS for Beef Fat Flavour (P<0.05), Oily Flavour (P<0.01) and Buttery Flavour (P<0.05)
1
2
3
4
LL BH LJ JH SH
Han
woo LH H
H
WH AH JJ
Beef Fat Flavour Oily Flavour Buttery Flavour
Calculated from Model 1
Table 5.4 - Least Square Means for IMFb, Tendernessc, pHc and individual Flavoursc
BS IMF%b
(P<0.001)WBSFc
(P<0.001) pHc
NS IJUICEc
NS SJUICEc
NS BEEFc
NS BFATc
NS OILYc
(P<0.05) BUTTERYc
NS CHICKENc
NS CORNc
NS GRASSYc
NS RANCIDc
NS SXB Heifers
AHFa 3.50 3.72
5.67 5.12 5.28 4.64 3.38 2.76 2.82 1.68 1.84 1.64 1.20BHFa 2.65 3.44 5.54 3.88 4.06 4.34 3.18 2.24 2.20 1.84 1.34 1.68 1.26HHFa 3.52 3.65 5.70 4.98 4.86 4.16 3.66 3.02 2.70 2.00 1.54 1.60 1.52JHFa 3.33 3.39 5.52 4.02 3.98 4.02 3.08 2.62 2.48 1.82 1.52 1.66 1.10LHFa 2.94 3.81 5.61 4.38 4.70 4.48 3.38 2.50 2.22 1.60 1.78 1.70 1.16SHFa 3.68 3.54 5.72 4.32 4.42 4.26 3.18 2.24 2.08 1.70 1.46 1.74 1.38WHFa 3.91 3.46 5.65 4.56 4.40 4.96 3.52 2.58 2.26 1.54 1.46 1.68 1.26±SE 0.64 0.21 0.05 0.35 0.30 0.26 0.25 0.24 0.23 0.17 0.14 0.17 0.13
SXB Steers AHMa 5.17 3.33
5.51 5.76 5.74 5.08 4.04 2.96 2.90 2.00 1.86 1.80 1.18BHMa 3.19 3.27 5.52 3.94 4.26 4.00 3.02 2.52 2.42 1.96 1.70 1.84 1.30HHMa 4.66 3.42 5.56 4.50 4.68 4.18 3.22 2.30 2.02 1.48 1.32 1.76 1.16JHMa 5.89 3.16 5.52 4.56 5.00 4.72 3.42 2.40 2.50 1.72 1.28 1.58 1.26LHMa 2.95 2.98 5.58 4.24 4.08 4.22 3.42 2.36 2.30 1.92 1.54 1.82 1.30SHMa 5.03 3.09 5.57 4.70 4.76 4.40 3.60 2.58 2.30 1.54 1.66 1.52 1.40WHMa 4.74 2.90 5.53 4.90 5.08 4.52 3.54 2.92 2.94 1.74 1.70 1.92 1.22
±SE 0.64 0.21 0.05 0.35 0.30 0.26 0.25 0.24 0.23 0.17 0.14 0.17 0.13DGM Steers
JJMa 6.80 3.09
5.49 5.12 5.48 5.14 3.90 2.88 2.66 1.74 1.34 1.52 1.04LJMa 6.16 2.83 5.47 4.48 4.86 4.04 3.18 2.38 2.44 1.88 1.58 1.80 1.32LLMa 3.36 3.10 5.44 4.14 4.36 4.24 2.94 2.12 2.30 1.92 1.50 1.62 1.52±SE 0.64 0.21 0.05 0.35 0.30 0.26 0.25 0.24 0.23 0.17 0.14 0.17 0.13
Hanwoo F 9.43 3.55 5.50 5.40 5.53 3.80 2.93 1.87 1.93 2.40 1.33 1.33 1.13±SE 1.17 0.39 0.10 0.63 0.55 0.48 0.46 0.43 0.42 0.31 0.26 0.30 0.24
Hanwoo M 10 60 4 62 5 47 5 35 5 32 4 49 3 86 3 13 3 03 2 09 1 78 1 82 1 34±SE 0.35 0.13 0.03 0.19 0.16 0.14 0.14 0.13 0.13 0.09 0.08 0.09 0.07
aA=Angus, B=Belgian Blue, H=Hereford, J=Jersey, L=Limousin, S=South Devon, W=Wagyu, F=Female, M=Male. bModel : IMF = breed by sex group c Model 1 : Flavour, Tenderness, pH = sex, breed, sex by breed – adjusted for session, group nested within session, taster nested within group by session (ANOVA – Table 5.3)
100
101
Table 5.5 - Least Squares Means for Flavours which were significant for breed by sex class for Model 2 (IMF% fitted as covariate).
BS OILYb BUTTERYb FACCEPTb FACCEPTc
SXB Heifers AHFa 2.88 2.91 5.88 5.93 BHFa 2.42 2.34 5.44 5.44 HHFa 3.14 2.79 5.44 5.47 JHFa 2.76 2.58 5.65 5.63 LHFa 2.66 2.34 5.65 5.71 SHFa 2.35 2.16 5.41 5.42 WHFa 2.68 2.33 5.82 5.82 ± SE 0.24 0.24 0.23 0.23
SXB Steers AHMa 2.97 2.91 6.11 6.07 BHMa 2.66 2.53 5.08 5.04 HHMa 2.35 2.05 5.43 5.41 JHMa 2.36 2.47 6.10 6.01 LHMa 2.52 2.42 5.65 5.55 SHMa 2.60 2.32 5.75 5.67 WHMa 2.96 2.97 5.87 5.74 ± SE 0.24 0.24 0.22 0.22
DGM Steers JJMa 2.78 2.59 6.43 6.34 LJMa 2.32 2.40 5.38 5.24 LLMa 2.25 2.40 5.43 5.35 ± SE 0.24 0.24 0.23 0.23
Hanwoo F 1.59 1.72 5.88 5.86 ± SE 0.44 0.43 0.42 0.42
Hanwoo M 2.77 2.76 5.44 5.57 ± SE 0.19 0.19 0.18 0.22
aA=Angus, B=Belgian Blue, H=Hereford, J=Jersey, L=Limousin, S=South Devon, W=Wagyu, F=Female, M=Male
b Model 2 : IMF, Sex, Breed, Sex by Breed (adjusted for session, group and taster) c Model 3 : IMF, WBSF, Sex, Breed, Sex by Breed (adjusted for session, group and taster)
As shown in Table 5.4 and Figure 5.8, steers had significantly (P<0.001) more intramuscular
fat (5.32%) than heifers (4.70%) due to being fattened to a greater degree in the feedlot.
Figure 5.6 shows that the Hanwoo animals had the highest level of IMF% (10.50%), which
was significantly different (P<0.001) from all other breeds in the study. Of the Australian
breeds, JJ steers had an intramuscular fat content (6.80%), which was significantly different
(P<0.01) from all other breeds apart from LJ. The breed with the lowest intramuscular fat
102
content was BH (2.90%), which was only significantly different to JH (P<0.05), LJ, JJ and
Hanwoo (P<0.001).
Figure 5.6 - Breed least square means for Intra-muscular Fat percentage
123456789
1011
BH LH LL HH
WH AH
SH JH LJ JJ
Han
Intr
amus
cula
r fat
(IM
F%)
Calculated from Model 1 : IMF = sex, breed, sex by breed
Table 5.6 - Estimate of the Slope for IMF%
ESTIMATE IMF%
± SE IMF% Significance
IJUICE 0.128 0.016 *** SJUICE 0.112 0.016 *** BEEF 0.019 0.013 NS BEEF FAT 0.078 0.014 *** OILY 0.073 0.013 *** BUTTERY 0.079 0.013 *** CHICKEN 0.036 0.011 ** CORN 0.012 0.009 NS GRASSY -0.003 0.010 NS RANCID 0.002 0.007 NS FACCEPT 0.044 0.013 **
** P<0.01, *** P<0.001
Solutions for intramuscular fat content were calculated and the estimate of IMF% (slope) is
shown in Table 5.6. Intramuscular fat was significant for initial and sustained juiciness, beef
fat flavour, oily flavour, buttery flavour, chicken flavour and flavour acceptability. For every
percentage increase in IMF%, these flavours also increase by the amount (estimate) shown.
103
When IMF% was fitted as a covariate to the flavour data, as shown in the analysis of variance
(Table 5.3), none of the flavours were significantly different between steers and heifers, and
breed group was only significant (P<0.05) for initial juiciness, sustained juiciness, beef
flavour, corn flavour and flavour acceptability. (Figure 5.1, Figure 5.2, Figure 5.3, Figure 5.4
and Figure 5.7). This suggests that for these flavours, other factors besides fatness were
accounting for differences in flavour. Figure 5.5 shows that only oily and buttery flavour,
were significant for the two way interaction of breed by sex, after adjusting for IMF%.
Figure 5.7 - Breed LSMEANS for Flavour Acceptability
1
2
3
4
5
6
7
BH LL HH LJ LH SH WH JH
Han
woo A
H JJ
model 1 model 2 (IMF) model 3 (IMF, WB)
Calculated from Models 1 (P<0.001), 2 (P<0.001), & 3 (P<0.01)
JJ animals achieved the highest scores for flavour acceptability of 6.5, which was significantly
higher than all other breeds, whilst BH had the lowest flavour acceptability of 5.15 (which
was only significantly different (P<0.01) to WH, JH, Hanwoo, AH and JJ). After adjusting for
intramuscular fat content, JJ maintained the highest score for flavour acceptability (6.43),
however it was no longer significantly different to AH or JH animals (5.99 and 5.87
104
respectively). Hanwoo animals achieved a high flavour acceptability score of 5.66, however
this was still significantly lower (P<0.001) than the JJ animals (6.43).
Figure 5.8 shows that, in general, as intramuscular fat content increases, flavour acceptability
is improved. However, the results also demonstrate that intramuscular fat content was not the
only factor determining flavour acceptability, since breed group differences were still seen
when the data was adjusted to the same level of intramuscular fat content, particularly for the
Hanwoo breed. Additionally, it was thought that tenderness may have influenced the
panellist’s judgment of flavour acceptability, since this attribute was not scored.
Figure 5.8 - Relationship between breed-sex class least square means for bIMF% and cFlavour Acceptability
aA=Angus, B=Belgian Blue, H=Hereford, J=Jersey, L=Limousin, S=South Devon, W=Wagyu, F=Female, M=Male. bModel : IMF = breed by sex group cModel : Flavour = sex, breed, sex by breed – adjusted for session, group nested
within session, taster nested within group by session (ANOVA – Table 5.3)
WHHHLH
BH
JHAH
SHWH
HHSH
JH
AHLH
BH
JJ
LJ
LL
HAN
4.6
4.8
5.0
5.2
5.4
5.6
5.8
6.0
6.2
6.4
6.6
1 3 5 7 9Intramuscular Fat %
Flav
our A
ccep
tabi
lity
Steers SXB Heifers SXBSteers DGM Hanwoo
11
105
Tenderness, (Warner-Bratzler shear force) was not significant for sex, however was
significant for breed (P<0.001) and significant (P<0.05) for breed by sex interaction (Table
5.4). LJ steers were the most tender, with an average shear force value of 2.83, which
compared to the less tender samples of the Hanwoo steers and heifers (4.62 and 3.55
respectively. Despite this, all samples were classed as tender (since no values were above 5kg
shear force). When included in the model, WBSF was not significant for any of the flavours
apart from flavour acceptability. For this reason, it was included in the final model (along
with IMF%, sex, breed and sex*breed) to adjust flavour acceptability scores to account for
differences in both IMF% and tenderness. Despite these corrections, there were still
significant differences (P<0.01) between breeds for flavour acceptability (Figure 5.7), but not
between breed by sex class. Figure 5.7 and Table 5.5 show the residual least squares means
for flavour acceptability for breed and breed by sex class, for the three models fitted to the
data.
When pH was fitted to the model, there were significant differences (P<0.001) in pH between
sexes. Heifers had a significantly higher pH (5.61±0.05), than steers (5.52±0.05). There were
no significant differences in pH between breeds, or breed by sex class and it was therefore not
included in the final model. Residual least square means for the 19 breed by sex classes can
be seen in Table 5.4.
When principal components were formed between the flavours, (Table 5.7), it became
apparent that the first three principal components accounted for 70% of the variation in
flavour.
106
Table 5.7 - Principal Component Eigenvectors and variation accounted for
Prin1 Prin2 Prin3 Prin4 Prin5 Prin6 Prin7 Prin8 Prin9 Prin10Ijuice 0.60 -0.38 0.17 -0.22 0.23 -0.19 -0.30 0.17 0.39 0.24 Sjuice 0.52 -0.42 -0.03 -0.02 -0.26 0.19 0.25 -0.15 -0.49 -0.35 Beef 0.25 0.35 -0.64 -0.16 0.47 0.09 -0.15 0.22 -0.23 -0.14 Bfat 0.30 0.47 -0.08 -0.34 -0.57 0.02 0.30 0.20 0.34 0.00 Oily 0.26 0.47 0.34 0.07 -0.13 0.07 -0.60 -0.39 -0.20 -0.07 Buttery 0.26 0.33 0.36 0.32 0.33 -0.46 0.47 0.07 -0.17 0.03 Chicken 0.06 0.03 0.22 0.45 -0.08 0.49 -0.15 0.66 -0.02 0.00 Corn 0.04 0.05 0.16 0.05 0.35 0.33 0.17 -0.20 0.51 -0.62 Grassy 0.04 0.09 0.17 -0.28 0.25 0.57 0.27 -0.29 -0.07 0.53 Rancid 0.00 0.01 0.19 -0.18 0.11 0.13 0.10 0.26 -0.24 0.12 Faccept 0.26 -0.01 -0.42 0.62 -0.09 0.10 0.09 -0.29 0.23 0.34 Cumulative % 46 59 70 77 82 87 90 93 96 98
aPrincipal components were calculated from flavours (adjusted for session, group nested within session, taster nested within group by session)
Principal component 1 primarily reflected juiciness and to a lesser extent beef fat flavours.
(As shown in Table 5.7, numbers in bold indicate the flavours that are contributing the most
to each principal component). When an analysis of variance was performed (Table 5.8), sex
and breed differences in principal component 1 were significant (P<0.01 and P<0.001,
respectively). When IMF% was added as a covariate, only breed remained significantly
different for principal component 1. Although principal component 2 appeared to reflect
fatness (beef fat, oily and buttery flavours), sex and breed differences were not significant.
Principal component 3 was significant for breed (P<0.01) and the two-way breed by sex
interaction (P<0.05). This component reflects oily and buttery flavours, and is strongly
negatively related to beef flavour and flavour acceptability. When IMF% was added as a
covariate, only breed remained significantly different for principal component 3.
Table 5.8 - Analysis of Variance Table for Breed Sex Class for Principal Components
% Variation Accounted for (Type I SS) IMF% Sex Breed Sex by Breed Total SS R2
Principal Component 1 3** 21*** 3 636 0.32 17* 0 12** 3 636 0.32Principal Component 3 0 11** 8* 155 0.19 2 0 10* 8 155 0.19
107
Figure 5.9 and Figure 5.10 show the breed least squares means for principal components 1
and 3. It can be seen that WH, Hanwoo, JJ and AH are all positive for principal component 1,
which would suggest that the meat from these animals were juicier, had more beef fat flavour
and had a more desirable flavour (flavour acceptability) than the other breeds, which were
negative for principal component 1. When intramuscular fat was fitted to the model, Hanwoo
animals dropped to a negative value, probably due to being confounded with fat content.
Figure 5.9 - Breed LSMEANS for PRIN1 (Calculated using Model 1 and 2)
-2
-1
0
1
2
B H L L L H J H S H L J H H W H H a n J J A H
M o d e l 1 M o d e l 2 ( IM F )
For principal component 3, BH, Hanwoo, HH and LJ were all positive for principal
component 3, which would suggest that these breeds have low scores for beef flavour and
acceptability and show oily and buttery characteristics. Conversely, JJ was negative for this
principal component and was significantly lower than all other breeds in the study. This
would suggest that this breed is highly related to beef flavour and acceptability and has less
oily and buttery characteristics than the other breeds in the study.
108
Figure 5.10 - Breed LSMEANS for PRIN3 (Calculated using Model 1 and 2)
-1
0
1
JJ W H JH S H L H A H L L B H H a n H H L J
M o d e l 1 M o d e l 2 (IM F )
Residual correlations were calculated from the model containing sex, breed, and sex by breed
(flavours were already adjusted for session, group and taster). Table 5.9 shows that many of
the flavours were significantly correlated with each other (P<0.05). The highest correlation
was between initial juiciness and sustained juiciness (0.77), which is not surprising since both
measure similar properties. Initial juiciness is normally defined as the amount of juice
released by the product on mastication, whereas sustained juiciness is often perceived as the
ability of the product to stimulate salivation, thereby increasing the amount of fluid in the
mouth. Sustained juiciness tends to reflect fat content and may include the mouth-feels of
oiliness (liquid fat) and greasiness (solid fat). The significant (P<0.001) correlation between
buttery and oily flavour was 0.50, indicating that these two flavours may be reflecting the
same type of flavour or texture.
Flavour acceptability was significantly positively correlated with initial and sustained
juiciness, beef flavour, beef fat flavour, oily flavour, buttery flavour and chicken flavour and
negatively correlated with rancid flavour (see Table 5.9).
Table 5.9 - Residual Correlationsa between Flavours
SJUICE BEEF BFAT OILY BUTTERY CHICKEN CORN GRASSY RANCID FACCEPT IMF pH WBSFIJUICE
0.77 ***
0.27 ***
0.30 ***
0.31 ***
0.37 ***
0.39 ***
0.37 ***
SJUICE
0.24 ***
0.32 ***
0.21 **
0.25 ***
0.50***
0.38 ***
BEEF
0.42 ***
0.21 **
0.19 **
0.43***
BFAT
0.49 ***
0.36 ***
0.28***
0.30 ***
OILY
0.50 ***
0.24 ***
0.20**
0.18**
0.30 ***
BUTTERY
0.18 *
0.17 *
0.24***
0.30 ***
CHICKEN
0.19**
0.16 *
CORN
0.28 ***
GRASSY
0.27 ***
RANCID
-0.26 ***
FACCEPT
0.19 **
-0.16*
IMF
-0.18 *
0.31 ***
aResidual correlations were calculated from the Model : Sex, Breed, Sex by Breed (adjusted for session, group nested within session, and taster nested within group by session)
109
110
5.4 DISCUSSION
In general, the leaner European breeds and some heifers (those with lower intramuscular fat)
were less juicy than the more marbled breeds, in particular the longer-fed steers and the
Hanwoo animals.
Steers had significantly more intramuscular fat content than heifers (P<0.05) with the
exception of LH, BH and LL steers. This was to be expected since steers were fed on grain
for a longer period (180 vs 80 days) and were more mature (25 vs 15 months). This is in
agreement with Harrison et al. (1978) who found that flavour became more desirable as
feeding period increased. Overall, the steers performed better than the heifers possibly due to
the longer period on grain. This is in agreement with Westerling and Hedrick (1979), who
showed that meat from animals fed a concentrated diet for 112 days was more desirable in
flavour than animals fed for 56 days. The study by Westerling and Hedrick, (1979) also
indicated that the increase in flavour was paralleled by an increase in marbling score.
It was interesting to note that the breed with the highest flavour acceptability was the purebred
Jersey (JJ), which is normally considered unsuitable for beef production in Australia because
of poor visual appearance of the meat, poor carcase conformation and low meat yields.
Despite this, the Jersey may be useful in crossbreeding systems to improve the flavour
acceptability of leaner well-muscled European breeds, such as the Belgian Blue and Limousin
crosses which scored very poorly for most flavours recorded in this study.
There was little evidence from this Australian taste panel, to support the theory that Wagyu
animals produce superior quality meat to other breeds as previously suggested in the literature
111
(Busboom et al., 1993). Despite this, Wagyu cross animals were intermediate for flavour
acceptability when compared with other breeds, and still had above average scores. The fact
that they did not resemble the Hanwoo in terms of taste and texture indicates that the Hereford
component in the cross may have been influencing the flavour, and/or the environmental and
nutritional conditions in Australia were not sufficient to produce the superior quality of beef
produced in Japan and America.
The Hanwoo animals had the highest level of intramuscular fat in the study (10.5%), which
was significantly different (P<0.001) from all other breeds in the study. The Hanwoo was
lower in flavour acceptability than many of the other breeds (Figure 5.8), despite the high
level of fat. This is an indication that at this high level of fat, the flavour was unacceptable to
Australian panellists, who are used to leaner beef and different flavours. Hanwoo samples
also had the lowest intensity of beef flavour, after adjusting for IMF%, and a high intensity of
chicken flavour. This type of flavour profile would be quite unusual to an Australian panel,
and hence for flavour acceptability it was rated lower than many of the Australian breeds.
Flavour acceptability was more highly correlated with beef flavour (0.43, Table 5.9) and with
beef fat flavour (0.28), than with chicken flavour (0.19). Flavour acceptability was also
moderately correlated with initial and sustained juiciness (0.39 and 0.50 respectively) and
negatively correlated with rancid flavours (-0.26).
It is apparent from the study presented in this chapter, that even after adjusting for tenderness,
flavour acceptability increases as intramuscular fat content increases. However, at levels of
IMF% above 10%, flavour acceptability to an Australian panel is reduced. Despite this, if the
same set of animals were tested using a Japanese or Korean panel, quite different results may
112
be obtained, since they are more accustomed to, and prefer beef with high levels of
intramuscular fat.
It was noted that the Korean animals were significantly tougher than the Australian animals.
This could be due to a number of reasons including the fact that Asians prefer their meat
sliced thinly to eat and therefore little emphasis would have been placed on selecting for
tender beef. Rhee and Kim (2001) state that “Hanwoo has been regarded as a premium beef in
Korea because of its high palatability and desirable chewiness”, which suggests that the
Korean palate prefers a more textured meat. Alternatively, it is unknown whether the animals
in Korea were electrically stimulated or not, whereas the Australian animals were all
stimulated at slaughter which would have had the effect of increasing tenderness.
The fact that beef fat flavour, became non-significant for breed after adjustment for IMF%
indicates that these flavours are heavily influenced by the amount of intramuscular fat within
the meat. It is not surprising that both beef fat flavour and oily flavour reflected IMF% since
they are both characterised by oily and fatty mouth-feels as well as the flavour of beef fat.
Additionally in this study, buttery taste seems to be related to a creamy fatty texture.
In the initial training sessions, it was noted that the Hanwoo beef had a characteristic chicken
skin flavour and had a low intensity of beef flavour. It is disappointing that this wasn’t more
prevalent in the results of this trial. The Hanwoo samples were much higher in intramuscular
fat content (5.3% to 16.5%) than the Australian breeds. This raises the question of whether
intramuscular fat content and chicken skin flavour are truly related or whether they are just
confounded. Table 5.9 shows that chicken flavour was significantly correlated with IMF%
(P<0.001), however the correlation was only 0.16. Chicken flavour was significantly
113
(P<0.05) correlated with buttery flavour and oily flavour (0.24 and 0.18 respectively), which
suggests that panellist’s may have been scoring these flavours similarly, although the
correlations are low.
Another point to note is that the first shipment of Hanwoo samples (heifers) had the highest
level of chicken skin flavour, whereas the second shipment, although still having the second
highest score for chicken skin flavour, weren’t as high as the first shipment. The fat content
of the meat was also higher in the first shipment than the second. Additionally, these
differences may have been shown to be significant, had there been more samples supplied by
ELDERS Limited from Korea.
Another factor was that the taste panel only had minimal training and this may have resulted
in less confidence in the scores for the flavours being recorded. Initial juiciness, sustained
juiciness, beef flavour and flavour acceptability scores were all normally distributed, whereas
beef fat flavour, oily flavour, buttery flavour, chicken flavour, corn flavour, grassy and rancid
flavour scores were all skewed towards 1 (not detectable). This is an indication that panellists
found these flavours harder to score having the effect of pulling down the averages for these
particular flavours and preventing real differences being detected.
Currently, with the trend for high levels of marbling for beef destined for the Japanese market,
Australian producers are placing increasing selection pressure on animals with superior
marbling ability. Many studies have shown a lack of evidence for the benefit of marbling to
palatability (Goll et al., 1965; Kregel et al., 1986; Crouse et al., 1989; Wheeler et al., 1994;
Rymill et al., 1997). Other studies, however, have reported marbling score is highly
correlated with flavour (Smith et al., 1983; Berry et al., 1980), which is in agreement with the
114
study presented in this chapter. Table 5.9 shows correlations between intramuscular fat
content and the various flavours scored. IMF% was significantly correlated (P<0.001) with
initial juiciness, sustained juiciness, beef fat, oily flavour, buttery flavour, chicken flavour and
flavour acceptability. Despite this, the correlations are quite low, ranging from 0.16 for
chicken flavour to 0.38 and 0.37 for sustained and initial juiciness, respectively.
Other studies that have shown that fat levels were highly related to meat flavour include those
of Beilken et al., 1990 and Dolezal et al., 1982. Dolezal et al. (1982) reported that steaks
from carcasses with at least 5mm of fat were superior to steaks from carcasses with less than
this amount of fat, however steaks with greater than 7mm of fat did not further improve
cooked beef palatability. This shows similarities to the present study, where the trend
indicated that as intramuscular fat increased, so did flavour acceptability, However, when fat
content was above 10%, flavour acceptability declined.
Berry et al., (1980) reported that higher mean marbling scores were also classed as having
more desirable flavours, characterised by sweet and browned flavours. It can be seen from
our results that in general, the breeds with higher fat contents, in particular the long fed steers,
were characterised by higher buttery flavour scores than the leaner breeds. This is supported
by the fact that the correlation between IMF% and buttery flavour was 0.30 (P<0.001).
Overall, all of the analyses performed in this chapter support the influence of intramuscular
fat on juiciness, beef, oily and buttery flavour and to a lesser extent chicken flavour and
flavour acceptability. The next chapter will examine whether other factors such as the fatty
acid composition can further explain the various flavours and differences seen in flavours
between breed.
115
5.5 CONCLUSIONS
It appears that whilst flavour acceptability is positively enhanced by increased levels of
intramuscular fat, it is not the sole determinant of flavour acceptability. Breed groups were
significantly different for juiciness, beef flavour, buttery flavour and flavour acceptability,
even after adjusting data to a constant level of intramuscular fat, suggesting that some of the
variation in flavour may be genetic. Both the Angus Hereford steers and heifers performed
well for a majority of the flavour attributes, both before and after adjusting for IMF% and
since they were fed the same diets as the other breeds in their respective groups, this indicates
some evidence for genetic differences in flavour.
The Hanwoo displayed some unusual flavour characteristics, which were different to those of
the Australian breeds, most notably a numerically higher intensity of chicken score and lower
intensity of beef flavour. Unfortunately, this was not demonstrated clearly in the statistical
analysis possibly due to the confounding with the Hanwoo breed having such a high level of
IMF% in comparison to the Australian breeds, which may have masked some of the true
differences between the Korean and Australian breeds. After adjusting for IMF%, the
Hanwoo animals rankings often changed from being amongst the top scoring breeds for a
particular flavour, to being not significantly different from the lowest scoring breed. Chapter
6 will look at the effect of fatty acid composition on the flavour, in particular to determine
whether the differences in flavour between the Hanwoo and Australian animals can be
explained by the fatty acid composition, in addition to the amount of intramuscular fat.
117
in Australia have been
entified as having superior marbling ability, although, not to the extent of that recorded in
the high levels of
onounsaturated fatty acids (MUFA’s), achieved by Japanese (Sturdivant et al, 1992) and
in Japan. Consequently, it is believed
at there may also be an effect of feeding regime (management and ration composition) on
termine the relationship between fatty acid composition
ze the fatty acid profile of both the
ustralian breeds and the Hanwoo.
6.1 INTRODUCTION
Currently, Australian beef does not meet the specifications for premium markets in Asian
countries (in particular Japan and Korea). Chiller assessed marbling score is the main criteria
used in the selection of animals for these markets. Certain breeds
id
Asian breeds such as the Japanese Wagyu and the Korean Hanwoo.
Intramuscular fat levels (a chemical measure of marbling) increase with length of time cattle
are raised on high-energy diets. However, the biochemical processes involved are not fully
understood. Additionally, there is controversy over whether it is the fat content or fatty acid
composition that has the most influence on palatability. It seems that
m
Korean cattle, may be influencing flavour more than the level of fatness.
Another interesting result was that Japanese cattle fattened in America did not reach the same
level of MUFA’s as their counterparts fed in Japan for the same length of time (Sturdivant et
al, 1992). Busboom et al (1993) stated that an increase in MUFA is responsible for an
increase in palatability traits of Wagyu beef produced
th
the biochemical processes involved in fat metabolism.
The aim of the present study was to de
and the flavour of beef striploins and to characteri
A
118
S
g and
avies Gene Mapping Projects, cited in Chapter 5. Further details of these animals can be
ethods, section 3.2.2 and Chapter 5, section 5.2.1.
B)
) 30 steers selected from the Davies Gene Mapping Project (DGM)
ploins imported into Australia
centage), melting point of fat and identification of
iacylgyceride fatty acids were carried out on all samples (see Chapter 3 - Materials and
s 3).
as that described in the previous chapter, section
5.2.2 and a comprehensive description of the taste panel methodology is presented in Chapter
3 – Materials and Methods, section 3.6.2.
6.2 MATERIALS AND METHOD
6.2.1 Selection of Animals for study
Beef striploins were collected from 170 animals, that represented many of the breeds
produced in Australia. A further 37 striploins were collected from Hanwoo animals and
imported into Australia from Korea, to make a total of 207 striploins to be used in the present
study. The Australian animals comprised the same subset of the Southern Crossbreedin
D
found in Chapter 3 – Materials and M
The animals used were as follows :
1) 70 heifers selected from the Southern Crossbreeding Project (SXB)
2) 70 steers selected from the Southern Crossbreeding Project (SX
3
4) 37 Korean Hanwoo (HAN) stri
6.2.2 Fat Measurements
Muscle fat content (intramuscular fat per
tr
Method , sections 3.3.1, 3.3.2 and 3.3.
6.2.3 Taste Panel Evaluation
The taste panel evaluation was the same
119
6.2.4 Statistical Analysis
Analysis of variance was carried out using the GLM procedure (SAS, 1996). A basic model
was developed to adjust the data for individual taster. Since taster name was not recorded,
taster had to be fitted within session by group. The effect of session, in addition to group
within session was also tested. Residuals for flavours were calculated and used for all
subsequent analysis. Following this, the effect of sex (cohort), breed and breed by sex group
(19 combinations) on the individual fatty acids were tested, with (Model 2) and without
(Model 1) intramuscular fat content (IMF%) as a covariate.
Model 1 : Residual fatty acid = Sex Breed Sex x Breed
Model 2 (IMF) : Residual fatty acid = IMF Sex Breed Sex x Breed
When fatty acid composition was tested for the effect of sire (within breed sex class), it was
shown that sire was significant (P<0.05) for myristic, palmitic and vaccenic acid. This
analysis was performed on Australian animals only, since sire information was unavailable for
the Hanwoo animals. Very few conclusions can be drawn from this analysis, due to the fact
that there are only 2-4 sires represented for each breed. Since only 10 animals were used per
breed sex class, this means that there are only 2-5 animals from each sire.
6.3 RESULTS
Table 6.1 shows the analysis of variance, which quantified the effect of sex, breed and the two
way interaction between sex and breed on the individual fatty acids. It can be seen that sex
was significant for myristic acid, trans-vaccenic acid and intramuscular fat (P<0.001),
myristoleic acid and linoleic acid (P<0.01) and oleic acid (P<0.05). Steers had 5.3%
120
intramuscular fat, which was significantly (P<0.001) higher than the value of 4.7% for
heifers. Steers also had a significantly higher (P<0.001) level of trans-vaccenic acid than
heifers (3.1% compared to 1.4% respectively). Heifers had significantly higher levels of
myristic (4.5%), myristoleic (1.7%), linoleic (1.5%) and oleic acid (40.3%) than the steers
(4.1%, 1.4%, 1.3% and 39.9% respectively).
Table 6.1 - Analysis of Variance Table - the effect of sex and breed on fatty acids
% Variation Accounted for (Type I SS) Model R2 IMF% Sex Breed Sex*Breed Total SS
Myristic (14:0) 1 23 9*** 11** 4 184 2 24 6 5 9 4 184
Myristoleic (14:1) 1 21 4** 15*** 2 68 2 21 2 7 10 2 68
Palmitic (16:0) 1 25 8 22*** 2 1659 2 27 18 1 7 2 1659
Palmitoleic (16:1) 1 24 0 22*** 2 369 2 24 7 0 15 2 369
Stearic (18:0) 1 39 0 36*** 3 1994 2 40 18 3 15 3 1994
Oleic (18:1n-9c) 1 29 2* 26*** 1 4691 2 34 26 1 5 2 4691
Trans-vaccenic (18:1n-7c)
1 34 7*** 23*** 4 1132
2 34 4 14 12 4 1132 Vaccenic (18:1n-7c) 1 40 0.3 40*** 0.5 58
2 41 18 6 16 0 58 Linoleic (18:2) 1 18 3** 7 8* 19
2 18 1 6 4 8 19 MUFA 1 33 0 31*** 2 5776
2 37 28 2 6 2 5776 MPt 1 58 1 55*** 3 4579
2 60 37 3 18 2 4579 IMF % 1 65 15*** 49*** 1 2233
All fatty acids were moderately to highly significant for breed, apart from linoleic acid (Table
6.1). Despite this, the two-way interaction between breed and sex was only significant
(P<0.05) for linoleic acid. (Table 6.2 shows the analysis of variance table for breed by sex
121
class). When the model was adjusted for intramuscular fat, all fatty acids became non-
significant for breed. Consequently, only least squares means for breed, calculated from
model 1 (Sex, Breed, Sex by Breed), are reported (Figure 6.1, Figure 6.2, Figure 6.3, Figure
6.4 and Figure 6.5).
Purebred Limousin (LL) steers had the highest level of palmitic acid, of 30.6% (±0.82). This
was significantly different (P<0.05) from JH (28.5%) and Hanwoo (25.2%) animals only. At
the other end of the scale, the Hanwoo had significantly lower levels of palmitic acid than all
other breeds (P<0.05). Of the Australian breeds, JH heifers had the lowest value of 28% for
palmitic acid, which was significantly higher than the value for the Hanwoo (P<0.05) and
significantly lower (P<0.05) than BH (30.1%) and LL (30.6%).
JJ animals had significantly higher (P<0.05) levels of palmitoleic acid (6.6%), to all other
breeds apart from the Hanwoo (6.1%). The lowest value for palmitoleic acid was 4.4% for
HH, which was significantly different from the Hanwoo, JJ (P<0.001), JH (P<0.01), and LJ
(P<0.05), which had values of 6.1, 6.6, 5.4 and 5.5 respectively.
Model 1 accounted for a large amount of the variation (65%) in intramuscular fat, a moderate
amount of variation in saturated (23 to 39%) and monounsaturated (21 to 40%) fatty acids,
but accounted for very little (7%) of the variation in linoleic acid (polyunsaturated fatty acid).
122
Figure 6.1 - Breed Least squares means (%) for Stearic acid and Oleic acid
05
101520253035404550
LL BH LH JJ HH
WH AH SH LJ JH H
an
% o
f tot
alStearic*** Oleic***
(*** P<0.001)
Hanwoo animals had the lowest value for stearic acid of 9.3% (±0.76), which was
significantly different (P<0.05) from all Australian breeds apart from JJ (10.7%). HH animals
had the highest levels of stearic acid of 14.6% (±0.57), which was significantly different
(P<0.05) to the Hanwoo, JH, WH, JJ and LH animals.
For oleic acid, the Hanwoo animals had the highest reported value of 46.5%, which was
significantly different (P<0.001) from all of the Australian breeds. The lowest value was
36.0% for LL animals, which was not significantly different to BH, LH or JJ, but significantly
different from all of the other breeds (P<0.05).
123
Figure 6.2 - Breed Least squares means (%) for Myristic acid and Myristoleic acid
0
1
2
3
4
5
6
LH LL HH
BH SH AH
WH LJ JH JJ
Han
% o
f tot
alMyristic** Myristoleic***
(** P<0.01 and *** P<0.001)
JJ animals had significantly higher (P<0.05) levels of myristic acid (5.1%), than all other
breeds apart from LL (4.6%), SH (4.37%) and WH (4.4%). Hanwoo animals had the lowest
value for myristic acid of 3.54, however this was not significantly different from AH, HH and
LJ animals, but significantly lower than all other breeds (P<0.05).
Hanwoo animals had the highest value (P<0.01) for myristoleic acid (1.9%), which was
significantly different (P<0.05) from all other breeds apart form JH (1.7%), WH (1.5%), JJ
(1.8%) and LJ (1.6%). The lowest value for this fatty acid was 1.3% for LH steers, which was
significantly different from Hanwoo (P<0.01), JJ and JH animals (P<0.05), which had values
of 1.9, 1.8 and 1.7 respectively.
124
Figure 6.3 - Breed Least squares means (%) for Palmitic and Palmitoleic acid
0
5
10
15
20
25
30
35
HH LH SH AH
WH
BH LL JH LJ
Han JJ
% o
f tot
alPalmitic*** Palmitoleic***
(*** P<0.001)
LL steers had the highest level of palmitic acid, of 30.6%. This was significantly different
(P<0.05) from JH (28.5%) and Hanwoo (25.2%) animals only. At the other end of the scale,
the Hanwoo had significantly lower levels of palmitic acid than all other breeds (P<0.05). Of
the Australian breeds, JH animals had the lowest value of 28.5% for palmitic acid, which was
significantly higher than the value for the Hanwoo (P<0.05) and significantly lower (P<0.05)
than BH (30.2%) and LL (30.6%).
JJ animals had significantly higher (P<0.05) levels of palmitoleic acid (6.6%), to all other
breeds apart from the Hanwoo (6.1%). The lowest value for palmitoleic acid was 4.4% for
HH, which was significantly different from the Hanwoo, JJ (P<0.001), JH (P<0.01), and LJ
(P<0.05), which had values of 6.10, 6.58, 5.43 and 5.46 respectively.
125
Figure 6.4 - Breed Least squares means for monounsaturated fatty acids (MUFA’s) (%) and Melting Point (°C) (MPT)
0
10
20
30
40
50
60LL LH BH
HH SH AH
WH LJ JJ JH H
an
% o
f tot
alMUFA*** MPT***
(*** P<0.001)
The total percentage of monounsaturated fatty acids showed a similar trend to that of oleic
acid, with Hanwoo animals having a value of 56.6% (±1.36), which was significantly higher
than all Australian breeds (P<0.001), which averaged 47.3% (±1.01). The lowest value
reported was for the LL steers. This value of 44.0% was not significantly different from LH,
BH or HH animals but significantly lower (P<0.05) than SH, AH, WH, LJ and JJ, lower than
JH (P<0.01) and of course lower (P<0.001) than the Hanwoo animals.
Melting point was significantly (P<0.001) negatively correlated with MUFA (-0.66) and thus
follows the inverse trend of MUFA, with the Hanwoo animals having a significantly
(P<0.001) lower melting point than all Australian breeds, with a value of 31.2°C (±1.0). For
the Australian breeds, JH steers had the lowest melting point of 37.7°C (±0.7). LL animals
had the highest melting point of 41.3°C, which was significantly different to Hanwoo
(P<0.001), JH (P<0.01) and JJ (P<0.05) only.
126
Figure 6.5 - Breed Least Squares means for trans-vaccenic and vaccenic acid
0
1
2
3
4
5
6
LL LH AH LJ HH
WH
BH JH H
an SH JJ
% o
f tot
altrans-vaccenic*** vaccenic***
(*** P<0.001)
LL steers had the highest level of trans-vaccenic acid of 4.9%, which was significantly higher
(P<0.05) than all other breeds. JJ animals had the lowest level (1.3%) of this acid, however, it
was only significantly different from AH, LH (P<0.05) and LL (P<0.001). Vaccenic on the
other hand, was significantly higher for Hanwoo animals (P<0.001), than any of the
Australian breeds, with a value of 2.7%. SH and HH had the lowest levels (1.6%) (P<0.05)
level of vaccenic acid, but this was not significantly different from any of the other Australian
breeds, however was significantly different (P<0.001) from the Hanwoo.
Table 6.2 - L east Squares Means for Breed Sex Classb
BS 14:0 14:1 16:0 16:1 18:0 18:1(n-9c) 18:1(n-7t) 18:1(n-7c) 18:2 MUFA MPT SXB Heifers
AHFa 4.63 1.68
29.58 4.81 11.93 39.96 2.01 1.87 1.39 48.40 38.50BHFa 4.46 1.64 30.07 5.21 12.22 38.45 0.84 1.83 1.52 47.30 38.50HHFa 4.65 1.68 28.99 4.75 12.69 39.45 1.61 1.82 1.43 47.70 37.90JHFa 4.55 1.71 27.99 5.19 12.88 40.79 0.67 1.90 1.42 49.60 37.90LHFa 4.80 1.56 29.50 4.96 13.37 37.74 2.64 1.84 1.48 46.10 39.30SHFa 4.65 1.53 28.90 4.70 13.42 39.31 0.72 1.84 1.44 47.40 39.60WHFa 4.83 1.73 30.35 4.79 12.60 38.63 1.00 1.89 1.31 47.10 39.00±SE 0.27 0.17 0.82 0.39 0.80 1.33 0.63 0.14 0.10 1.44 1.00
SXB Steers AHMa 3.70 1.21
28.67 4.43 14.56 39.67 4.33 1.41 1.38 46.75 41.00BHMa 4.02 1.11 30.36 4.54 13.95 38.46 3.44 1.51 1.35 45.59 40.90HHMa 3.68 1.03 28.71 3.95 16.43 39.54 3.24 1.45 1.34 45.98 42.40JHMa 4.07 1.66 28.93 5.67 12.09 40.31 3.57 1.70 1.27 49.35 37.40LHMa 3.59 0.98 28.61 4.20 15.31 39.50 4.09 1.49 1.37 46.17 41.50SHMa 4.09 1.27 29.36 4.49 14.99 40.47 1.91 1.42 1.34 47.67 40.50WHMa 4.01 1.32 29.16 4.51 13.26 40.61 3.49 1.67 1.20 48.09 38.30
±SE 0.27 0.17 0.82 0.39 0.80 1.33 0.63 0.14 0.10 1.44 1.00DGM Steers
JJMa 5.05 1.76
29.82 6.58 10.65 38.77 1.26 1.90 1.38 48.99 38.00LJMa 4.15 1.57 29.35 5.46 11.74 40.02 2.75 1.86 1.37 48.90 38.90LLMa 4.55 1.31 30.57 4.96 13.59 36.00 4.94 1.79 1.30 44.04 41.30±SE 0.27 0.17 0.82 0.39 0.80 1.33 0.63 0.14 0.10 1.44 1.00
Hanwoo F 3.30 2.03 24.17 6.47 9.07 47.83 2.00 2.87 2.27 59.33 31.67±SE 0.50 0.31
1.49 0.71 1.47 2.43 1.15 0.25 0.17 2.61 1.84Hanwoo M 3.77 1.76 26.30 5.75 9.47 45.12 0.80 2.44 1.41 53.92 30.68
±SE 0.15 0.09 0.44 0.21 0.43 0.72 0.34 0.07 0.05 0.78 0.55aA=Angus, B=Belgian Blue, H=Hereford, J=Jersey, L=Limousin, S=South Devon, W=Wagyu, F=Female, M=Male
b Model 1 : Residual fatty acid = Sex Breed Sex x Breed
127
Table 6.3 - Correlations between fatty acids 14:1 16:0 16:1 18:0 18:1(n-9c) 18:1(n-7t) 18:1(n-7c) 18:2 MUFA MPT IMF pH WBSF
14:0
0.50 ***
0.77 ***
0.38 ***
-0.25 **
-0.73 ***
-0.24 **
-0.19 **
-0.52 ***
0.28 ***
14:1
0.29 ***
0.74 ***
-0.69 ***
-0.16 *
0.20 **
-0.29 ***
16:0
0.15 *
-0.78 ***
0.15 *
-0.49 ***
-0.31 ***
-0.65 ***
0.44 ***
-0.19**
-0.28***
16:1
-0.83 ***
-0.15 *
0.38 ***
0.35 ***
-0.49 ***
18:0
-0.16 *
-0.41 ***
-0.49 ***
0.63 ***
-0.22 **
18:1(n-9c)
-0.39 ***
0.25 ***
0.90 ***
-0.51 ***
0.26 ***
0.27***
18:1(n-7t)
-0.41 ***
-0.21 **
18:1(n-7c)
0.21 **
0.39 ***
-0.61 ***
0.32 ***
18:2
MUFA
-0.66 ***
0.24 ***
0.25***
MPT
-0.17*
IMF
-0.18*
0.31 ***
aResidual correlations (Model = Sex , Breed, Sex by Breed) * P<0.05, ** P<0.01, *** P<0.001
128
129
Many of the fatty acids are significantly correlated with each other (Table 6.3), which since
they sum to 100% happens by definition. The highest correlation of 0.90 was seen between
oleic acid and MUFA% (P<0.001). This was expected, since oleic acid makes up the majority
of the monounsaturated fatty acid component in animal tissue. Melting point (MPT) was
significantly (P<0.001) negatively correlated with both MUFA% and IMF% (-0.66 and –0.17
respectively). This was also expected, since MUFA tends to increase with IMF% and melting
point tends to decrease as unsaturation increases. Other high correlations included the
negative correlation between oleic acid and palmitic acid (-0.78). Table 6.4 shows that oleic
acid was significantly positively correlated with initial juiciness, sustained juiciness, beef fat
flavour and buttery flavour (P<0.05). These correlations were 0.21, 0.17, 0.16 and 0.17
respectively. Chicken flavour was weakly negatively correlated with palmitic acid (-0.17,
P<0.05), and positively correlated (P<0.01) with vaccenic acid and intramuscular fat content
(0.19 and 0.10 respectively).
Buttery flavour seemed to be significantly (P<0.05) related to low levels of stearic (-0.17) and
linoleic acid (-0.16), low melting point (-0.20) and high levels of oleic acid (0.17) and
MUFA% (0.16). Additionally, buttery flavour was significantly (P<0.001) positively
correlated with IMF% (0.20). The highest (P<0.001) correlations were seen between IMF%
and intramuscular and sustained juiciness (0.25 and 0.23 respectively). It appears that
juiciness increases as intramuscular fat increases. This is also true for the other flavours, such
as beef fat flavour, oily flavour, buttery flavour, chicken flavour and flavour acceptability,
which were all significantly (P<0.001) correlated with IMF% (0.18, 0.18, 0.20, 0.10 and 0.11
respectively).
Table 6.4 - Correlations between fatty acids and residual flavours
14:0 14:1 16:0 16:1 18:0 18:1(n-9c) 18:1(n-7t) 18:1(n-7c) 18:2 MUFA MPT IMFIJUICE -0.16
* 0.21
** 0.18
** -0.16
* 0.25 ***
SJUICE -0.15 *
-0.15 *
0.17 *
0.21 **
0.17 *
-0.18 *
0.23 ***
BEEF
BFAT -0.15 *
0.16 *
0.17 *
-0.15 *
0.18 ***
OILY -0.14 *
-0.15 *
0.18 ***
BUTTERY -0.16 -0.17 *
0.17 * *
0.16 *
-0.20 **
0.20 ***
CHICKEN -0.17 *
0.19 **
0.10 ***
CORN
GRASSY -0.15*
-0.14 *
RANCID
FACCEPT 0.15 *
0.11 ***
a Correlations calculated between raw fatty acid data least square means for id, and residual least square means of flavours (calculated from the basic model which was adjusted for session, group nested within session and taster nested within group by session)
130
131
The Hanwoo had lower levels of saturated fats such as palmitic (16:0) stearic (18:0), and
myristic (14:0) acids, in addition to having higher levels of mono-unsaturated fatty acids
including oleic acid (18:1(n-9c)), myristoleic acid (14:1) and palmitoleic acid (16:1) than all
Australian breeds studied in this trial. Additionally, Hanwoo samples had the highest level of
vaccenic acid (18:1(n-7c)). Consequently, it was hypothesised that these fatty acids were
important in determining flavour differences between Australian and Korean animals.
Unfortunately, this theory was only weakly supported by the correlations between these fatty
acids and the flavours recorded by the taste panel (shown in Table 6.4).
6.4 DISCUSSION
The results of fat analysis in this experiment indicate that Australian cattle breeds differ in
fatty acid composition from each other and also from that of the Korean Hanwoo. Hanwoo
animals had a much higher level of mono-unsaturated fatty acid than that of the Australian
breeds. There is also some indication that there are differences in fatty acid composition
within the Australian breeds studied, with animals containing Jersey (JH, JJ, and LJ) having
slightly higher mono-unsaturated fat than the other Australian breeds, in particular LL, LH,
BH and HH animals. It is important to note that, whilst fat content is reported as a proportion
of muscle (wet weight), fatty acid composition is reported as a percentage of the total present.
This means that it is possible to have more of one fatty acid in one sample, yet less on an
absolute basis, than in another.
In chapter 5, it was demonstrated that chicken scores were highest for the Hanwoo animals,
although not significantly different from all Australian breeds. It was thought that this
particular flavour was more characteristic of Hanwoo animals than of the Australian breeds,
132
The two-way interaction between breed and sex was not significantly different for any of the
fatty acids, besides linoleic acid, indicating that within each breed, the individual fatty acids
follow the same trends for males and females.
and was examined closely for correlations with fatty acids. The weak negative correlation
between chicken flavour and palmitic acid in part supports the theory that animals with low
levels of palmitic acid such as the Hanwoo, have higher scores for chicken flavour, although
oleic acid and MUFA% were not related to chicken flavour.
Limousin animals (LL and LH) were the most different in fatty acid profile to the Hanwoo,
having the lowest level of mono-unsaturated fatty acids. LL had the lowest level of oleic acid
and LH had the lowest level of myristoleic acid, whilst pure Hereford (HH) had the lowest
level of plamitoleic acid. Despite this, the breeds did not mimic the same trend for the
saturated fatty acids, with HH having the highest level of stearic acid, LL having the highest
level for palmitic acid and JJ having the highest value for myristic acid.
Differences were seen in fatty acid composition between heifers and steers (cohort groups),
with steers having lower levels of myristic, myristoleic, linoleic and oleic acid and higher
levels of trans-vaccenic acid. The fact that heifers had higher levels of oleic acid is surprising
since normally unsaturation increases with fat content and length of time on feed. However,
this is in agreement with Westerling and Hedrick (1979), who demonstrated that fat from
steers had less stearic and oleic acid than did fat from heifers. It should also be noted that in
the present study, heifers and steers were fed differently, which may also have affected the
results.
133
It is well documented that the type of fat (fatty acid composition) contributes significantly to
meat flavour differences between animals. Tables 2.5, 2.6 and 2.7 report various correlations
between individual fatty acids and flavour from four independent studies. Recent studies have
indicated that increased levels of mono-unsaturated fatty acids are not only beneficial from a
nutritional point of view, but are also associated with more desirable flavour characteristics
(Schroeder et al., 1980; Busboom et al., 1993; May et al., 1993). The level of mono-
unsaturated fatty acids in this study was significantly positively correlated with increased
desirable flavour characteristics such as initial juiciness (0.18), sustained juiciness (0.17), beef
fat flavour (0.17) and buttery flavour (0.16), which is in agreement with these studies.
After intramuscular fat was added to the model, breed differences were no longer seen,
suggesting that fatty acid composition was highly related to fat content. The Hanwoo, which
was the most extreme for fatty acid composition, was also the most extreme for intramuscular
fat content, and therefore was adjusted down significantly to account for fat percentage.
Despite this, it is highly unlikely that animals of high fat content in Australia would reach
such high levels of mono-unsaturation. Animals from another study which contained 18%
intramuscular fat (8.5% higher in intramuscular fat content than the Hanwoo in this study),
produced similar fatty acid profiles to the other Australian breeds in this study (Siebert, pers
comm, 2000). However, the results of this study indicate that the effects of fatty acid cannot
truly be separated from that of fat content.
Oleic acid was the major fatty acid contributing to the total level mono-unsaturated fatty acid,
and consequently showed similar correlations with flavours as just mentioned, which supports
the finding of Melton et al. (1982b) who reported that higher concentrations of 18:1, in the
neutral lipid, and water soluble carbohydrates were positively correlated with flavour score.
134
Our results also support the study by Dryden and Marchello (1970) that increased
concentrations of 18:1 and 15:0 in the m.longissimus dorsi were generally scored higher by
panel evaluation, while 14:1, 16:0, 18:0 and 18:2 acids were less desirable when present in
increased quantities.
Additionally, this study showed that flavour acceptability was significantly correlated with the
mono-unsaturated fatty acid palmitoleic acid and intramuscular fat only (0.15 and 0.11
respectively). Jersey (JJ) animals had the highest levels of palmitoleic acid (significantly
different from all breeds apart from the Hanwoo) and were also the rated the most favourably
for flavour acceptability.
Based on the review of literature, it was expected that the Wagyu cross animals would have a
similar fatty acid composition to the Hanwoo animals, and most closely resemble the Hanwoo
in terms of taste and texture. Despite this, Wagyu cross heifers had the second lowest
percentage of MUFA’s (47%) out of all animals within the SXB heifer cohort. Wagyu cross
steers performed better, having the second highest level of MUFA’s (48%) after Jersey cross
steers (49%). Hanwoo animals achieved a level of 57% MUFA, which is approximately 10%
higher than any of the Australian animals. These results suggest that the Hereford dam, as
well as the environmental and nutritional conditions in Australia, may have exerted an
influence on Wagyu animals in the present study.
It has been shown previously that when Bos taurus breeds such as Murray Grey were
imported from Australia and fed for 12 months on a high energy ration in Japan, they
exhibited fat compositions similar to the other local Japanese and Wagyu breeds that could
not be achieved in Australia (Yang et al., 1999). The average level of MUFA’s exhibited in
135
Japanese cattle from this trial was 58%, which was similar to the level in Hanwoo cattle
imported from Korea in the present trial. Additionally, Australian cattle in the trial by Yang
et al., (1990) only achieved 46% MUFA, which again was similar to the levels of animals
within the present trial.
Since the Murray Grey steers imported into Japan exhibited fat composition similar to
Japanese animals, Yang et al., (1999) concluded that the differences in physical and chemical
compositions of carcase fat from Japanese and Australian long-term grain fed steers were
unlikely to be due to genetics alone. The high proportions of myristoleic, palmitoleic and
oleic acids may be due to Japanese animals having a greater adipose tissue desaturase activity
and hence the authors postulated a nutritional and/or environmental effect on this enzyme.
Wagyu cattle are well recognised for their superior fat characteristics, namely their extremely
soft (low melting point) fat due to having lower levels of saturated fatty acids, in particular
stearic acid and high levels of unsaturated fatty acids. Japanese Wagyu animals were reported
to have melting points of 22°C compared to 45°C for animals produced under Australian
conditions (Yang et al., 1999). Our results show similar trends with the imported Hanwoo
animals having a melting point of 31°C, which was much lower than any of the Australian
breeds (39°C). Wagyu cattle in the present trial had a melting point of 39°C, which was
slightly higher than Jersey and Jersey cross animals (38°C). Limousin animals had the
highest melting point of 41°C.
The results from the present study indicate that some of the differences in fatty acid
composition can be attributed to differences in genetics since Limousin, Jersey and Limousin
cross Jersey steers from the DGM project were all fed on the same diet, slaughtered together
136
and yet were markedly different in fatty acid composition. Despite this, these animals also
differed markedly in IMF%, with 3%, 6% and 7% for LL, LJ and JJ respectively. Therefore,
when fatty acids were adjusted for IMF% (Table 6.1), breed became non-significant
demonstrating that there was a strong relationship between the level of intramuscular fat and
fatty acid composition.
In Chapter 5, breed differences were demonstrated for initial juiciness, sustained juiciness,
beef flavour, corn flavour and flavour acceptability after adjusting for IMF% suggesting that
IMF% was not the only factor influencing the flavour of beef and that differences in these
flavours at least are partly genetic. Despite this, all of the flavours were highly correlated
with intramuscular fat percentage, which suggests that most of the panellists judgements of
flavour were heavily influenced by the amount of intamuscular fat present in the beef
samples. Additionally, IMF% was also significantly (P<0.001) correlated with all fatty acids
with the exception of linoleic acid. This raises the question of whether the effect of the fatty
acids on flavour is real or whether it is purely due to the influence of the amount of IMF%.
Westerling and Hedrick (1979) showed that meat from animals fed a concentrate diet for 112
days was more desirable in flavour than animals fed for 56 days. The percentage of total
saturated fatty acids was negatively correlated (P<0.01) with flavour scores, whilst oleic acid
had a positive effect on flavour. This is in agreement with the present study, which showed
that negative correlations were seen between palmitic acid and initial juiciness, sustained
juiciness, beef fat flavour, chicken flavour and also grassy flavour. Apart from grassy
flavour, all of these flavours are desirable, therefore the less palmitic acid, the more desirable
the flavour. In this study, Hanwoo animals had the lowest levels of palmitic acid than all
other breeds, whereas JH had the lowest value out of the Australian breeds.
137
6.5 CONCLUSIONS
It was shown in the present study, that the Hanwoo animals were markedly different in fatty
acid composition to all other breeds, being nearly 10% higher in mono-unsaturated fatty acids
than the average of all of the Australian breeds. It was also shown that Australian cattle
breeds differ markedly in fatty acid composition with animals containing Jersey more closely
resembling the Hanwoo in fatty acid profile and animals containing Limousin differing
significantly from the Hanwoo.
The trends for increased mono-unsaturation within the Australian breeds, reflect a similar
trend to the increases seen in Chapter 5 for flavour acceptability, with animals containing
Jersey having a higher flavour acceptability score, in addition to having a higher degree of
mono-unsaturated fatty acids.
It also appears that improved flavour characteristics were partly due to differences in fatty
acid composition (particularly due to an increase in mono-unsaturated fatty acids). However,
it was difficult to separate out the effect of individual fatty acids from the effect of
intramuscular fat, which was also highly correlated with total mono-unsaturated fatty acids,
oliec acid and negatively related to palmitic acid.
139
7.1 INTRODUCTION
Recognisable flavours or odours arise from the specific combination of complex mixtures of
many odorous molecules (Bartlett et al, 1997). The use of chemical sensors in sensory
evaluation of meat flavour has been used in a number of meat flavour studies (Annor-
Frempong et al.,Young et al., 1997).
The aim of this study was to analyse beef samples from Korean Hanwoo animals and a sub-
sample of Australian animals to determine whether differences could be established, using a
chemical sensor. A secondary aim was to test if the results from the chemical sensor could be
used for prediction of flavours in beef.
7.2 MATERIALS AND METHODS
7.2.1 Selection of Animals for study
Thirty striploins, representative of the Korean native breed (Hanwoo), were collected at a
Korean abattoir and imported into Australia by ELDERS Limited. (Only 30 of the 37
samples originally imported into Australia were used for this study due to limited quantity).
Beef striploins were also collected from 30 Australian animals, which comprised of a subset
of those from the Southern Crossbreeding and Davies Gene Mapping Projects (see Chapter 3
– Materials and Methods, section 3.2.3). The Australian animals were slaughtered after 180
days (steers) on a grain ration. Steers were approximately 25 months of age at slaughter
(mean carcass weight 326kg) as outlined below.
140
• 10 Limousin (LL) steers selected from the Davies Gene Mapping Project (DGM) based
at Martindale, Mintaro, SA.
• 10 Hereford (HH) steers selected from the Southern Crossbreeding Project (SXB) based
at Struan Research Centre, Naracoorte, SA.
• 10 Jersey (JJ) steers selected from the Davies Gene Mapping Project (DGM) based at
Martindale, Mintaro, SA.
• 30 Hanwoo steers imported into Australia from Korea.
7.2.2 Fat Measurements
Muscle fat content (intramuscular fat percentage) was carried out on all samples as described
previously (Chapter 3 - Materials and Methods, section 3.3.1).
7.2.3 Taste Panel Evaluation
A comprehensive description of the taste panel methodology is presented in Chapter 3 –
Materials and Methods, section 3.6.2, in addition to Chapter 5 – section 5.2.2.
7.2.4 Chemical Sensor
From the 60 selected striploins, a sample of chopped meat from each animal was run through
an automated headspace sampler coupled to a quadrupole mass sensor (HP 4440 Chemical
sensor). More details are presented in the Materials and Methods Section 3.4. After the
samples had been analysed in this manner, the means of the abundances of ions 35 to 180
were calculated from the 10 replicates from each of the 60 samples.
141
7.2.5 Statistical Analysis
Means of the ionic abundances (I35 to I180) for the 60 samples were transferred into SAS and
combined with the mean sensory data and the fatty acid data. Analysis of variance was carried
out using the GLM procedure (SAS, 1990). The following model was fitted to the data to
determine the effect of country of origin, project and breed on the individual ions.
Ion35...Ion180 = Country (Korea, Australia)
Project (Korea, Australia x SXB, Australia x DGM)
Breed (Hanwoo x Korea, HH x Australia x SXB, LL x Australia x
DGM, JJ x Austalia x DGM)
Least squares means and standard errors were calculated for each ionic mass. Additionally,
principal components were calculated between ions I35 to I180. Correlations between Ions
35 to 180 and the individual flavours and fatty acid compositions were formed using the
PROC CORR procedure in SAS (SAS, 1996).
7.3 RESULTS
The least square means of the ion abundances for those ions that were significant (P<0.05) for
country, project nested within country and breed nested within project x country (Table 7.1)
were calculated and are shown in Appendix 2, Table 7.2 and Table 7.3 . Ions that were highly
significant (P<0.001) for country are shown in red, those significant for project nested within
country are shown in green and those significant for breed nested within project x country are
shown in blue.
142
Table 7.1 - Ions significant for country, project or breed
R2 Country Project Breed TOTAL SS(x106)
Ion 36 14 ** 193Ion 38 14 * * 194Ion 39 13 * 863Ion 40 18 * 4053841Ion 42 18 ** 37993 Ion 43 * 110441017 Ion 44 17 ** 62962460Ion 45 13 * 1730506Ion 47 21 ** 870Ion 48 21 ** 2393Ion 49 13 * 81Ion 55 29 *** 1971Ion 56 32 *** 1879Ion 57 35 *** 3880Ion 58 16 ** 697Ion 60 37 *** 2756Ion 61 17 ** 159Ion 66 13 * 201Ion 67 24 *** 217Ion 69 21 *** 260Ion 70 30 *** 414Ion 71 32 *** 443Ion 72 14 ** 160Ion 77 14 ** 61Ion 78 12 33 * Ion 80 9 * 61Ion 82 19 *** 128Ion 83 15 ** 207Ion 85 22 *** 349Ion 86 34 *** * 849Ion 91 26 *** 274Ion 92 9 * 90Ion 94 11 * 1623Ion 95 24 *** 48
Model I36….I180 = country, country x project, breed x country x project. Note : *P<0.05, ** P<0.01, ***P<0.001
Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
143
Table 7.1(cont) - Ions significant for country, project or breed
R2 Country Project Breed TOTAL SS(x106)
Ion 101 19 ** 28Ion 102 17 ** 28Ion 104 11 * 22Ion 107 10 * 24Ion 110 14 * 23Ion 117 14 * 22Ion 118 16 ** 16Ion 124 10 * 24Ion 125 19 ** 26 Ion 126 18 ** 30Ion 130 10 * 16Ion 133 15 * * 25Ion 137 8 * 20Ion 144 15 ** 20Ion 145 10 * 26Ion 148 18 ** 25Ion 149 17 ** 21Ion 152 11 * 27Ion 153 13 ** 20Ion 154 14 * 26Ion 156 14 ** 24Ion 160 9 * 23Ion 162 16 * * 24Ion 163 12 * 29Ion 171 10 * 23Ion 175 13 ** 21Ion 177 12 * 25Ion 178 10 * 17
Model I36….I180 = country, country x project, breed x country x project Note : *P<0.05, ** P<0.01, ***P<0.001
Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
144
As can be seen in Table 7.1, a large number of ions were significant for country, indicating
that there were major differences between Korean and Australian samples. Only Ion 43 was
significant for country x project, whilst the effect of country was not significant. Ions 38, 86,
133 and 162 despite being significant for country, were also significant for breed x country x
project. Ions 77, 78 and 80 were all significant for breed x country x project only.
As seen in Appendix 2, the Korean samples had higher abundances (P<0.001) than Australian
samples for ions 55, 56, 57, 67, 69, 70, 71, 82, 85, 86, 91 and 95, whilst Australian samples
had a higher abundance (P<0.001) of Ion 60.
Table 7.2 shows that Korean samples had a higher abundance (P<0.05) for Ion 43 than
Australian samples from the SXB project, however this was not significantly different from
Australian samples from the DGM project. Appendix 3 shows the correlations of the
individual ions with fatty acids and flavours. Ion 43 was negatively correlated with trans-
vaccenic acid (-0.30), positively correlated with intramuscular fat (0.26), beef, beef fat and
oily flavours (0.39, 0.54 and 0.47 respectively). These results indicate that the samples from
the DGM project (LL and JJ steers) were more similar to Hanwoo animals, having more beef
fat and oily flavours than animals from the Southern Crossbreeding Project (HH steers).
Table 7.2 - LSMEANS for Ions significant for country x project
Ion 43 Australia, SXB 401720a
± 40496 Australia, DGM 506086b
± 28635 Korea 528049b
± 23381
145
Table 7.3 shows the least squares means for ions which were significant for breed nested
within project x country. For all ions except from Ion 133 and 162, the Hanwoo did not
significantly differ to the Jersey, however was significantly different to the Limousin samples.
The Hereford samples were not significantly different to the other Australian animals from the
DGM project except for Ion 80 in which Jersey samples were significantly higher in ionic
abundance. Figure 7.1 shows the ionic abundances for the Australian animals as a percentage
of the Hanwoo ionic abundance for ions 38, 86, 133, 162, 77, 78 and 80.
Table 7.3 - LSMEANS for Ions significant for breed x country x project
Ion 381 Ion 861 Ion 1331 Ion 1621 Ion 772 Ion 782 Ion 802 Hanwoo (Korean)
17142b 15074c 6320a 6131a 10124b 8873b 9489ab
Std. Err ±315 ±579 ±112 ±108 ±176 ±131 ±182 HxH
(Aust, SXB) 16113ab ab 10609ab 6747ab 6426ab 9804ab 8673 9302a
JJJJ (Aust, DGM)
17032b 13083bc 6915b 6809b 10521b 9038b
a
±1002 ±188 ±316
10206b
LLLL (Aust, DGM)
15396a 9819a 6326a 6126a 9304a 8242 9295a
Std. Err ±546 ±194 ±306 ±227 1Also significant for country, 2Not significant for country
Figure 7.1 - Ionic abundances for Australian animals shown as a % of Hanwoo
0
20
40
60
80
100
120
Ion 38 Ion 86 Ion 133 Ion 162 Ion 77 Ion 78 Ion 80
% ο
φ Η
ανω
οο ιο
νιχ
αβυ
νδα
νχ
HH JJJJ LLLL
146
Ion 38 was positively correlated with myristoleic acid (0.30), MUFA’s (0.26) and
intramuscular fat (0.39), whilst it was negatively correlated with stearic acid (-0.26) and trans-
vaccenic acid (-0.30). The Korean animals had similar ionic abundances of Ion 38 to the
Jersey animals.
Ion 86 indicates that the Hanwoo and Jersey are not significantly different from one another
however both are significantly higher in ion abundance from the Limousin breed. The Jersey
and Hereford are not significantly different and the Hereford and the Limousin are not
significantly different. This ion was negatively correlated with all saturated fatty acids and
melting point, and positively correlated with oleic acid, MUFA’s, IMF%, initial and sustained
juiciness, beef fat flavour and chicken skin.
As already mentioned, the Korean animals had a significantly (P<0.001) higher abundance of
ions 55, 56, 57, 70, 71 and 91 than all of the Australian breeds. All of these ions were
positively correlated (P<0.05) with mono-unsaturated fatty acids and apart from ion 91, were
all correlated with beef flavour, buttery flavour and corn flavour.
In chapter 5, it was suggested that Hanwoo animals had a higher intensity of chicken skin
flavour (although not significant). Other ions (besides ion 86), which were significantly
correlated with chicken skin flavour, included ions 36, 40, 101, 118, 133, 156, 107 and 147.
Ion 133 was negatively correlated with chicken flavour (-0.31) and therefore as the abundance
of ion 133 decreased, chicken flavour increases. Table 7.3 shows that the Hanwoo had
significantly lower abundances (and possibly higher chicken flavour) than Jersey samples,
however did not differ significantly from Hereford and Limousin samples.
147
For ion 60, the Korean Hanwoo had a significantly (P<0.001) lower abundance than the
Australian breeds. Ion 60 was positively correlated with palmitic acid, myristic acid, melting
point and stearic acid and negatively correlated with IMF%, oleic acid, MUFA, vaccenic acid,
linoleic acid, beef fat flavour, buttery flavour and rancid flavour.
Additionally, the Hanwoo sammples were significantly (P<0.01) higher than the Australian
breeds for ion 67, ion 85 and ion 95. These ions were also positively correlated to the total
level of monounsaturated fatty acids and negatively correlated to melting point. They were
also positively correlated to corn flavour.
For ion 82, the Hanwoo was significantly (P<0.01) higher than the Australian samples. Ion
82 was positively correlated to intramuscular fat, and myristoleic acid and corn flavour and
negatively correlated with stearic acid and melting point.
When principal components were formed between ionic abundances (I35 to I180), the first
three principal components explained 99% of the variation in meat (Table 7.4). The ions
contributing the most to each principal component are shown in bold. Principal Component
1, 3 and 8 were significant for breed (Table 7.5). Overall for the principal component
analysis, the main ions that seem to be important include ions 40, 41, 42, 43, 44, 45, 47, 48
and 60. Only these ions are shown in Table 7.4, since they seemed to have the most influence
on the principal components. Figure 7.2, Figure 7.3, and Figure 7.4 show the least square
means of the abundances of these ions for each breed. Appendix 4 shows all of the ions (ion
35 to ion 180) that make up the first ten principal components.
148
Table 7.4 - Variation accounted for by each Principal Component
Prin1 Prin2 Prin3 Prin4 Prin5 Prin6 Prin7 Prin8 Ion 40 0.07 0.95 -0.30 0.05 0.00 -0.04 0.03 -0.03 Ion 41 0.00 0.05 0.07 -0.08 0.63 -0.19 -0.02 0.27 Ion 42 0.01 0.08 0.07 0.11 0.07 0.69 -0.50 0.21 Ion 43 -0.07 0.30 0.89 -0.29 -0.13 -0.02 0.03 -0.03 Ion 44 0.98 -0.05 0.04 -0.17 -0.03 0.02 0.00 0.00 Ion 45 0.16 0.03 0.31 0.91 0.11 -0.16 0.03 -0.02 Ion 46 0.01 0.00 0.02 0.01 0.22 0.26 -0.15 -0.46 Ion 47 0.00 0.01 0.01 -0.01 0.08 0.02 0.22 0.35 Ion 48 0.00 0.01 0.01 0.02 0.12 0.29 0.39 0.33 Ion 60 0.00 0.01 -0.01 0.02 -0.21 -0.06 -0.28 0.56
Variation% 93.3 5.8 0.6 0.1 0.01 0.009 0.005 0.002 Cumulative % 93.3 99.1 99.7 99.8 99.81 99.82 99.82 99.83
aPrincipal components were adjusted for sex, breed, sex by breed, IMF% and sire
Table 7.5 - Principal Components which were significant for breed % Variation (Type I SS) Total SS
Breed (x106)Prin 1 17* 65258570Prin 3 21** 464945Prin 8 26*** 1505
Ionic mass 44 was the main component (0.98) of principal component 1. This ion was
negatively correlated (P<0.01) with beef fat flavour, oily flavour, buttery flavour, IMF%,
sustained juiciness, initial juiciness, beef flavour and flavour acceptability (-0.49, -0.39, -0.29,
-0.37, -0.35, -0.33, -0.30 and -0.28 respectively), which indicates that the first principal
component is reflecting fatness to a large degree, juiciness, beef flavour and flavour
acceptability.
Principal component 2 was not significant for breed, however, it did seem to reflect fatty acid
composition. Ionic Mass 40 (although not significant for breed) was the main component
(0.95) of principal component 2 and was significantly negatively correlated with chicken skin
flavour (-0.29), significantly negatively correlated with vaccenic acid, linoleic acid, total
monounsaturated fatty acids, oleic acid and palmitoleic acid (-0.48, -0.37, -0.37, -0.34 and –
149
Principal component 3 was significant for breed, with the main component being Ion 43
(0.89). This ion was positively correlated with beef fat flavour, oily flavour, beef flavour,
intramuscular fat and negatively correlated with trans-vaccenic acid (0.54, 0.47, 0.39, 0.26
and -0.30 respectively).
0.29 respectively) and positively correlated with myristoleic and stearic acid (0.29 and 0.31
respectively).
Principal component 4 was made up of predominantly ion 45, principal component 5 reflected
ion 41, principal component 6 comprised mainly of ion 42 and principal component 7
contained mostly ion 42.
Principal component 8, which was significant for breed (P<0.001), mostly reflected Ion 60
(0.56). This Ion is principally reflecting fatty acid composition, being positively correlated
with palmitic acid, myristic acid, melting point and stearic acid (0.59, 0.49, 0.54 and 0.26
respectively) and negatively correlated with IMF%, oleic acid, MUFA, vaccenic acid, linoleic
acid, beef fat flavour, buttery flavour and rancid flavour (-0.42, -0.55, -0.46, -0.39, -0.29, -
0.30, -0.28 and -0.26 respectively).
Table 7.6 shows the correlations between principal components 1, 3 and 8 with the fatty acids
and flavours. It can be seen from this table that principal component 1 was correlated
similarly with flavours and fatty acids to Ion 44. This is not surprising since ion 44 made up
98% of principal component 1. For principal components 3 and 8, several ions make up the
principal components and therefore don’t mirror any one particular ion. Despite this,
principal component 3 reflected beef fat flavour, having a positive correlation of 0.30. It also
150
had a negative correlation with melting point (-0.37) and stearic acid (-0.28). Principal
component 8 was positively correlated with flavour acceptability (0.27) and was positively
associated with fatty acids myristic, palmitic and melting point. This was unusual, since in
previous chapters, it was shown that increasing levels of saturated fatty acids were associated
with decreasing flavour acceptability.
Table 7.6 - Correlation between fatty acids and flavours with the Principal Components that were significant for breed
PRIN1 PRIN3 PRIN8 14:0 0.41
** 14:1
16:0 0.26 *
0.32 *
18:0 -0.28 *
IMF -0.37 **
MPT 0.30 *
-0.37 **
0.26 *
Initial Juiciness -0.33 *
Sustained Juiciness -0.35 **
Beef Flavour -0.30 *
Beef Fat Flavour -0.49 ***
0.30 *
Oily Flavour -0.39 **
Buttery Flavour -0.29 *
Flavour Acceptability -0.28 *
0.27 *
Ions 40 and 44 were responsible for the majority of the variation for principal components 2
and 1 respectively, and are shown in Figure 7.2. Ion 40 and ion 60 tended to reflect fatty acid
composition and were significantly different for the Korean and Australian animals,
151
irrespective of fatness. The Korean Hanwoo breed had significantly lower abundances of
both of these ions than the Australian breeds.
Figure 7.2 - Breed Comparison of Ion abundances (I40* & I44*)
bbba
cdd
cdc
01000000200000030000004000000
Hanwoo JJJJ HxH LLLL
5000000600000070000008000000
I40 I44
In Figure 7.3, ion 41 was not significantly different between any of the breeds, however is
cluded, since it makes up a major proportion of principal component 2. The Hanwoo
nimals had a significantly (P<0.05) lower abundance of ion 42 compared to JJ and HH
nimals, but was not significantly different to LL animals. Hanwoo animals had a
significantly (P<0.05) higher abundance of ion 43, than HH animals, however they were not
signifi 0.05)
different for ion 43. Hanwoo animals were not significantly different from JJ or LL animals
for ion 45, however, were again significantly different from HH animals.
in
a
a
cantly different from JJ or LL animals. HH and LL were not significantly (P<
152
Figure 7.3 - Breed Comparison of Ion abundances (I41ns, I42*, I43* & I45*)
a aaacd ddc
ef eff
gh h
ghg
0 20000400006000080000
100000120000140000
Hanwo JJJJ HxH LLLL
I41 I42 I43 I45
Hanwoo animals had significantly (P<0.01) lower abundance of ion 47 than JJ animals,
however this was not significantly different from that of HH or LL (Figure 7.4). A similar
trend was seen for ion 48, with Hanwoo only being significantly different (P<0.01) to JJ
animals. For ion 60 however, the ion abundance for Hanwoo animals was significantly
<0.001) lower that the other Australian breeds.
Figure 7.4 - Breed Comparison of Ion abundances (I47**, I48** & I60***)
(P
ababba
cdcddc fff
e
300004000050000
7000060000
01000020000
Hanwoo JJJJ HxH LLLL
I47 I48 I60
153
of the sensory characteristics of meat, the fact that a large number of ions were
gnificant for country, indicated that there were major differences between Korean and
rrelated with ion 86 (0.27), Hanwoo
nd Jersey had higher levels of this ion than Hereford and Limousin animals, which would
dicate that they may also have higher levels of chicken skin flavour than the other breeds.
dditionally, when the Hanwoo was significantly higher than all other Australian breeds for a
articular ion, it was determined that a majority of these ions were associated with increased
ids, lower levels of saturated fatty acids and positively
orrelated with buttery and corn flavours. These factors support the findings in Chapter 6 in
7.4 DISCUSSION
In the analysis
si
Australian samples in ionic abundances. The significant difference in ionic abundances
between the Korean and Australian samples can largely be attributed to the differences in
fatty acid composition between these samples. The Korean samples had a greater proportion
of MUFA’s and a higher incidence of chicken-skin flavour (not significant) and a lower
incidence of beef flavour than the Australian samples. The correlations between the various
fatty acids and flavours with the ions support this difference seen with the ion abundances
(Appendix 3).
Where chicken skin flavour was significantly (P<0.05) co
a
in
A
p
levels of mono-unsaturated fatty ac
c
which the Hanwoo was significantly different to all other Australian breeds for fatty acid
composition (in particular mono-unsaturated fatty acid).
154
the ions.
Although many ions stood out as being significantly different for Hanwoo animals compared
Australian animals, no single ion was isolated as being of sole importance. For this reason,
incipal components were formed from the ions and from this it was shown that the first
three principal components explained 99% of the variation between the meat samples and that
incipal components 1, 3 and 8 were significant for breed. The next chapter will use
principal components 1 to 8 in combination with fatty acids, in an attempt to predict flavour.
7.5 CONCLUSIONS
In summary, it was of note that the use of a chemical sensor was able to establish significant
differences between Korean Hanwoo and Australian breeds (Jersey, Limousin and Hereford).
These differences predominantly mirrored the differences in fatty acid composition and to
some extent flavour, as demonstrated by the individual correlations between these traits and
to
pr
pr
156
8.1 INTRODUCTION
hilst, sensory testing of beef is a useful tool to evaluate flavour changes due to genetic,
ive and time consuming exercise and is
often difficult to carry out in scientific experiments, let alone in the commercial environment.
For this reason, the ability to predict flavour, based on an objective measurement would be of
this study (initial and sustained juiciness, beef flavour, beef fat flavour, buttery flavour,
id flavour and flavour acceptability).
8.2.1 Selection of Animals for study
This study utilises the taste panel results from the main study (Chapter 5) and fatty acid
analysis results from the main study (Chapter 6). The selection of these animals was
escribed in Chapter 3 – Materials and Methods, section 3.2.3. Additionally, the results from
(Section 7.2.1 describes the animal selection for this study).
uscle fat content (intramuscular fat percentage), melting point of fat and identification of
triacylgyceride fatty acids were carried out on all samples. See Chapter 3 - Materials and
Methods, sections 3.3.1, 3.3.2 and 3.3.3.
W
nutritional or post slaughter differences, it is an expens
benefit to the industry.
The aim of this study was to use information from both the chemical sensor and also from
fatty acid composition, to develop prediction equations for individual flavours measured in
chicken flavour, corn flavour, ranc
8.2 MATERIALS AND METHODS
d
the chemical sensor were utilised
8.2.2 Fat measurements
M
157
8.2.3 Taste Panel Evaluation
om tion of the taste panel methodology is presented in Chapter 3 –
a) Cluster Analysis
ter analyses were carried out (SAS, 1996) on the fatty acid data,
ast square means of the flavour data, ionic abundances (I35 to I180) and the principal
A c prehensive descrip
Materials and Methods, section 3.6.2, in addition to Chapter 5 – section 5.2.2.
8.2.4 Chemical Sensor
A comprehensive description of the chemical sensor methodology is presented in Chapter 7 –
section 7.2.4.
8.2.5 Statistical Analysis
A cluster analysis was carried out using the CLUSTER procedure (SAS, 1996). This was
performed on both sets of fatty acid data, in addition to the least square means of the flavour
data for all 18 breed by sex groups. The least squares means for flavour were adjusted for
session, group within session and taster within group by session (Chapter 5 – section 5.2.5,
TABLE 5.1). The purpose of this analysis was to determine whether breed similarities
existed between fatty acid and flavour. If this was the case, it would indicate that fatty acid
composition could be used to predict flavour with a high degree of accuracy.
Additionally, a series of clus
le
component data (1-10 calculated from ionic abundances) for the breeds used in the chemical
sensor analysis. These included : SXB project HH steers, DGM project JJ and LL steers and
imported Hanwoo animals. The reason for looking at these animals in isolation was that the
Australian animals represented here had all been fed the same diet for the same period of
time, slaughtered together and are all purebred animals. This meant that there was no
158
s described previously in Chapter 7 – section 7.2.5, the means of the ionic abundances (I35
ombined with the mean sensory
data and the fatty acid data. Analysis of variance was carried out using the GLM procedure
(SAS, 1990). Breed x Country x Project (Hanwoo, (Korea), Hereford (Australia, SXB),
Limousin (Australia, DGM), Jersey (Austalia, DGM)) least squares means and standard errors
were calculated for each ionic mass. Additionally, principal components were calculated
between ions I35 to I180. The final breed (by country by project) least square means from
this model were used in the cluster analysis.
b)
The PROC REG procedure (SAS, 1996) was used to develop prediction equations for the
individual flavours (initial juiciness, sustained juiciness, beef flavour, buttery flavour, chicken
flavour, corn flavour, rancid flavour and flavour acceptability) using fatty acid data and
chemical sensor data. PROC REG sequentially dropped non-significant variables out of the
model for each flavour, until the final model was obtained. This was done for each flavour
using: a) fatty acid data only,
b) chemical sensor data only and finally
c) a combination of both fatty acid data and chemical sensor data.
It was decided to use the principal components calculated from chemical sensor data (as
determined in Chapter 7) for prediction of the flavours, rather than individual ions, due to the
fact that there were not large numbers of samples tested. If too many ions were fitted there
confounding effects of sex, finishing regime or hybrid vigour, and thus could be compared to
the Korean Hanwoo breed.
A
to I180) for the 60 samples were transferred into SAS and c
Development of Prediction Equations for Flavour
159
al degrees of freedom for testing. Consequently the principal
components used were Prin1 – Prin8 (Appendix 4).
8.3 RESULTS
The results in Chapter 6 indicated that the Australian cattle breeds differed markedly in fatty
acid composition between each other and also compared with Korean Hanwoo cattle. Within
the Australian animals, those containing Jersey (JH, JJ, and LJ), had slightly higher mono-
unsaturated fat than the other Australian breeds, in particular LL, LH, BH and HH animals.
When a cluster analysis was performed on this data, there was a distinct separation between
the Hanwoo and the Australian breeds for the cluster analysis of the fatty acids (Figure 8.1),
in d
to the results in Chapter 6. There were three distinct clusters for fatty acid composition. The
s (WHM), JH heifers (JHF), SH heifers (SHF), BH heifers (BHF), WH heifers
(WHF), AH heifers (AHF) and HH heifers (HHF).
eifers (LHF), HH steers (HHM), AH steers (AHM)
and LH steers (LHM).
would be insufficient residu
Once each model had been determined using the PROC REG procedure, the variables left in
the model were run through PROC GLM, in order to obtain the solutions for the model and
therefore to obtain the actual prediction equation.
ad ition to some separation between the Australian breeds, which followed a similar trend
Hanwoo (Han - cluster 1) was clearly very different from all Australian animals. Within the
remaining Australian animals there were two clusters, as follows:
• Cluster 2 – JJ steers (JJM), JH steers (JHM), LJ steers (LJM), SH steers (SHM), WH
steer
• Cluster 3 – BH steers (BHM), LH h
160
Unfortunately, however, the same clusters were not apparent for taste panel assessment of
flavours (Figure 8.2), indicating that it is unlikely that fatty acids alone were the sole source
of r d to reliably predict
fla u
Th l
• ers (JHF), LH steers (LHM), LL steers (LLM), BH steers (BHM),
BH heifers (BHF) and SH heifers (SHF)
• Cluster 2 – AH heifers (AHF), AH steers (AHM), JJ steers (JJM)
• Cluster 3 – Hanwoo (Han) and HH heifers (HHF)
• Cluster 4 – LH heifers (LHF) and WH heifers (WHF)
• Cluster 5 – JH steers (JHM), SH steers (SHM), HH steers (HHM), LJ steers (LJM) and
WH steers (WHM)
A subsequent cluster analysis used only HH, JJ, LL steers (Australian) and Hanwoo (Korean)
animals (Figure 8.3). It showed similar results to those of Figure 8.1, with a distinct
sep a ysis.
The Jersey steers were clustered near the Hanwoo animals and the LL steers were furthest
way.
HH steers (HHM - SXB project) and LL steers (LLM - DGM project)
pe ceived flavour differences and consequently could not be use
vo r.
e c usters for flavour were as follows:
Cluster 1 – JH heif
ar tion between the Korean animals and the Australian animals for fatty acid anal
a
Figure 8.4 indicated that there was a separation in flavour between the different breeds
(cohorts), with the Jersey animals again clustering with the Hanwoo animals.
• Cluster 1 –
• Cluster 2 – Hanwoo (Han - Korea) and JJ steers (JJM - DGM project)
161
Figure 8.1 - Average Linkage Cluster Analysis for Breed – Fatty Acids
|XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
a |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
c 1.25 +. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
C |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX . XXXXXXXXXXXXX l |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX . XXXXXXXXXXXXX u |. XXXXXXX XXXXXXXXXXXXXXXXXXXXXX . XXXXXXXXXXXXX s |. XXXXXXX XXXXXXXXXXXXXXXXXXXXXX . XXXX XXXXXXX t 0.5 +. XXXXXXX XXXXXXXXXXXXXXXXXXXXXX . XXXX XXXXXXX e |. . XXXX XXXX XXXXXXXXXXXXXXXX . XXXX XXXXXXX r |. . XXXX XXXX XXXXXXXXXXXXXXXX . XXXX XXXXXXX
H J J L S W J S B W A H L B L H A L A J H J H H H H H H H H L H H H H H N M M M M M F F F F F F M M F M M M 2.25 + | |
2 +. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX A |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX v |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX e |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX r |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX a 1.75 +. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX g |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX e |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX D |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX i 1.5 +. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX s |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX t |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
n |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
e |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX B |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX e |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX t 1 +. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX w |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX e |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX e |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXX n |. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXX 0.75 +. XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX . XXXXXXXXXXXXX
s |. . XXXX XXXX XXXX XXXXXXXXXX . XXXX . XXXX |. . XXXX . . XXXX . XXXXXXX . XXXX . . . 0.25 +. . . . . . . . . . XXXX . . . . . . |. . . . . . . . . . . . . . . . . .
162
Figure 8.2 - Average Linkage Cluster Analysis for Breed – Flavour
|XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
v |XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX
e |XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX
s |XXXXXXXXXXXXXXXX XXXXXXX XXXX XXXXXXXXXXXXXXXXXXX
a |XXXXXXXXXXXXXXXX XXXXXXX XXXX XXXXXXXXXXXXXXXXXXX
c |XXXXXXXXXXXXXXXX . XXXX XXXX XXXXXXXXXXXXXXXXXXX
+XXXXXXXXXXXXXXXX . XXXX XXXX XXXX XXXXXXXXXXXXX B |XXXXXXXXXXXXXXXX . XXXX . . XXXX XXXXXXXXXXXXX
w |XXXX XXXXXXXXXX . . . . . XXXX XXXXXXXXXXXXX e |XXXX XXXXXXXXXX . . . . . XXXX XXXX XXXXXXX e |XXXX . XXXXXXX . . . . . XXXX XXXX XXXXXXX n |XXXX . XXXXXXX . . . . . XXXX XXXX . XXXX 0.4 +. . . . XXXX . . . . . XXXX XXXX . XXXX |. . . . XXXX . . . . . . . . . . XXXX
l |. . . . . . . . . . . . . . . . XXXX u |. . . . . . . . . . . . . . . . XXXX s |. . . . . . . . . . . . . . . . XXXX t |. . . . . . . . . . . . . . . . . . e |. . . . . . . . . . . . . . . . . . r |. . . . . . . . . . . . . . . . . . s 0.2 +. . . . . . . . . . . . . . . . . .
J L L B B S A A J H H L W J S H L W H H L H H H H H J A H H H H H H J H F M M M F F F M M N F F F M M M M M 1.2 +XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
|XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX |XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX A 1 +XXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
e |XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX r |XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX a |XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX g |XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX
|XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX D 0.8 +XXXXXXXXXXXXXXXX XXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXX i |XXXXXXXXXXXXXXXX XXXXXXX XXXX XXXXXXXXXXXXXXXXXXX
t |XXXXXXXXXXXXXXXX XXXXXXX XXXX XXXXXXXXXXXXXXXXXXX
n |XXXXXXXXXXXXXXXX . XXXX XXXX XXXXXXXXXXXXXXXXXXX
e |XXXXXXXXXXXXXXXX . XXXX XXXX XXXX XXXXXXXXXXXXX 0.6
e |XXXX XXXXXXXXXX . XXXX . . XXXX XXXXXXXXXXXXX t |XXXX XXXXXXXXXX . . . . . XXXX XXXXXXXXXXXXX
C
163
ntry – Fatty Acids
1.4 +
|
e |. XXXXXXXXXXXXX
a |. XXXXXXXXXXXXX
e |. XXXXXXXXXXXXX
D |. XXXXXXXXXXXXX
t 0.6 +. . XXXXXXX
s |. . . . t |. . . . e |. . . . r 0.2 +. . . . s |. . . .
. |. . . .
.
. . . .
Figure 8.3 - Average Linkage Cluster Analysis for Project by CouH J H L
A J H L N M M M
| |
| A |XXXXXXXXXXXXXXXXXXX v 1.2 +. XXXXXXXXXXXXX
r |. XXXXXXXXXXXXX
g |. XXXXXXXXXXXXX
1 +. XXXXXXXXXXXXX
i |. XXXXXXXXXXXXX s |. XXXXXXXXXXXXX t |. XXXXXXXXXXXXX a |. XXXXXXXXXXXXX n 0.8 +. XXXXXXXXXXXXX c |. XXXXXXXXXXXXX e |. . XXXXXXX |. . XXXXXXX B |. . XXXXXXX e |. . XXXXXXX
w |. . XXXXXXX e |. . . . e |. . . . n |. . . . |. . . . C 0.4 +. . . . l |. . . . u |. . . .
|. . .
|. . . . . |. 0 +.
164
Flavours
e |XXXXXXX XXXXXXX
a |XXXXXXX XXXXXXX
|XXXXXXX XXXXXXX
s |XXXXXXX XXXXXXX
n |XXXXXXX . .
0.6 +. . . .
e |. . . .
e |. . . .
0.4 +. . . . C |. . . . l |. . . . u |. . . . s |. . . .
Figure 8.4 - Average Linkage Cluster Analysis for Breed – H L H J H L A J M M N M 1.2 + | | |XXXXXXXXXXXXXXXXXXX |XXXXXXX XXXXXXX |XXXXXXX XXXXXXX |XXXXXXX XXXXXXX |XXXXXXX XXXXXXX A 1 +XXXXXXX XXXXXXX v |XXXXXXX XXXXXXX
r |XXXXXXX XXXXXXX
g |XXXXXXX XXXXXXX e |XXXXXXX XXXXXXX
D 0.8 +XXXXXXX XXXXXXX i |XXXXXXX XXXXXXX
t |XXXXXXX XXXXXXX a |XXXXXXX . .
c |XXXXXXX . . e |XXXXXXX . .
B |. . . .
t |. . . . w |. . . .
e |. . . . n |. . . .
t |. . . . e |. . . . r |. . . . s 0.2 +. . . . |. . . . |. . . .
165
igure 8.5 shows the cluster analysis for breeds for principal components 1-10, where there is
with SXB animals separated
from the Hanwoo and DGM project animals.
initial
nd sustained juiciness, and was responsible for explaining 14% of the variation. Only 8% of
e variation in flavour acceptability was explained by the fatty acids palmitoleic, stearic,
yristic, linoleic and IMF%. For all of the other flavours, the R2 value was below 8%. It was
interesting to note that intramuscular fat (IMF%) was contained in all of the models with the
xception of corn and rancid flavour, suggesting its’ importance in influencing flavour.
F
a distinct separation between animals from the SXB project (HH animals) and the other
animals (Korean and DGM animals). Again JJ steers (JJM) were clustered near the Hanwoo
animals. Figure 8.6 for ions 35 to 180 shows a similar pattern
In another attempt to predict flavour from fatty acid composition, prediction equations were
developed from fatty acid data (Table 8.1). The accuracy of predicting flavour from fatty acid
composition, however, was extremely low (R2 = 3% to 17%).
For buttery flavour, 17% of the variation in flavour was explained by the individual fatty
acids – palmitic, palmitoleic, stearic, myristic, t-vaccenic, vaccenic, linoleic and IMF%. The
fatty acids palmitic, oleic, myristoleic, linoleic and IMF% explained 10% of the variation in
chicken flavour, whilst palmitic, palmitoleic, vaccenic and IMF% explained 10% of the
variation in beef fat flavour. IMF% was the only variable left in the model to predict
a
th
m
e
16 6
Figure 8.5 - Average Linkage Cluster Analysis for Breed – Principal Component 1-10
|
|. XXXXXXXXXXXXX
C |. . XXXXXXX l |. . XXXXXXX u 0.4 +. . XXXXXXX
t |. . XXXXXXX
. . . .
. . .
H H J L H A J L M N M M 1.4 + | |
|XXXXXXXXXXXXXXXXXXX
|. XXXXXXXXXXXXX 1.2 +. XXXXXXXXXXXXX A |. XXXXXXXXXXXXX v |. XXXXXXXXXXXXX e |. XXXXXXXXXXXXX r |. XXXXXXXXXXXXX a |. XXXXXXXXXXXXX g |. XXXXXXXXXXXXX e 1 +. XXXXXXXXXXXXX |. XXXXXXXXXXXXX D |. XXXXXXXXXXXXX i |. XXXXXXXXXXXXX s |. XXXXXXXXXXXXX t |. XXXXXXXXXXXXX a |. XXXXXXXXXXXXX n 0.8 +. XXXXXXXXXXXXX c |. XXXXXXXXXXXXX e |. XXXXXXXXXXXXX |. XXXXXXXXXXXXX B |. XXXXXXXXXXXXX e |. . XXXXXXX t |. . XXXXXXX w 0.6 +. . XXXXXXX e |. . XXXXXXX e |. . XXXXXXX n |. . XXXXXXX |. . XXXXXXX
s |. . XXXXXXX e |. . . r |. . . s |. . .
. . |. 0.2 +.
167
Figure 8.6 - Average Linkage Cluster Analysis for Project – Ions I35 to I180
n |. . . 0.4 +. . . C |. . . l |. . . . . . . 0.2 . .
H LJ H M MM N
H LJ A
1.2 + | |XXXXXXXXXXXXX |. XXXXXXX |. XXXXXXX |. XXXXXXX |. XXXXXXX |. XXXXXXX A 1 +. XXXXXXX v |. XXXXXXX e |. XXXXXXX r |. XXXXXXX a |. XXXXXXX g |. XXXXXXX e |. XXXXXXX |. XXXXXXX D 0.8 +. XXXXXXX i |. XXXXXXX s |. XXXXXXX t |. XXXXXXX a |. XXXXXXX n |. XXXXXXX c |. XXXXXXX e |. XXXXXXX 0.6 +. XXXXXXX B |. XXXXXXX e |. . . t |. . . w |. . . e |. . . e |. . .
u |. . .
s
|. .
t e
|. |.
. .
r s
|. +.
. .
|. .
168
abl - P tion s using f cids
R2 MEAN Variables ESTIMATE
T e 8.1 redic of flavour atty a
± SE Initial Juiciness 14 4.71 Intercept
IMF 4.01 0.13
0.14 0.02
Sustained Juiciness IMF
14 4.82 Intercept 4.18 0.12
0.13 0.02
Beef Flavour 3 4.44 Intercept Palmitic Myristic
Myristoleic IMF
3.28 0.05 -0.17 0.22 0.03
0.80 0.03 0.11 0.13 0.02
Beef Fat Flavour 10 3.46 Palmitic
0.89
0.16
Intercept
Palmitoleic Vaccenic
IMF
4.33 -0.04 0.06 -0.29 0.07
0.03 0.05
0.02 Buttery Flavour 17 2.53 Intercept
Pal ic
Vaccenic
Palmitic mitole
Stearic Myristic
T-vaccenic
Linoleic IMF
6.82 -0.08 -0.08 -0.08 0.11 0.05 -0.34 -0.54 0.06
1.44 0.04 0.07 0.04 0.09 0.02 0.16 0.18 0.02
Chicken Flavour 10 1.83 1.28 Intercept Palmitic
Oleic Myristoleic
Linoleic IMF
6.25 -0.09 -0.04 0.12 -0.31 0.02
0.02 0.01 0.07 0.13 0.01
Corn Flavour 4 1.59
ristic
Intercept Pa c lmitolei
Steric Oleic
My
1.44 -0.10 -0.03 0.02 0.09
1.00 0.05 0.02 0.01 0.06
Rancid Flavour 5 1.28 Intercept Palmitic Stearic
Myristic Myristoleic Vaccenic
-0.06 -0.02 0.13 -0.13 -0.14
0.02 0.02 0.05 0.08 0.09
3.10 0.66
Flavour Acceptability Palmitoleic
Myristic
IMF
0.15
0.07
0.06
0.07
0.06
0.02
8 5.64 Intercept
Stearic
Linoleic
3.83
0.06
-0.23
0.86
0.03
0.17
169
hen principal components 1 to 8 (calculated from the chemical sensor data – individual
i
flavour because the m of e principal components is zero (by definition).
Additionally, the estimates and standard errors for the princ nts ound
zero. Table 8.2 shows moderate R2 values for the prediction of flavours.
Table 8.2 - Prediction of flavours using princ mpo e c sor d
R2 MEAN Variables ESTIMATE ±
W
ons) were used to predict flavours, the intercept was the same value as the mean value of the
ean all th
ipal compone were all ar
ipal co nents from th hemical senata
SE Initial Juiciness 20 5.04
In t
-0.00000037 -0.000
0.000015463 0.0
tercepPrin1 Prin2 Prin3 Prin5
5.04
000576 0.000002529
0.14 0.00000014 0.00000054 0.00000161
000096 Sustained Juiciness 19 5.17 In t
0.00 84 0.0 47
tercepPrin1 Prin2 Prin3
5.17 -0.000000361 -0.000000573
00027
0.13 0.00000012 0.0000005
00001Beef Flavour 25 4.52 Intercept
Prin1 Prin2 Prin3 Prin5 Prin6
4.52 -0.00000024 0.000000847 0.00 49 0.000012048
00011
0.000015021
0.10 0.0000001
0.00000038 0.0 13 0.00000675
00001
0.00000957 Beef Fat Flavour 35 3.60 Intercept
Prin1 Prin2 Prin3
3.60 -0.000000411 0.000000497 0.000002937
0.09 0.00000009 0.00000036 0.00000106
Buttery Flavour 12 2.61 Intercept Prin1 Prin5 Prin6
-0.000000224 0.000007421 -0.00001133
0.0000001 0.00000689 0.00000977
2.61 0.10
Chicken Flavour 9 1.93 Intercept
Prin3
1.93
0.000001132
0.08
0.00000096 Prin2 -0.000000679 0.00000032
Flavour Acceptability 20 .75 Intercept Prin1 Prin3 Prin7
5.75 -0.000000202 0.000001364 -0.000015305
0.09 0.00000009 0.00000104 0.00001175
5
Prin8 0.000041174 0.00001827
170
e appearing in the prediction equation. Principal component 1 was contained
prediction equations for initial and sustained juiciness, beef flavour, beef fat flavour,
buttery flavour and flavour acceptability. This principal component was negatively correlated
ith intramuscular fat, beef flavour, beef fat flavour, oily flavour, buttery flavour, flavour
cceptability, initial and sustained juiciness, and positively correlated with palmitic acid and
elting point.
When both the fatty acid data and the chemical sensor data (principal components 1-8) were
used to predict flavour (Table 8.3), reasonable R2 values were obtained (24% for chicken
flavour to 43% for beef fat flavour). Once again, intramuscular fat content was included in
the prediction equations of all flavours.
The variables used to predict beef fat flavour included, principal components 1, 2 and 3,
palmitic, palmitoleic, vaccenic acids and IMF% which explained 43% of the variation in beef
fat flavour. For chicken flavour, principal components 2 and 3, palmitic, oleic, myristoleic,
linoleic and IMF% explained 24% of the variation in this flavour (Table 8.3).
Principal components 1, 2 and 3 seemed to be important in predicting individual flavours,
with one or mor
in
w
a
m
171
Table 8.3 - Prediction of flavours using principal components in addition to fatty acid data
R2 MEAN Variables ESTIMATE ± SE Initial Juiciness 32 5.04 Intercept
Prin1 Prin2
4.06 -0.000000197 -0.000000422 0.0000001541
0.33 0.00000014 0.00000051 0.00000152 Prin3
Prin5 IMF
0.000005068 0.121373203
0.00000948 0.03825631
Sustained Juiciness 28 5.17 In t tercepPrin1 Prin2 Prin3 IMF
4.47 -0.000000238 -0.000000463 0.000002081 0.086381330
0.30 0.00000013 0.00000047 0.00000143 0.03369992
Beef Flavour 28 4.52 -0.000000234 0.000000735 0.00000043
Intercept Prin1 Prin2 Prin3 Prin5 Prin6
Palmitic Palmitoleic Vaccenic
IMF
3.20
0.000001293 0.000012177 0.000017821 0.057022781 -0.150799721 0.210872368 0.002540975
1.24 0.00000011
0.00000122 0.00000753 0.00001067 0.05365794 0.19804971 0.18794805 0.03500448
Beef Fat Flavour 43 3.60 Intercept Prin1 Prin2 Prin3
Palmitic Palmitoleic V accenic
IMF
4.06 -0.000000328 0.000000502 0.000002835 -0.027548723 0.074392061 -0.272004806 0.058528530
1.59 0.00000010 0.00000039 0.00000113 0.04544688 0.07684948 0.24744598 0.02832594
Buttery Flavour 38 2.61 Prin1
Pa ic
T- ic
0.000005080
-0.107435709 0.06805661
Intercept
Prin5 Prin6
Palmitic lmitoleStearic
Myristic vaccen
Vaccenic Linoleic
IMF
7.30 -0.000000107
-0.000003345
0.064351144 -0.096095502 0.094957007 0.090275489 -0.411414598 -0.706624935 0.044491334
2.51 0.00000010 0.00000786 0.00001072
0.12607749 0.06077594 0.18950675 0.05293556 0.26784210 0.28312616 0.03703866
Chicken Flavour 24 1.93 Intercept Prin2 Prin3
P
almiticOleic
Myristoleic Linoleic
IMF
7.94 -0.000000916 0.000001097 -0.095206398 -0.073453693 0.197955946 -0.605566104 0.035682993
2.34 0.00000035 0.00000096 0.04448108 0.02916067 0.14047172 0.24880237 0.02840233
Flavour Acceptability 27 5.75
Prin3 Prin7 Prin8
Palmitoleic Stearic
IMF
0.000001315 -0.000017858 0.000029209 0.136862568 0.036383395
0.028853549
0.00000110 0.00001260 0.00002043 0.11012113 0.05458619
0.03392344
Intercept Prin1
Myristic Linoleic
0.128269208 -0.262584740
0.12446597 0.25231373
4.21 -0.000000182
1.46 0.00000010
172
at the cluster analysis for the Principal components and
dividual ions. Unfortunately, this did not provide conclusive results since the effect of
ently high correlations were apparent between
dividual fatty acids and the sensory panel traits. The present study is in agreement with this
flavour characteristics were partly due to
differences in fatty acid composition (particularly due to an increase in mono-unsaturated
fatty acids), it was difficult to separate out the effect of individual fatty acids from the effect
8.4 DISCUSSION
The results of the cluster analysis demonstrated that whilst the entire data set was not
conclusive in showing the relationship between fatty acid and flavour, a smaller less
confounded data set (containing purebred Australian animals fed on a similar diet to Korean
animals) was able to show distinct trends which support the findings in Chapter 7. Jersey
animals most closely matched the fatty acid composition and flavour of the Hanwoo animals.
Additionally, Hereford (SXB) animals were distinctly different to DGM (LL and JJ steers)
and Korean animals when looking
in
intramuscular fat was still evident, ranging from 3.36% in LL steers, 4.66% in HH steers,
6.80% in JJ steers and 10.50% in Hanwoo animals. What was apparent was the cluster
groupings tended to reflect fatness.
Dryden and Marchello (1970) showed that an increased lipid content (IMF%) had a desirable
influence on tenderness, juiciness, overall acceptability and to a smaller extent, flavour.
However, few significant and no consist
in
finding in that despite being difficult to show a good relationship between individual flavours
and fatty acids, intramuscular fat content seemed to be a major influence on individual
flavours and also on flavour acceptability.
In Chapter 6, it was shown that whilst improved
173
ting with R2 values
ss than 15%. More promising was the prediction equations formed from the fatty acid data
fatty acids and positive correlations
ith monounsaturated fatty acids. Ion 40 also had a negative correlation with chicken skin
avour. Principal component 3 is positively correlated with beef fat flavour and negatively
orrelated with stearic acid and melting point Table 7.5.
of intramuscular fat, which was also highly correlated with total mono-unsaturated fatty acids,
oleic acid and negatively related to palmitic acid.
Attempts to predict flavour from individual fatty acids was quite disappoin
le
combined with the principal components (formed from the individual ion data from the
chemical sensor). These equations resulted in R2 values of 24% to 43%.
In chapter 5, it was suggested that the Korean Hanwoo animals had a numerically (although
not statistically) higher chicken skin flavour and a lower intensity of beef flavour than the
Australian breeds. In the final prediction equation for chicken flavour, the model (containing
principal components 2 and 3, palmitic, oleic, myristoleic, linoleic and intramuscular fat)
explained 24% of the variation in this flavour. Principal component 2 reflected fatty acid
composition, as demonstrated by the fact that it was predominantly made up of Ion 40 (95%)
which had moderate negative correlations with saturated
w
fl
c
174
ble to be achieved for certain flavours when
oth fatty acid data and chemical sensor data were combined (R2 values of 24% to 43%).
F%
arries with it more fatty acid in absolute terms (as distinct from a percentage of the total) and
so that increased IMF% affects the texture of the food (‘mouth-feel’). Despite this,
intramuscular fat was contained in all prediction equations for individual flavours, which
monstrates its’ importance in influencing individual flavours and flavour acceptability.
8.5 CONCLUSIONS
In conclusion, it appears that it is extremely difficult to predict flavour from objective
assessment of other traits such as fatty acid composition and chemical sensor data.
Moderately accurate prediction equations were a
b
Additionally, cluster analysis showed that Jersey animals were clustered with Hanwoo
animals suggesting that they were similar for both fatty acid composition and flavour. This
supports earlier findings from Chapters 5 and 6.
It was virtually impossible to separate the effects of intramuscular fat from the effect of fatty
acid composition, due to the fact that the Hanwoo, which was significantly different for the
majority of fatty acids (P<0.001), was also significantly different from all of the Australian
breeds for intramuscular fat. This meant that there was a continual confounding effect
between fatty acid composition and fat. This may well be due to the fact that increased IM
c
al
de
176
Since this research project commenced, the retail sector in Korea has been unsettled following
market liberalisation on 1st January 2001, the carry-over of beef stocks, a fall in beef
onsumption as a result of negative publicity surrounding Foot and Mouth/BSE, and the
ite this, the temperatures used were quite mild and
nly a single muscle was used instead of a whole carcase which would have taken longer to
ur
iscolouration and rancidity could occur.
stored for various lengths of time in frozen storage
efore tasting (up to one year) and therefore a secondary aim of the freeze-thaw trial was to
c
slowing of the national economy (Tilley, 2001). Historically, Australia has had a reputation
for supplying cheap lean beef to Korea and whilst there is an increasing amount of chilled
product being exported, Australia still predominantly exports frozen grass-fed product
destined for use in wet cooking, soup, mass catering and processing (Lugsdin, 2000).
The handling methods of Australian frozen product in Korea is still of major concern,
particularly as freezing and thawing is being repeated at different distribution points
throughout Korea. The results from the freeze-thaw trial in the present study (aimed to
demonstrate a detrimental effect of this practice), however showed that tenderness was
enhanced through accelerated aging. Desp
o
thaw. Bacterial populations were not tested so it is not known what effect the freezing and
thawing had on hygiene. There were no significant effects on colour or fatty acid
composition, however it was hypothesised that at more extreme temperatures colo
d
If this practice continues, in order to satisfy hygiene and meat quality standards for Australian
beef entering Korea, a code of practice, such as “the optimisation of the thawing and
processing of Australian frozen quarter beef in Korea” (Powell, 1993), should be followed.
Samples used throughout this study were
b
177
determine whether fatty acid composition changed as a result of freezing and thawing
treatments. Since no changes were seen in fatty acid composition, even when thawed three
times, it was assumed that fatty acid composition in frozen samples for the main study would
be similar to those measured at slaughter.
The literature reviewed in Chapter 2 indicates the complexity of meat flavour and describes
d 8 follow a logical progression from the taste panel
valuation of individual flavours, objective measurements of tenderness, intramuscular fat and
ated fatty acids and a more desirable flavour.
espite this, Wagyus raised in America have not been able to reach the same levels of
monounsaturation and palatability as their relatives in Japan. Sturdivant et al, (1992) showed
the various factors that influence the fatty acid composition and other flavours of beef,
including environment, genetics, nutrition and post-slaughter factors. It can be seen from this
review that it is unlikely a single factor determines beef flavour, however the fatty acid
hypothesis seems to show the most promising correlations with various flavours.
In the present study, Chapters 5, 6, 7 an
e
fatty acid composition, through to use of a chemical sensor and finally the development of
prediction equations for flavour based on a combination of traits. The majority of the
discussion and conclusions are contained within each chapter, however a number of
observations can be made in summary.
If genetics played a major role in determining fatty acid composition and flavour, it would be
expected that the Wagyu cross animals would have a similar fatty acid profile and flavour to
the Korean Hanwoo. Both the Japanese Wagyu and the Korean Hanwoo are similar type
animals that have extremely high levels of intramuscular fat (marbling), monounsaturated
fatty acids (MUFA’s), low levels of satur
D
178
eed alone, is that of Yang et al. (1999), in which
urray Grey animals imported from Australia and fed under Japanese feedlot conditions in
H, JJ, and LJ)
aving slightly higher mono-unsaturated fat than the other Australian breeds, in particular LL,
that purebred Wagyu cattle from Japan had elevated MUFA’s in their adipose tissues and
suggested a genetic basis for this compositional difference. Despite this, they could not rule
out the efffect of environmental influences.
Another study which supports the theory that the environment influences fatty acid
composition, rather than genetics and f
M
Japan achieved similar extreme levels of MUFA’s to the Japanese Wagyu. This is the
strongest evidence in the literature that supports the effect of an environmental rather than
genetic effect on fatty acid composition.
The results of the fat analysis from the present trial indicate that the Australian cattle breeds
differed markedly in fatty acid composition between each other and that of Korean Hanwoo
cattle. The latter had 57% mono-unsaturated fatty acids, which was significantly higher than
all of the Australian breeds (47%). The Hanwoo results are comparable to the levels of
monounsaturation seen in Japanese and Murray Grey steers fed in Japan, whilst the levels in
the Australian animals were also comparable to those animals raised under Australian
conditions. There was some indication that there were differences in fatty acid composition
within the Australian breeds studied, with animals containing Jersey genes (J
h
LH, BH and HH animals. However, when intramuscular fat was added to the model, breed
was not significant for fatty acid. Despite this, the results showed that Jersey animals more
closely resembled the Hanwoo in fatty acid profile than Wagyu cross animals.
179
orean and Australian breeds. After
djusting for IMF%, the Hanwoo animals ranking’s often changed from being amongst the
aused some inconsistency and unreliability of the sensory
sults due to the fact that panellists did not have sufficient experience in detecting the
acids indicate that improved flavour
haracteristics were partly due to differences in fatty acid composition (particularly due to an
The Hanwoo in the present study displayed some unusual flavour characteristics, which were
different to those of the Australian breeds, most notably a numerically higher intensity of
chicken score and lower intensity of beef flavour. Unfortunately, this was not demonstrated
clearly in the statistical analysis possibly due to the confounding with the Hanwoo breed
having such a high level of IMF% in comparison to the Australian breeds, which may have
masked some of the true differences between the K
a
top scoring breeds for a particular flavour, to being not significantly different from the lowest
scoring breed. The flavour acceptability, by an Australian panel, for Hanwoo samples was
quite low (even after adjusting for tenderness and intramuscular fat content), which suggested
that it tasted distinctly different to Australian samples.
One of the major limiting factors of this trial was the fact that the resources of a fully trained
sensory panel was not able to be used. As a compromise, a semi-trained panel was utilised in
the main trial. This may have c
re
individual flavours. On the other hand, the scoring of initial and sustained juiciness, beef
flavour and flavour acceptability were quite consistent, suggesting that panellists were more
confident in scoring these attributes. This was reflected in the level of significance achieved
for breed in these latter flavours.
The correlations between flavours and fatty
c
increase in mono-unsaturated fatty acids, which is consistent with the literature). Despite this,
it was difficult to separate out the effect of individual fatty acids from the effect of
180
itively enhanced by increased levels of intramuscular fat,
F% was not the sole determinant of flavour acceptability. This may be partly due to
oo,
hich was significantly different for the majority of fatty acids (P<0.001), was also
significantly different from all of the Australian breeds for intramuscular fat. This meant that
there was a continual confounding effect between fatty acid composition and fat. Despite
this, intramuscular fat was contained in all prediction equations for individual flavours, which
demonstrates its’ importance in influencing individual flavours and flavour acceptability.
d for a period of 80 days. This meant that there was a continal confounding
etween sex and management (cohort) group. Despite this, the animals within the projects
intramuscular fat, which was also highly correlated with total mono-unsaturated fatty acids,
oleic acid and negatively related to palmitic acid.
Whilst flavour acceptability was pos
IM
increased fat altering the absolute amount of fatty acids and food texture. Breed groups were
significantly different for juiciness, beef flavour, buttery flavour and flavour acceptability,
even after adjusting data to a constant level of intramuscular fat, suggesting that some of the
variation in flavour may be genetic.
Throughout all of the results in this study, it was virtually impossible to separate the effects of
intramuscular fat from the effect of fatty acid composition, due to the fact that the Hanw
w
Another confounding factor was the influence of cohort group on the results. The DGM
steers and SXB steers were raised separately initially, however were fed on the same finishing
diet for 180 days and were slaughtered at the same time. These animals were older, had
heavier and fatter carcases than the SXB heifers which were a later drop from the SXB steers
and were only fe
b
were all reported separately throughout this thesis so that this could be taken into account
when drawing conclusions. In the final chapters 7 and 8, only purebred animals HH, LL and
181
Hanwoo animals compared to
Australian animals, no single ion was isolated as being of sole importance. For this reason,
principal components were formed from the ions and from this it was shown that the first
three principal components explained 99% of the variation between the meat samples and that
principal components 1, 3 and 8 were significant for breed. It was the principal components
from this analysis that were used in the prediction equations.
as the fact that moderate prediction equations were able to be derived for certain flavours
JJ steers were used in comparison to the Hanwoo, to try and eliminate some of this
confounding.
Use of a chemical sensor was able to establish significant differences between Korean
Hanwoo and Australian breeds (Jersey, Limousin and Hereford). These differences
predominantly mirrored the differences in fatty acid composition and to some extent flavour,
as demonstrated by the individual correlations between these traits ion abundance. Although
many ions stood out as being significantly different for
In the final chapter, all of the data was drawn together in an attempt to develop prediction
equations for the various flavours. It was disappointing that prediction equations using fatty
acid data alone was unable to predict individual flavours. However, based on the literature
explaining the complexities of beef flavour, it is hardly a surprising result. More heartening
w
when both fatty acid data, intramuscular fat and chemical sensor data were combined (R2
values of 24% to 43%). The flavours which were predicted with the most accuracy were those
that the panellists found most easy to score. This suggests that had the panellists been more
professionally trained and experienced, better prediction equations could have been achieved.
In summary, beef flavour was shown to be an extremely complex attribute to assess, however
throughout this study valuable information has been gained which will help add to the
182
bility of beef striploins. Additionally, increased levels of MUFA’s seem to have a
ositive effect on flavour as demonstrated by their positive correlations with individual
s a result of this trial, further work with ELDERS Limited and Adelaide University is aimed
t looking more closely at the backgrounding, nutritional and environmental effects on fatty
cid compositon, with the aim of increasing marbling and monounsaturated fatty acids of
ef.
knowledge base of the industry. This trial has clearly demonstrated that marbling
(intramuscular fat) is one of the biggest influences on individual flavours and flavour
accepta
p
desirable flavours and flavour acceptability. Another positive result has been that animals
containing Jersey have high levels of monounsaturated fatty acids, have a high degree of
marbling and produce highly palatable beef, which was superior to all other breeds in this
study.
A
a
a
be
184
APPENDIX 1
Figure 1 - Ambient chiller temperature and mean chilling rate of carcasses throughout 21 hour period
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
0 10 21
Tem
pera
ture
( 0 C)
Time Elapsed (hours)
Carcass MeanAmbient Chiller
Figure 2 - Ambient chiller and freezer temperatures during initial freezing and aging of samples (4 day period)
-25
-20
-15
-10
-5
0
5
10
2:54 PM 3:24 AM 3:54 PM 4:24 AM 4:54 PM 5:24 AM 5:54 PM 6:24 AM
Freezer TemperatureChiller Temperature
185
Table 1 - Treatment allocation for Preliminary Trial
Animal ID
Carcass Side
Ageing Tmt
Tmt Tmt Tmt
APPENDIX 1 (cont)
1 LS A 1 2 3 1
1 2 A
5 RS 3 1 2
6 1 2 3
1
LS A 9 RS F 10 LS F 10 RS A 3 1 2
RS F 3 1 2 2 LS F 2 3 1 2 RS A 1 2 3 3 LS A 3 1 2 3 RS F 2 3 1 4 LS F 1 2 3 4 RS A 3 5 LS 2 3 1
F 1 2 3 6 LS F
RS A 2 3 1 7 LS A 7 RS F 3 1 2 8 LS F 2 3 8 RS A 1 2 3 9 3 1 2
2 3 1 1 2 3
186
APPENDIX 2 for Ions ignificant for co
Australia Korea Signi nce
- LSMEANS s untry
fica
Ion 36 23371 22145 ** Ion 38 16164 17142 * Ion 39 25215 27724 * Ion 40 2880010 2660936 * Ion 42 336921 316370 ** Ion 44 6192992 5397061 ** Ion 45 1050490 929194 * Ion 47 30654 27825 ** Ion 48 55289 50754 ** Ion 49 12300 11704 * Ion 55 22762 28824 *** Ion 56 19345 25803 *** Ion 57 23266 32570 *** Ion 58 20586 23155 ** Ion 60 49314 41567 *** Ion 61 11588 10506 ** Ion 66 18715 10469 * Ion 67 9626 11443 *** Ion 69 11313 12998 *** Ion 70 11541 14401 *** Ion 71 10587 13629 *** Ion 72 10943 12145 ** Ion 82 9755 11060 *** Ion 83 8800 10206 ** Ion 85 9069 11338 *** Ion 86 *** 11030 15074 Ion 91 9105 11049 *** Ion 92 7545 8217 * Ion 94 11241 14520 * Ion 95 8013 8808 ***
A a Korea Significance ustrali
Ion 101 7350 6808 ** Ion 102 7085 6520 ** Ion 104 7142 6804 * Ion 107 6826 6434 * Ion 110 7190 6876 * Ion 117 7010 6669 * Ion 118 7155 6741 ** Ion 124 6761 6369 * Ion 125 6896 6392 ** Ion 126 7075 6461 ** Ion 130 6599 6280 * Ion 133 6684 6320 * Ion 137 6602 6298 * Ion 144 6529 6174 ** Ion 145 6528 6183 * Ion 148 6630 6145 ** Ion 149 6487 6168 ** Ion 152 6547 6128 * Ion 153 6448 6087 ** Ion 154 6585 6188 * Ion 156 6489 6045 ** Ion 160 6521 6131 * Ion 162 6447 6131 * Ion 163 6433 6111 * Ion 171 6472 6127 * Ion 175 6458 6078 ** Ion 177 6489 6066 * Ion 178 6397 6051 *
Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
187
16:1 18:1
APPENDIX 3 - Correlations of Ions with fatty acids and flavours
R2 14:0 14:1 16:0 18:0 18:1 (n-9c) (n-7t)
18:1 (n-7c)
18:2 MUFA IMF MPT Ijuice Sjuice Beef Beef fat
Oily Buttery Chicken Corn Rancid FlavAccept
Ion 36 14 28 0.26*
0.30 -0.26 * *
-0.35**
-0.31*
-0.29 *
0.28*
-0.*
Ion 38 14 -0.26 0.30* *
-0.30*
0.26 0.39 * **
Ion 39 13 0.36***
-0.27*
-0.32*
0.39**
0.27 *
Ion 40 18 0.*
-0.34 -0.37 **
29 0.29*
-0.29*
31 **
-0.48***
-0.37**
-0.**
Ion 42 18 0.28*
0.33*
0.28 -0.35* **
-0.27*
-0.37 -0.27** *
0.29 *
Ion 43 17 30 0.26 -0.*
*
0.39 0.54** ***
0.47***
Ion 44 17 0.26*
-0.37**
0.29 *
-0.33 *
-0.35 **
-0.30*
-0.49***
-0.39**
-0.29 *
-0.28*
Ion 45 13 -0 -0 7
.34**
.2 -0.29 * *
-0.26*
-0.43***
-0.31*
-0.28 *
Ion 47 21 41 0.33*
0.30*
-0.**
0.31 *
Ion 48 21
-0.26*
Ion 49 13 33 0.30*
0.29*
-0.*
0.26*
Ion 55 29
0.39 **
-0.27 *
-0.43***
0.38 **
-0.34**
0.37**
0.37 **
0.44***
-0.45 ***
0.30 *
0.30 0.29 * *
0.30*
0.29*
Ion 56 32
0.46 ***
-0.46***
0.37 **
-0.39**
0.27*
0.37 **
0.47***
-0.44 ***
0.29 *
0.26*
0.28*
0.37**
Ion 57 35
0.39 **
-0.31 *
-0.43***
0.44***
-0 .41**
0.26*
0.40 **
0.47***
-0.48 ***
0.33 *
0.27*
0.36**
0.42***
Ion 58 16
0.41 **
-0.31*
-0.32*
0.38 **
0.37**
R2 - amount of variation explained by the Model I36….I180 = country, country x project, breed x country x project. *P<0.05, ** P<0.01, ***P<0.001 Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
APPENDIX 3 (cont) - Correlations of Ions with fatty acids and flavours
R2 14:0 14:1 16:0 16:1 18:0 18:1(n-9c)
18:1 (n-7t)
18:1 (n-7c)
18:2 MUFA IMF MPT Ijuice Sjuice Beef Beef fat
Oily Buttery Chicken Corn Rancid FlavAccept
Ion 60 37 0.49 ***
0.59 ***
-0.42 0.54 30 0.26 -0.55* ***
-0.39**
-0.29*
-0.46*** *** ***
-0.*
-0.28*
-0.26 *
Ion 61 17 32 26 28 0.31*
0.*
-0.*
-0.26*
-0.*
Ion 66 13 -0 3 0.27 -0.26*
.4***
0.30*
0.36**
-0.33** *
Ion 67 24 -0.34 0.*
0 0.30 -0.31 * * **
32 .27 0.35 * **
0.30 *
0.38 **
-0.42 ***
0.30 0.32 * *
0.26*
Ion 69 21 0.27*
-0.26*
0.35 **
0.35 -0.36 ** **
0.30 *
0.26 *
0.28 *
0.26 *
Ion 70 30 0.34 -0.36 ** **
-0.40**
0.44***
-0.35**
0.32 *
0.42 ***
0.49***
-0.49 ***
0.28 *
0.26 0.33 * **
0.29*
0.28*
Ion 71 32 0.39**
-0.42***
0.42***
0.39 0.49** ***
-0.45 ***
0.32 *
0.26 0.27 * *
0.37**
0.43***
Ion 72 14 28 0.39**
-0.*
0.28 *
0.28*
0.39**
Ion 77 14
0.27*
Ion 78
12 0. 0.26*
0.28 0.28 * *
28 *
0.25 *
0.32 *
Ion 80
9 0.26*
Ion 82 19 0.40**
-0.36**
0.32 -0.27 * *
0.35**
Ion 83 15 37 0.27*
-0.28*
0.34 **
-0.32*
0.28 0.36 ** *
0.**
Ion 85 22 0. -0.33 *
40 **
-0.39**
0.35 **
-0.33**
0.35**
0.48***
0.34**
0.32*
Ion 86 -0**
0.36 34 -0.29 *
-0.48 ***
-0.35**
0.47***
.39 0.46 ***
0.40 **
-0.49 ***
0.28 * **
0.26*
0.27*
R2 - amount of variation explained by the M I36….I180 = country, c ry x oj bree o x project. *P<0.05, ** P<0.01, ***P<0.001
odel ount pr ect, d x c untryRed = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
188
APPENDIX 3 (cont) - Correlations of Ions with fatty acids and flavours
18:1 (n-9c)
Buttery Corn R2 14:0 14:1 16:0 16:1 18:0 18:1 18:1 18:2 MUFA IMF MPT Ijuice Sjuice Beef Beef fat
Oily Chicken Rancid Flav(n-7t) (n-7c) Accept
Ion 91 26 0.36 0.27 -0.28 0.28 0.32 *
-0.34 ** * * * **
Ion 92 9 27 -0.*
Ion 94 11
-0.43 ***
Ion 95 24 0.32 -0.37 0.34**
-0.32 0.34 0.26 -0.39**
0.28 0.35 0.26 0.26 ** * * * * * ** * *
Ion 101 19 0.35 29 -0.35**
-0.34**
0.33**
-0.* **
Ion 102 17 0.28 0.32 0.34 -0.34 0.37 **
-0.32 -0.31 -0.35 -0.28* ** * * * ** * **
Ion 104 11 27 -0. *
Ion 107 10 34 -0.27*
0.37 -0.32 -0.35 -0.28 0.34 -0. ** * ** * ** **
Ion 110 14
Ion 117 14 0.28*
Ion 118 16 32 0.26 -0.28* *
-0.28*
-0.35**
0.31*
-0.*
Ion 124
10 29 -0.27 -0.30 * *
0.28 *
-0.*
Ion 125 19 0. 9 -0.27 2 -0.27 * *
0.26 *
-0.27*
-0.34**
-0.30*
-0.30 *
0.31*
-0.30* *
-0.33*
R2 - amount of variation explained by the Model I36….I180 = country, country x project, breed x country x project. *P<0.05, ** P<0.01, ***P<0.001 Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
189
APPENDIX 3 (cont) - Correlations of Ions with fatty acids and flavours
18:1 Sjuice Rancid R2 14:0 14:1 16:0 16:1 18:0 18:1
(n-9c)18:1
(n-7t) (n-7c)18:2 MUFA IMF MPT Ijuice Beef Beef
fat Oily Buttery Chicken Corn Flav
Accept Ion 126 18 28 0.32
* 0.28
* -0.
*
Ion 130
10 26 -0.*
Ion 133 15 31 -0.27 0.26 * *
-0.28*
-0.*
Ion 137
8 26 -0.32*
-0.38 0.32 ** *
-0.26*
-0.27*
-0.*
Ion 144 15 -0.31*
-0.30*
0.27*
Ion 145
10 30 0.28*
-0.*
-0.27*
Ion 148 18 0.27 -0.28 0.28 * * *
Ion 149 17 0.35 0.33*
-0.32 -0.33* *
0.26 * **
Ion 152
11 -0.28 *
Ion 153 13 -0.25
-0.29 *
0.29 * *
0.25*
Ion 154 14
-0.29*
0.30*
Ion 156 14 35
-0.26 *
-0.**
Ion 160
9 26 -0.*
Ion 162 16
2 0. 6 0.27 * *
R2 - amount of variation explained by the Model I36….I180 = country, country x project, breed x country x project. *P<0.05, ** P<0.01, ***P<0.001 Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
190
APPENDIX 3 (cont) - Correlations of Ions with fatty acids and flavours
R2 14:0 14:1 16:0 16:1 18:0 18:1
(n-9c)18:1
(n-7t)18:1
(n-7c)18:2 MUFA IMF MPT Ijuice Sjuice Beef Beef
fat Oily Buttery Chicken Corn Rancid Flav
Accept Ion 163
12 29 -0.
*
Ion 171
10 30 -0.*
Ion 175
13 29 -0.29*
-0.28 0.28 * *
0.30*
-0.*
Ion 177
12 -0 6 0.29 -0.37**
.3 0.40 ** **
*
Ion 178
10 27 -0.31*
0.31 -0.26* *
-0.31*
-0.*
R2 - amount of variation explained by the Model I36….I180 = country, country x project, breed x country x project. *P<0.05, ** P<0.01, ***P<0.001 Red = P<0.0001 for country, Green = significant for project nested within country, Blue = significant for breed nested within project x country
191
192
APPENDIX 4 - Eigenvectors from the Principal Components Analysis showing the amount of variation explained by each ion making up Principal components 1 to 10.
I39 -0.0013 -0.0001 0.0094 -0.0329 0.1701 0.0879 -0.0395 0.0626 -0.2595 -0.2905I40 0.0732 0.9466 -0.3018 0.0470 0.0017 -0.0362 0.0303 -0.0321 0.0017 0.0252I41 0.0010 0.0463 0.0666 -0.0803 0.6308 -0.1869 -0.0162 0.2718 0.0042 -0.2313I42 0.0101 0.0755 0.0747 0.1135 0.0653 0.6867 -0.4962 0.2148 -0.2992 -0.0848I43 -0.0734 0.3023 0.8937 -0.2872 -0.1349 -0.0191 0.0334 -0.0299 0.0199 -0.0045I44 0.9821 -0.0536 0.0395 -0.1724 -0.0270 0.0176 0.0019 0.0007 -0.0009 -0.0049I45 0.1565 0.0292 0.3077 0.9123 0.1110 -0.1563 0.0271 -0.0205 -0.0030 0.0355I46 0.0057 0.0034 0.0228 0.0150 0.2226 0.2593 -0.1479 -0.4587 0.4753 -0.0200I47 0.0016 0.0054 0.0081 -0.0115 0.0770 0.0284 0.2213 0.3545 0.2571 -0.0622I48 0.0011 0.0131 0.0084 0.0202 0.1199 0.2905 0.3919 0.3363 0.2065 -0.0799I49 0.0004 0.0026 0.0016 0.0025 0.0241 -0.0042 0.0233 0.0612 0.0727 -0.0125I50 0.0002 0.0039 0.0047 0.0061 0.0453 0.0422 0.0013 0.0318 0.0756 -0.0336I51 -0.0002 0.0015 0.0042 -0.0049 0.0556 0.0467 -0.0159 -0.0253 0.0423 -0.0293I52 0.0001 0.0023 0.0018 -0.0005 0.0239 0.0000 -0.0515 -0.0285 0.0526 0.0018I53 -0.0003 0.0019 0.0043 -0.0178 0.0850 0.0055 -0.0809 0.0149 0.1297 0.1099I54 0.0002 0.0066 0.0035 -0.0152 0.0867 0.0358 -0.2902 -0.1146 0.3404 0.0626I55 -0.0020 -0.0017 0.0188 -0.0638 0.2718 -0.0154 -0.1508 -0.0035 0.0687 0.2028I56 -0.0019 -0.0001 0.0139 -0.0738 0.2405 -0.1264 -0.1300 0.0612 -0.0756 0.0827I57 -0.0028 -0.0053 0.0255 -0.0996 0.3450 -0.1648 -0.0160 -0.0366 -0.2787 0.4235I58 -0.0002 0.0041 0.0076 -0.0489 0.1172 -0.1102 -0.0336 0.0571 0.0388 -0.2503I59 0.0001 0.0029 0.0048 -0.0072 0.0156 -0.0393 -0.0239 0.0342 -0.0511 -0.0454I60 0.0031 0.0117 -0.0058 0.0246 -0.2086 -0.0616 -0.2842 0.5642 0.2577 0.3063I61 0.0004 0.0031 0.0024 0.0110 0.0007 -0.0161 -0.0217 0.0735 0.0626 0.0729I62 0.0002 0.0005 0.0046 -0.0040 0.0567 0.0338 0.0619 -0.0557 0.0931 0.0802I63 -0.0003 0.0009 0.0078 0.0062 0.0656 0.0821 0.1417 -0.0462 -0.0154 0.0037I64 -0.0004 0.0005 0.0142 0.0151 0.1200 0.3169 0.3741 -0.0145 0.0944 0.1045I65 -0.0004 0.0005 0.0075 0.0025 0.0699 0.0963 0.1125 -0.0391 -0.0206 -0.0056I66 -0.0005 -0.0017 0.0067 -0.0029 0.0395 0.0989 0.0958 -0.0432 -0.0874 0.0672I67 -0.0006 -0.0001 0.0061 -0.0208 0.0892 0.0196 -0.0137 -0.0345 0.0194 0.0891I68 -0.0002 0.0015 0.0039 -0.0105 0.0697 0.0510 -0.0603 -0.0737 0.0944 0.0377I69 -0.0007 0.0001 0.0058 -0.0180 0.0948 0.0317 -0.0593 -0.0882 0.0142 0.0991I70 -0.0010 -0.0007 0.0093 -0.0241 0.1203 -0.0064 -0.0520 -0.0252 0.0141 0.1426I71 -0.0009 0.0002 0.0083 -0.0307 0.1059 -0.0910 -0.0081 -0.0258 -0.1484 0.1583I72 -0.0003 0.0013 0.0043 -0.0206 0.0585 -0.0619 -0.0159 0.0411 0.0047 -0.0192I73 0.0001 0.0017 0.0020 -0.0015 0.0025 -0.0151 0.0004 0.0290 -0.0511 -0.0581I74 0.0001 0.0013 0.0012 0.0051 0.0085 -0.0013 -0.0092 0.0078 0.0254 -0.0107I75 0.0001 0.0012 0.0010 0.0078 0.0071 0.0149 0.0022 -0.0123 0.0095 -0.0144I76 0.0001 0.0015 0.0018 0.0036 0.0137 0.0095 0.0083 -0.0052 -0.0235 -0.0034I77 -0.0003 0.0009 0.0040 0.0035 0.0245 0.0403 -0.0107 -0.0257 0.0504 0.0517I78 -0.0001 0.0007 0.0029 0.0039 0.0247 0.0283 -0.0035 -0.0262 0.0296 0.0228I79 0.0002 0.0006 0.0010 -0.0022 0.0368 -0.0076 -0.0043 0.0154 0.0680 0.0559I80 0.0001 0.0008 0.0031 0.0072 0.0343 0.0222 0.0265 -0.0013 0.0704 -0.0322I81 -0.0006 0.0001 0.0049 -0.0163 0.0763 0.0132 0.0113 0.0093 -0.0638 0.1140I82 -0.0003 0.0004 0.0039 -0.0158 0.0555 -0.0227 -0.0652 0.0129 0.1079 0.0000I83 -0.0004 0.0014 0.0013 -0.0273 0.0678 -0.0572 0.0316 0.0216 0.0185 -0.0238
Prin1 Prin2 Prin3 Prin4 Prin5 Prin6 Prin7 Prin8 Prin9 Prin10I35 0.0005 0.0043 -0.0010 0.0075 0.0256 -0.0101 0.0675 0.0425 0.0784 0.0227I36 0.0005 0.0059 -0.0012 0.0100 0.0253 0.0141 0.0213 0.0253 -0.0024 0.0065I37 0.0001 0.0039 0.0028 -0.0009 0.0468 0.0043 -0.0097 0.0134 0.0457 -0.0183I38 -0.0001 0.0033 0.0048 -0.0109 0.0565 -0.0271 -0.0278 -0.0473 0.1031 -0.1430
193
APPENDIX 4 (cont) - Eigenvectors from the Principal Components Analysis showing the amount of variation explained by each ion making up Principal components 1 to 10.
Prin1 Prin2 Prin3 Prin4 Prin5 Prin6 Prin7 Prin8 Prin9 Prin10I84 0.0000 0.0014 0.0020 -0.0048 0.0242 -0.0113 0.0230 -0.0045 0.0200 0.0399I85 -0.0005 0.0014 0.0029 -0.0286 0.0932 -0.1045 -0.0471 0.0401 0.0264 -0.0518I86 -0.0009 -0.0077 0.0229 -0.0087 0.0561 0.0648 0.0737 0.0415 0.0784 0.2290I87 0.0001 0.0012 0.0021 0.0018 0.0070 -0.0149 0.0098 0.0148 -0.0073 -0.0603I88 0.0002 0.0012 0.0016 0.0093 -0.0038 -0.0051 -0.0021 0.0063 -0.0174 -0.0146I89 0.0001 0.0013 0.0020 0.0051 0.0106 0.0025 0.0050 -0.0047 -0.0165 -0.0099I90 0.0000 0.0016 0.0006 0.0053 0.0050 0.0008 0.0055 -0.0091 -0.0009 -0.0285I91 -0.0002 0.0006 0.0016 -0.0192 0.0822 -0.0420 -0.0356 -0.1051 0.0228 -0.2780I92 0.0000 0.0012 0.0003 -0.0085 0.0410 -0.0271 0.0074 -0.0236 -0.0203 -0.2174I93 0.0001 0.0018 0.0015 0.0042 0.0053 -0.0028 -0.0021 -0.0173 -0.0147 -0.0342I94 -0.0019 -0.0083 0.0156 -0.0110 0.0932 0.2574 0.2802 -0.0737 -0.2020 0.2726I95 -0.0003 -0.0004 0.0052 -0.0031 0.0309 0.0227 0.0037 -0.0178 -0.0240 0.0420I96 0.0000 0.0010 0.0030 -0.0009 0.0185 0.0058 -0.0234 -0.0117 0.0145 -0.0139I97 -0.0001 0.0008 0.0027 0.0027 0.0226 0.0177 -0.0038 -0.0382 0.0145 -0.0258I98 0.0001 0.0013 0.0021 0.0003 0.0244 0.0112 -0.0244 -0.0275 -0.0182 0.0356I99 0.0000 0.0008 0.0010 0.0027 0.0059 -0.0060 0.0052 -0.0084 -0.0253 0.0081I100 0.0001 0.0015 0.0008 0.0032 0.0070 -0.0014 0.0099 -0.0144 -0.0077 -0.0412I101 0.0001 0.0019 -0.0009 0.0053 -0.0010 -0.0004 -0.0006 -0.0149 -0.0020 -0.0077I102 0.0001 0.0016 -0.0006 0.0060 0.0004 0.0074 -0.0037 -0.0041 0.0194 -0.0238I103 0.0000 0.0014 0.0020 0.0058 0.0087 0.0158 0.0016 -0.0099 -0.0321 0.0075I104 0.0000 0.0013 0.0005 0.0060 0.0038 0.0122 0.0008 0.0093 -0.0105 -0.0278I105 0.0001 0.0013 0.0026 0.0029 0.0117 0.0050 0.0042 -0.0121 0.0208 0.0021I106 -0.0001 0.0009 0.0006 0.0034 0.0144 0.0186 -0.0105 -0.0282 0.0288 -0.0064I107 0.0001 0.0014 0.0008 0.0048 0.0062 0.0142 0.0140 0.0109 -0.0545 -0.0245I108 0.0001 0.0015 -0.0009 0.0045 0.0036 0.0044 -0.0181 -0.0232 0.0177 -0.0130I109 0.0002 0.0012 -0.0006 0.0018 0.0057 0.0037 0.0051 -0.0092 0.0198 0.0193I110 0.0001 0.0015 -0.0002 0.0045 0.0009 -0.0015 0.0057 0.0168 -0.0162 0.0297I111 0.0001 0.0018 0.0019 0.0050 0.0170 -0.0110 0.0158 -0.0128 -0.0565 0.0322I112 0.0001 0.0015 0.0019 0.0025 0.0079 0.0050 0.0026 -0.0066 -0.0255 -0.0229I113 0.0000 0.0013 0.0009 0.0033 0.0042 0.0059 0.0055 -0.0060 -0.0120 0.0397I114 0.0001 0.0013 0.0018 0.0033 0.0091 0.0072 0.0107 0.0063 0.0106 -0.0506I115 0.0000 0.0016 0.0013 0.0060 0.0047 0.0070 -0.0055 -0.0135 0.0154 0.0052I116 0.0001 0.0010 0.0021 0.0063 0.0067 -0.0007 0.0077 -0.0181 -0.0336 -0.0049I117 0.0001 0.0014 0.0017 0.0050 0.0054 0.0043 0.0080 0.0108 -0.0121 -0.0052I118 0.0001 0.0014 0.0000 0.0023 0.0022 0.0009 0.0037 0.0001 0.0040 0.0156I119 0.0001 0.0016 0.0021 0.0033 0.0088 0.0020 0.0146 -0.0258 -0.0056 -0.0100I120 0.0002 0.0016 0.0008 0.0058 0.0126 -0.0007 -0.0087 -0.0086 -0.0233 0.0346I121 0.0001 0.0009 0.0011 0.0075 0.0054 0.0097 0.0034 -0.0391 -0.0295 -0.0276I122 0.0001 0.0016 0.0008 0.0080 0.0073 0.0072 0.0012 0.0023 -0.0186 0.0333I123 0.0002 0.0014 0.0000 0.0058 0.0082 -0.0027 0.0072 -0.0081 -0.0379 -0.0209I124 0.0001 0.0013 0.0007 0.0049 0.0022 -0.0005 0.0117 0.0007 0.0003 0.0115I125 0.0002 0.0015 -0.0012 0.0056 -0.0043 0.0006 0.0077 -0.0077 -0.0077 -0.0167I126 0.0002 0.0017 -0.0003 0.0048 0.0026 0.0128 -0.0005 -0.0140 0.0026 0.0142I127 0.0001 0.0014 0.0001 0.0047 0.0051 -0.0067 0.0256 -0.0135 -0.0190 0.0261I128 0.0001 0.0012 0.0011 0.0056 0.0059 0.0063 0.0099 -0.0129 -0.0529 -0.0286I129 0.0002 0.0011 0.0010 0.0067 0.0086 0.0012 0.0124 -0.0136 -0.0102 -0.0104I130 0.0001 0.0011 0.0009 0.0028 0.0015 0.0099 0.0039 -0.0043 -0.0225 -0.0083I131 0.0001 0.0010 0.0019 0.0064 0.0052 0.0081 0.0187 0.0125 -0.0088 0.0069I132 0.0002 0.0013 0.0015 0.0052 0.0039 -0.0024 0.0066 -0.0098 0.0205 0.0210
194
APPENDIX 4 (cont) - Eigenvectors from the Principal Components Analysis showing the amount of variation explained by each ion making up Principal components 1 to 10.
0.0013 0.0157 -0.0039 0.0214 -0.0136I135 0.0001 0.0014 0.0005 0.0054 0.003 0.0063 0.0145 -0.0103 -0.0086 -0.0016I136 0.0002 0.0012 0.0015 0.0036 -0.0001 -0.0056 0.0084 -0.0174 -0.0041 -0.0205I137 0.0001 0.0013 -0.0370 -0.0036I138 0.0001 0.0017 0.0020 0.0123I139 0.0001 0.0013 -0.0006 0.0160I140 0.0001 0.0013 -0.0426 -0.0043I141 0.0001 0.0013 0.0010 0.0050 0.0060 0.0038 0.0199 -0.0128 -0.0191 0.0097I142 0.0000 0.0013 0.0008 0.0066 0.0102 0.0027 0.0087 -0.0151 -0.0020 -0.0071I143 0.0001 0.0014 0.0017 0.0058 0.002 0.0014 0.0129 -0.0081 -0.0396 -0.0039I144 0.0001 0.0013 -0.0001 0.0035 0.002 0.0043 -0.0012 -0.0004 0.0101 -0.0159I145 0.0001 0.0016 0.0005 0.0058 0.003 -0.0011 0.0071 0.0169 0.0053 0.0125I146 0.0001 0.0010 0.0011 0.0048 0.0006 0.0006 0.0070 -0.0165 -0.0246 0.0043I147 0.0000 0.0015 -0.0003 0.0055 0.0038 0.0005 0.0135 -0.0237 -0.0128 -0.0271I148 0.0002 0.0015 0.0015 0.0054 0.0048 0.0088 0.0043 0.0117 0.0087 0.0004I149 0.0002 0.0011 0.0006 0.0062 0.0014 0.0035 -0.0001 0.0034 0.0212 0.0170I150 0.0001 0.0013 0.0017 0.0067 0.0040 0.0072 0.0160 -0.0039 -0.0326 -0.0326I151 0.0001 0.0016 0.0018 0.0049 0.0067 -0.0060 0.0059 -0.0268 -0.0095 0.0304I152 0.0003 0.0012 0.0016 0.0078 0.0020 0.0012 0.0062 -0.0072 -0.0032 0.0206I153 0.0002 0.0012 0.0005 0.0052 0.0057 0.0036 0.0132 0.0033 0.0226 0.0066I154 0.0002 0.0014 0.0012 0.0070 0.0069 -0.0015 0.0139 0.0009 -0.0091 -0.0170I155 0.0002 0.0011 0.0009 0.0052 0.0046 0.0039 0.0068 -0.0154 -0.0013 -0.0010I156 0.0001 0.0014 0.0011 0.0063 0.0029 0.0015 0.0041 0.0054 -0.0221 -0.0174I157 0.0000 0.0012 0.0013 0.0044 0.0017 0.0009 0.0076 -0.0122 -0.0288 0.0049I158 0.0001 0.0014 0.0012 0.0052 0.0078 -0.0022 0.0155 -0.0070 -0.0320 0.0254I159 0.0001 0.0012 0.0020 0.0073 -0.005 -0.0005 0.0109 0.0005 -0.0156 0.0042I160 0.0001 0.0013 0.0002 0.0051 0.007 -0.0007 -0.0056 0.0092 -0.0174 -0.0063I161 0.0001 0.0015 0.0011 0.0029 0.001 0.0032 0.0094 -0.0084 0.0152 -0.0089I162 0.0001 0.0014 0.0006 0.0058 0.002 0.0069 0.0051 -0.0101 0.0069 -0.0149I163 0.0001 0.0014 0.0017 0.0074 0.0053 -0.0028 0.0140 0.0302 0.0028 -0.0352I164 0.0000 0.0016 0.0015 0.0045 -0.0002 0.0020 0.0127 0.0044 0.0153 -0.0302I165 0.0001 0.0013 0.0016 0.0058 -0.000 0.0008 -0.0092 0.0095 -0.0047 -0.0002I166 0.0001 0.0015 0.0009 0.0062 0.004 -0.0012 0.0252 -0.0016 -0.0054 0.0075I167 0.0001 0.0017 0.0013 0.0045 -0.000 0.0094 0.0070 -0.0071 -0.0166 -0.0223I168 0.0000 0.0015 0.0005 0.0051 0.0037 0.0019 0.0069 -0.0216 -0.0268 -0.0073I169 0.0001 0.0014 0.0015 0.0051 0.0074 0.0077 0.0085 -0.0015 -0.0038 -0.0140I170 0.0001 0.0015 0.0014 0.0079 0.009 0.0112 0.0127 -0.0043 -0.0091 0.0027I171 0.0002 0.0012 0.0015 0.0074 -0.001 0.0070 0.0079 0.0135 -0.0186 -0.0112I172 0.0001 0.0013 0.0015 0.0064 0.005 0.0004 0.0191 -0.0118 0.0115 -0.0428I173 0.0001 0.0017 0.0001 0.0075 0.003 -0.0044 0.0029 -0.0192 -0.0180 -0.0267I174 0.0001 0.0017 0.0011 0.0056 0.0073 0.0084 0.0125 -0.0135 -0.0168 0.0008I175 0.0001 0.0014 0.0000 0.0055 0.0017 -0.0010 0.0205 0.0023 -0.0168 -0.0209I176 0.0001 0.0013 0.0006 0.0068 0.004 -0.0075 0.0048 -0.0025 -0.0308 -0.0470I177 0.0002 0.0014 0.0010 0.0075 -0.002 0.0136 0.0077 0.0001 0.0041 -0.0111I178 0.0001 0.0013 0.0001 0.0049 -0.001 0.0054 0.0065 -0.0133 -0.0046 0.0136I179 0.0001 0.0012 0.0006 0.0048 0.0004 -0.0016 0.0127 0.0053 -0.0047 -0.0386I180 0.0001 0.0009 0.0025 0.0064 0.0071 0.0068 0.0208 -0.0023 -0.0136 0.0180
Prin1 Prin2 Prin3 Prin4 Prin5 Prin6 Prin7 Prin8 Prin9 Prin10I133 0.0002 0.0014 0.0007 0.0036 -0.0040 -0.0027 0.0050 -0.0116 -0.0138 -0.0043I134 0.0001 0.0012 0.0011 0.0050 0.0013
7
-0.0005 0.0067 0.0063 -0.0040 0.0181 -0.00060.0018 0.0043 0.0011 0.0013 -0.0097 0.01200.0007 0.0055 0.0107 0.0018 0.0049 -0.00560.0006 0.0067 0.0036 0.0030 0.0171 -0.0009
183
8001
916
9856
527
196
Aalhus, J.L., Jones, S.D.M., Lutz, S., Best, D.R. and Robertson, W.M. (1994) The Efficacy of High and Low Voltage Electrical Stimulation under Different Chilling Regimes, Canadian ournal of Animal Science, 74 (3) : 433 - 442
A , 26 : 40 - 49
A ration Alarcon-Rojo, A.D. (1995) Alteration of Post-mortem Ageing in Beef by the Addition of Enzyme
Inhibitors and Activators, Meat Science, 41 (2): 163 - 178
rlag, New York, pp 218 - 227
, Springer-Verlag, New York, pp 218 - 227
merine, M.A., Pangborn, R.M. and Roessler, E.B. (1965a) Consumer Studies, In : Principles of Sensory Evaluation of Food, Academic Press., New York. pp 398 - 436
Amerine, M.A., Pangborn, R.M. and Roessler, E.B. (1965b) Laboratory Studies : Types and
Principles, In : Principles of Sensory Evaluation of Food, Academic Press., New York. pp 275
AMLC (1996) Beef Exports Korea, 1996, AMLC estimates
ivestock Grading Bureau, AMLC Seoul.
Boar Taint’ using a Sensory Panel and an Electronic Nose, Meat Science, 50 (2) : 139 -151
A ages, Food Technology, 18 (8) : 25 - 31
Art.
J
bbott, J.A. (1972) Sensory Assessment of Food Texture, Food Technology
gri-Marine Products Team (1996) Korea Meat Market, Ssangyong Corpo
Altner, H. (1978a) Physiology of Taste - Ch. 7, In: Fundamentals of Sensory Physiology,
Schmidt, R.F. ed., Springer-Ve Altner, H. (1978b) Physiology of Olfaction - Ch. 8, In: Fundamentals of Sensory Physiology,
Schmidt, R.F. ed. American Meat Institute Foundation (1960) Official Method Cc 3b-92 American OilChemist’s Society AOCS (1993) Beef Aroma, Some Volatile Constituents of
Cooked Beef, Journal of Agricultural and Food Chemistry, 8 (6) : 491 - 494
A
- 321
AMLC (1995) Korean Beef and Pork Grading Systems, Extract from : Implementation of Korean
Livestock Grading Systems, Ministry of Forestry, Fisheries and Agriculture and National Livestock Co-operatives Federation, L
Annor-Frempong, I.E., Nute, G., Wood, J.D., Whittington, F.W. and West. A. (1998) The
Measurement of the Responses to Different Odour Intensities of ‘
nonymous (1964) Sensory Testing Guide for Panel Evaluations of Foods and Bever
RC (1997-1999) Australian Research Council Grant Proposal : Genetics of Beef Quality Traits Fat Metabolism, awarded to Pitchford, Bottema and Siebe
Awad, A., Powrie, W.D. and Fennema, O. (1968) Chemical Deterioration of Frozen Bovine Muscle at -4°C, Journal of Food Science, 33 : 227 - 235
197
aghurst, K. (1997) Attitudes of Korean People to Protein-Based Foods and Their Relationship
Bailey, M.E. (1994) Maillard reactions and meat flavour development, In : Flavor of Meat and
Meat Products, Shahidi, F (Ed), Blackie, London, pp 155 - 173
B Noses and their Application in the Food Industry, Food Technology, 51 (12) : 44 - 48
B of Clinical Nutrition, 31 : 1068 - 1077
Bartoshuk, L.M. (1980) Separate Worlds of Taste, Psychology Today, 14 (4) : 48 - 57
at Science, 41 (1) : 1 - 5
Batzer, O.F., Santoro, T., Tan, M.C., landmann, W.A. and Schweigert, B.S. (1960) Meat Flavor
Chemistry, Precursors of Beef Flavor, Journal of Agricultural and Food Chemistry, 8 (6) : 498
eilken, S.L., Bouton, P.E. and Harris, P.V. (1986) Some Effects on the Mechanical Properties of
Beilken, S.L., Eadie, L.M., Griffiths, I., Jones, P.N. and Harris, P.V. (1991) Assessment of the
Textural Quality of Meat Patties: Correlation of Instrumental and Sensory Attributes, Journal of Food Science, 56 (6) : 1465 - 1469
erry, B.W. and Leddy, K.F. (1984) Effects of Fat Level and Cooking Method on Sensory and
erry, B.W. and Leddy, K.F. (1990a) Influence of Steak Temperature at the Beginning of
-298
erry, B.W., Maga, J.A., Calkins, C.R., Wells,L.H., Carpenter, Z.L. and Cross, H.R. (1980) Flavor Profile Analyses of Cooked Beef Loin Steaks, Journal of Food Science, 45 : 1113 - 1121
Bto Health, AMLC Trade Seminar - Beef and Human Nutrition, pp 3 – 16
artlett et al, (1997) Electronic
artoshuk, L.M. (1978) The Psychophysics of Taste, The American Journal
Barker, B.P., Mies, W.L., Turner, J.W., Lunt, D.K. and Smith, S.B. (1995) Influence of
Production System on Carcass Characteristics of F1 Wagyu x Angus Steers and Heifers, Me
- 501
BMeat Produced by Cooking at Temperatures between 500 and 600 C, Journal of Food Science, 51 (3) : 791 - 796
Beilken, S.L., Eadie, L.M., Jones, P.N. and Harris, P.V. (1990) Quality Assessment of
Manufactured Meat Products, Food Australia, 42 (11) : 532 - 536
BTextural Properties of Ground Beef Patties, Journal of Food Science, 49: 870 - 875
BBroiling on Palatability, Shear and Cooking Properties of Beef Loin Steaks Differing in Marbling, Journal of Foodservice Systems, 5 : 287
Berry, B.W. and Leddy, K.F. (1990b) Comparison of Restaurant vs Research-Type Broiling with
Beef Loin Steaks Differing in Marbling, Journal of Animal Science, 68 : 666 - 672
B
198
Bouton, P.E. and Harris, P.V. (1972a) The Effects of Cooking Temperature and Time on some
outon, P.E. and Harris, P.V. (1972b) A Comparison of Some Objective Methods used to Assess
B me Post-Slaughter Treatments on the
Mechanical Properties of Bovine and Ovine Muscle, Journal of Food Science, 37 : 539 - 543
Bdies, 9 (2) : 395 - 493
4 - 407
Journal of Food Technology, 8 : 39 - 49
Burnal of Texture Studies, 6 (3) : 315 -328
outon, P.E., Ford, A.L., Harris, P.V. and Shaw, F.D. (1980) Electrical Stimulation of Beef
B
ences, Meat Science, 2 : 301 - 311
Bourne, M.C. (1966) Measure of Shear and Compression Components of Puncture Tests, Journal of Food Science, 31 : 282 - 290
Mechanical Properties of Meat, Journal of Food Science, 37 : 140 - 144
BMeat Tenderness, Journal of Food Science, 37: 218 - 221
outon, P.E. and Harris, P.V. (1972c) The Effects of so
outon, P.E. and Harris, P.V. (1978) Factors Affecting Tensile and Warner-Bratzler Shear Values of Raw and Cooked Meat, Journal of Texture Stu
Bouton, P.E. and Harris, P.V. (1981) Changes in the Tenderness of Meat Cooked at 500 - 650 C, Journal of Food Science, 46 (2) : 475 - 478
Bouton, P.E., Carroll, F.D., Fisher, A.L., Harris, P.V. and Shorthose, W.R. (1973c) Effect of
Altering Ultimate pH on Bovine Muscle Tenderness, Journal of Food Science, 38 : 816 – 820
Bouton, P.E., Carroll, F.D., Harris, P.V. and Shorthose, W.R. (1973b) Influence of pH and Fiber Contraction State upon Factors Affecting the Tenderness of Bovine Muscle, Journal of Food Science, 38 : 40
Bouton, P.E., Fisher, A.L., Harris, P.V. and Baxter, R.I. (1973a) A Comparison of the Effects of
some Post-slaughter Treatments on the Tenderness of Beef,
outon, P.E., Ford, A.L., Harris, P.V. and Ratcliff, D. (1975a) Objective - Subjective Assessment of Meat Tenderness, Jo
Bouton, P.E., Ford, A.L., Harris, P.V. and Ratcliff, D. (1975b) Objective Assessment of Meat Juiciness, Journal of Food Science, 40 : 884 - 885
Bouton, P.E., Ford, A.L., Harris, P.V. and Shaw, F.D. (1978b) Effect of Low Voltage Stimulation
of Beef Carcasses on Muscle Tenderness and pH, Journal of Food Science, 43 (5) : 1392 - 1396
BSides, Meat Science, 4 : 145 - 155
outon, P.E., Ford, A.L., Harris, P.V., Shorthose, W.R., Ratcliff, D. and Morgan, J.H.L. (1978a) Influence of Animal Age on the Tenderness of Beef : Muscle Differ
199
Bouton, P.E., Harris, P.V. and Shorthose, W.R. (1973d) Changes in Shear Parameters of Meat
outon, P.E., Harris, P.V. and Shorthose, W.R. (1975c) Changes in Shear Parameters of Meat
outon, P.E., Harris, P.V. and Shorthose, W.R. (1976a) Dimensional Changes in Meat During
B al Contraction of Meat During
Cooking and its Possible Influence on Tenderness, Journal of Texture Studies, 7 : 193 - 203
Bouton, P.E., Harris, P.V. and Shorthose, W.R. (1976c) Peak Shear-Force Values Obtained for Veal Muscle Samples Cooked at 50 and 600C : Influence of Aging, Journal of Food Science,
Bouton, P.E., Harris, P.V. and Shorthose, W.R. (1976d) Factors Influencing Cooking Losses
B erature and Ultimate
pH on the Increase in Meat Toughness Resulting from Restraint During Cooking, Meat
outon, P.E., Harris, P.V., Macfarlane, J.J. and Snowden, J.McK. (1977) The Effects of Long
B osses and Eating Quality
of Further Pre-Slaughter Treatments, In : Studies on Beef Quality - Part VI, Technical Paper rch
Organisation, Melbourne, pp 5 - 23
B Carcasses, In : Studies on Beef
Quality - Part VII, Technical Paper no.8 for Div. of Food Preservation and Transport, Comm.
rms, Journal of Food Science, 45 (1) : 148 - 149
Bouton, P.E., Harris, P.V. and Ratcliff, D. (1981) Effect of Cooking Temperature and Time on the Shear Properties of Meat, Journal of Food Science, 46 (4) : 1082 - 1087
Associated with Structural Changes Produced by Aging, Cooking and Myofibrillar Contraction, Journal of Food Science, 40 : 1122 - 1126
BAssociated with Structural Changes Produced by Aging, Cooking and Myofibrillar Contraction, Journal of Food Science, 40 : 1122 - 1126
BCooking, Journal of Texture Studies, 7 : 179 - 192
outon, P.E., Harris, P.V. and Shorthose, W.R. (1976b) Therm
41 : 197 - 198
from Meat, Journal of Food Science, 41 (5) : 1092 - 1095
outon, P.E., Harris, P.V. and Shorthose, W.R. (1982) The Effect of Temp
Science, 6 : 235 - 241
BTerm Cooking on Shear Force-Deformation Curves, Journal of Texture Studies, 8 : 297 - 310
outon, P.E., Howard, A. and Lawrie, R.A. (1957) Effects on Weight L
no.6 for Div. of Food Preservation and Transport, Comm. Scientific and Industrial Resea
outon, P.E., Howard, A. and Lawrie, R.A. (1958) The Influence of Certain Holding Conditions on Weight Losses and Eating Quality of Fresh and Frozen Beef
Scientific and Industrial Research Organisation, Melbourne, pp 5 - 31 Bouton, P.E., Weste, R.R. and Shaw, F.D. (1978c) Electrical Stimulation of Calf Carcasses :
Response of Various Muscles to Different Wavefo
200
Bowers, J.A., Craig, J.A., Kropf, D.H. and Tucker, T.J. (1987) Flavor, Color, and Other Characteristics of Beef Longissimus Muscle Heated to Seven Internal Temperatures Between 550 and 850 C, Journal of Food Science, 52 (3) : 533 - 536
(1995) Lipid Content and Composition of Wagyu and Domestic Breeds of Beef, Journal of
oylston, T.D., Morgan, S.A., Johnson, K.A.,Wright Jr, R.W., Busboom, J.R., and Reeves, J.J.
: 1091 - 1095
tion, Biometrics, 9 : 22 - 38
B
raggins, T.J. and Frost, D.A. (1997) The Effect of Extended Chilled Storage in CO2 Atmosphere
rd ICOMST - Vitality of Meat, Auckland, New Zealand, 198 - 199
usboom, J.R., Jeremiah, L.E., Gibson, L.L., Johnson, K.A., Gaskins, C.T., Reeves, J.J. and
1 - 258
amfield, P.K. Brown, A.H.Jnr., Lewis, P.K., Rakes, L.Y. and Johnson, Z.B. (1997) Effects of
aul, J.F. (1957) The Profile Method of Flavour Analysis. Advances in Food Research, 7 : 1 - 40
C ation of Forage and Grain-Fed Beef, Journal of Food Science, 47 : 340 - 341
C
Cheraghi, T. and Roozen, J.P. (1994) Lipid Oxidation in Meat By-Products : Effect of Antioxidants and Maillard Reactants on Volatiles, In : Flavor of Meat and Meat Products, Shahidi, F (Ed), Blackie, London, pp 144 - 152
Boylston, T.D., Morgan, S.A., Johnson, K.A., Busboom, J.R., Wright, R.W. and Reeves, J.J.
Agricultural Food Chemistry, 43 : 1202 - 1207
B(1996) Volatile Lipid Oxidation Products of Wagyu and Domestic Breeds of Beef, Journal of Agricultural Food Chemistry, 44
Bradley,R.A. (1953) Some Statistical Methods in Taste Testing and Quality Evalua
rady, P.L. and Hunecke, E. (1985) Correlations of Sensory and Instrumental Evaluations of Roast Beef Texture, Journal of Food Science, 50 : 300 - 303
Bon the Odour and Flavour of Sheepmeat as Assessed by Sensory Panel and an Electronic Nose, Proc. 43
Buck, E.M., Hickey, A.M. and Rosenau, J. (1979) Low Temperature Air Oven vs a Water Bath
for the Preparation of Rare Beef, Journal of Food Science, 44 : 1602 - 1605
BWright, R.W. (1993) Effects of Biological Source on Cooking and Palatability Attributes of Beef Produced for the Japanese Market, Meat Science, 35 : 24
Callow, E.H. (1952) Frozen Meat, Journal of Science of Food and Agriculture,, 3 : 145 - 150
CFrame Size and Time-on-Feed on Carcass Characteristics, Sensory Attributes, and Fatty Acid Profiles of Steers, Journal of Animal Science, 75 : 1837 - 1844
C
hastain, M.F., Huffman, D.L. and Bertram, S. (1981) Sensory Evalu
heng, C. and Parrish, F.C. (1979) Heat-Induced Changes in Myofibrillar Proteins of Bovine Longissimus Muscle, Journal of Food Science, 44 (1) : 22 - 24
201
C . Proc. of the New Zealand
Society of Animal Production, 56 : 172 – 179
Civille, G.V. and Szczesniak, A.S. (1973) Guidelines to Training a Texture Profile Panel, Journal
ohen, T. (1984) Aging of Frozen Parts of Beef, Journal of Food Science, 49 : 1174 - 1177
C ducing and Processing Quality Beef from Australian Cattle Herds – Industry Outcomes of the Cooperative Research Centre for the Cattle and Beef Industry (Meat Quality) : 1993 – 2000, Dundon, P., Sundstrom, B. and Gaden, B.
richton, N. (1996) Executive Summary from Study Tour to Korea, ELDERS Limited, pers
ross, H.R., Bernholdt, H.F., Dikeman, M.E., Greene, B.E., Moody, W.G., Staggs, R. and West,
ross, H.R., Moen, R. and Stanfield, M.S. (1978b) Training and Testing of Judges for Sensory
Cross, H.R., Stanfield, M.S. and Koch, E.J. (1976) Beef Palatability as Affected by Cooking Rate
and Final Internal Temperature, Journal of Animal Science, 43 (1) : 114 - 121
C mparison of Roasting versus Broiling on the Sensory Characteristics of Beef Longissimus Steaks, Journal of Food
rouse, J.D. and Koohmaraie, M. (1990) Effect of Freezing of Beef on Subsequent Postmortem
rouse, J.D., Cundiff, L.V., Koch, R.M., Koohmaraie, M. and Seideman, S.C. (1989)
CSIRO (1995) South Korea - Red meat into Asia, MRC project CS 260 (Phase 1)
Dss Index in Adult Females,
Journal of the American Dietetic Association, 95 (9) : Supp. A-41
Dand, New Zealand, 46 - 50
Christie, W.W (1989) Gas Chromatography and Lipids, The Oily Press, UK
hrystall, B.B. and Daly, C.C, (1996) Processing for Meat Quality
of Texture Studies, 4 : 204 - 223
C
RC for the Cattle and Beef Industry – Meat Quality (2000) Pro
Ccomm.
CR.L. (1978a) Guidelines for Cookery and Sensory Evaluation of Meat, AMSA (American Meat Science Association), Chicago, IL, pp 1 - 24
CAnalysis of Meat Quality, Food Technology, 32 (7) : 48 - 54
ross, H.R., Stanfield, M.S., Elder, R.S. and Smith, G.C. (1979) A Co
Science, 44 (1) : 310 -311
CAging and Shear Force, Journal of Food Science, 55 (2) : 573 - 574
CComparisons of Bos Indicus and Bos Taurus Inheritance for Carcass Beef Characteristics and Meat Palatability, Journal of Animal Science, 67 : 2661 - 2668
abrila, G.M., Storrs, C.T., Bartoshuk, L.M. and Duffy, V.B. (1995) Preliminary Findings of Genetic Taste Status Association with Fat Intake and Body Ma
aly, C.C. (1997) Energy Metabolism in Post Mortem Muscle, Proc. 43rd ICOMST - Vitality of Meat, Auckl
202
awson, E.H., Brogdon, J.L. and McManus, S. (1963a) Sensory Testing of Differences in Taste -
Dawson, E.H., Brogdon, J.L. and McManus, S. (1963b) Sensory Testing of Differences in Taste -
d Congress of Genetics and Livestock Production,
ikeman, M.E., Tuma, H.J., Glimp, H.A., Gregory, K.E. and Allen, D.M. (1972) Evaluation of
nce, 47 : 397 - 401
Dransfield, E. (1997) When the Glue Comes Unstuck, Proc. 43rd ICOMST - Vitality of Meat,
ransfield, E., Etherington, D.J. and Taylor, M.A.J. (1992b) Modelling Post-mortem
, D.E., Joseph, R.L., Schon, I., Paardekooper, E.J.C. (1984) Beef Quality Assessed at European Research Centres, Meat Science, 10: 1 - 20
Dransfield, E., Wakefield, D.K. and Parkman, I.D. (1992a) Modelling Post-mortem Tenderisation - I: Texture of Electrically Stimulated and Non-Stimulated Beef, Meat Science,
Drewnowski, A. and Rock, C.L. (1995) The Influence of Genetic Taste Markers on Food
Acceptance, The American Journal of Clinical Nutrition, 62 : 506 - 511
D omposition upon the Palatability of Three Bovine Muscles, Journal of Animal Science, 31 (1) : 36 - 41
D(1) Methods, Food Technology, 17 : 45 - 51
(2) Selection of Panel Members, Food Technology, 17 : 39 - 44
Deland, M.P.B., Malau-Aduli, A.E.O., Siebert, B.D., Bottema, C.D.K. and Pitchford, W.S. (1998) Sex and Breed Differences in the Fatty Acid Composition of Muscle Phospholipids in Crossbred Cattle, Worl
de Roos (1997) How Lipids Influence Food Flavour, In : Overview – Outstanding symposia in
Food Science and Technology, Food Technology, 51 (1) : 60 - 62
Dthe Tenderometer for Predicting Bovine Muscle Tenderness, Journal of Animal Science, 34 (6) : 960 - 962
Dolezal, H.G., Smith, G.C., Savell, J.W. and Carpenter, Z.L. (1982) Comparison of
Subcutaneous Fat Thickness, Marbling and Quality Grade for Predicting Palatability of Beef, Journal of Food Scie
Dransfield, E. (1980) Eating Quality of DFD Beef, In : The Problem of Dark Cutting Beef, Hood, D.E. and Tarrant, P.V. (Eds), Martinuus Nijhoff, The Hague, 344 - 361
Dransfield, E. (1993) Modelling Post-mortem Tenderisation - IV: Role of Calpains and
Calpastatin in Conditioning, Meat Science, 34 : 217 - 234
Auckland, New Zealand, 52 - 63
DTenderisation - II: Enzyme Changes During Storage of Electrically Stimulated and Non-Stimulated Beef, Meat Science, 31 : 75 - 84
Dransfield, E., Nute, G.R., Roberts, T.A., Boccard, R., Touraille, C., Butcher, L., Casteels, M.,
Cosentino, E., Hood
31 : 57 - 73
ryden, F.D. and Marchello, J.A. (1970) Influence of Total Lipid and Fatty Acid C
203
9 - 33
utson, T.R., Savell, J.W. and Smith, G.C. (1980) Electrical Stimulation of Ante Mortem Stressed Beef, In : The Problem of Dark Cutting Beef, Hood, D.E. and Tarrant, P.V. (Eds),
E
eate Sunflower Seeds, Journal of Animal Science, 70 : 2574 - 22580
Elias Calles, J.A., Gaskins, C.T., Busboom, J.R., Duckett, S.K., Conrath, J.D. and Reeves, J.J. (2000) Sire Variation in Fatty Acid Composition of Crossbred Wagyu Steers and Heifers, Meat Science, 56 : 23 -29
pp 169 - 182
ed Beef, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 386 - 387
Farouk, M.M. and Swan, J.E. (1997b) Factors Affecting Protein Functionality in Frozen Beef, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 388 - 389
Fe Frozen and Thawed States, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 560 - 561
Fi likes, Nature, 191 : 1328
Fo
in and/or Sarcoplasmic Protein, Journal of Food Science, 43 (11) : 815 - 818
Time Analysis, Physiology and Behaviour, 35 : 239 - 242
aw Beef with Cooked Tenderness, Journal of Agricultural Food Chemistry, 21 (5) : 880 - 883
Duncan, W.R.H. and Garton, G.A. (1978) Differences in the Proportions of Branched-Chain Fatty Acids in Subcutaneous Triacylglycerols of Barley-fed Ruminants, British Journal of Nutrition, 40 : 2
D
Martinuus Nijhoff, The Hague, 253 - 268
ckeren, P.A., Smith, D.R., Lunt, D.K. and Smith, S.B. (1992) Ruminal Biohydrogenation of Fatty Acids from High-Ol
Farmer, L.J. (1992) Meat Flavour, In : The Chemistry of Muscle Based Foods, Ledward, D.A.,
Johnston, D.E. and Knight, M.K. (Eds), Royal Society of Chemistry, Cambridge, Farouk, M.M. and Swan, J.E. (1997a) Rigor Temperature and Functional Properties of
Electrically Stimulat
rrier, G.R. and Hopkins, D.L. (1997) Tenderness of Meat Cooked From Fresh,
scher, R., Griffin, F., England, S. And Garn, S.M. (1961) Taste Thresholds and Food Dis
gerty, A.C. (1989) Dietary Fatty Acids in Blood Lipids, Food Research Quarterly, 49 (3&4) : 36 - 45
Ford, A.L., Jones, P.N., MacFarlane, J.J. Schmidt, G.R. and Turner, R.N. (1978) Binding of Meat
Pieces: Objective and Subjective Assessment of Restructured Steakettes Containing Added Myos
Frank, R.A. and Korchmar, D.L. (1985) Gustatory Processing Diferences in PTC Tasters and
Non-Tasters: A Reaction Galloway, D.G., Tuomy, J.M. and Hinnergardt, L.C. (1973) Relationship of Penetrometer
Readings on R
204
erides, with Special Reference to the Effect of Propionate on the Presence of Branched-Chain Components, British Journal of
ault, N.F.S. (1992) Structural Aspects of Raw Meat, In : The Chemistry of Muscle Based Foods,
494
lanville, E.V. and Kaplan, A.R. (1965) Food Preference and Sensitivity of Taste for Bitter
Goll, D.E., Carlin, A.F., Anderson, L.P., Kline, E.A. and Walter, M.J. (1965) Effect of Marbling
and Maturity on Beef Muscle Characteristics, Food Technology, 19 (5) :163 - 167
G ing Quality and Sensory Acceptability of a New Lean Beef Product, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 202 - 203
Gry of Muscle Based Foods, Ledward, D.A., Johnston, D.E. and
Knight, M.K. (Eds), Royal Society of Chemistry, Cambridge, pp 145 - 168
G Off-flavours in Meat - Formation and Inhibition, In : Flavor of Meat and Meat Products, Shahidi, F (Ed), Blackie, London, pp 116 -
rinker, J. (1978) Obesity and Sweet Taste, The American Journal of Clinical Nutrition, 31 :
urr, M.I. and Harwood, J.L. (1996) Lipid Biochemistry – Roles of Dietary Lipids, Chapman and
all, M.J., Bartoshuk, L.M., Cain, W.S. and Stevens, J.C. (1975) PTC Taste Blindness and The
H
f Texture Studies, 3 : 146 -164
e Studies, 7 : 49 - 63
Garton, G.A., Hovell, F.D.DeB., Duncan, W.R.H. (1972) Influence of Dietary Volatile Fatty Acids on the Fatty-Acid Composition of Lamb Triglyc
Nutrition, 28 : 409 - 416
GLedward, D.A., Johnston, D.E. and Knight, M.K. (Eds), Royal Society of Chemistry, Cambridge, pp 79 –105
Gasser, U. and Grosch, W. (1988) Identification of Volatile Flavour Compounds with High
Aroma Values from Cooked Beef, Z. Lebensm. Unters. Forsch, 186 : 489 - Gasser, U. and Grosch, W. (1990) Primary Odorants of Chicken Broth, Z. Lebensm. Unters.
Forsch, 190 : 3 – 8
GCompounds, Nature, 205 : 851
raafhuis, A.E. and Scott, S.M. (1997) Eat
ray, J.I. and Crackel, R.L (1992) Oxidative Flavour Changes in Meats : Their Origin and Prevention, In : The Chemist
ray, J.I. and Pearson, A.M. (1994) Lipid Derived
143
G1078 - 1087
GHall, pp. 188
HTaste of Caffeine, Nature, 253 : 442 - 443
ansen, L.J. (1972) Development of the Armour Tenderometer for Tenderness Evaluation of Beef Carcasses, Journal o
Harris, P.V. (1976) Structural and Other Aspects of Meat Tenderness, Journal of Textur
205
9 - 172
ce, 47 (2) : 383 - 388
olstetler, R.L., Dutson, T.R. and Smith, G.C. (1982) Effect of Electrical Stimulation and Steak Temperature at the Beginning of Cooking on Meat Tenderness and Cooking Loss, Journal of Food Science, 47 : 687 - 688
Hopkins, J.W. (1953) Laboratory Flavour Scoring : Two Experiments in Incomplete Blocks,
organ, D.J., Kurth, L.B. and Kuypers, R. (1991) pH Effect on Thermal Transition Temperature
ornstein, I. and Crowe, P.F. (1960) Meat Flavor Chemistry, Flavor Studies on Beef and Pork,
Huff-Lonergan, E., Mitsuhashi, T., Beekman, D.D., Parrish Jnr, F.C., Olsen, D.G. and Robson,
R.M. (1996) Proteolysis of Specific Muscle Structural Proteins by µ-Calpain at Low pH and
Hunter and Harold (1987) The Measurement of Appearance, 2 Ed, New York, Wiley
Izutsu, T. and Wani, K. (1985) Food Texture and Taste : A Review, Journal of Texture Studies, 16 (1) : 1 - 28
Jakobsson, B. and Bengtsson, N. (1973) Freezing of B : Influence of Aging, Freezing Rate and Cooking Method on Quality and Yield, Journal of Food Science, 38 : 560 - 565
Harris, P.V. (1988) Meat Tenderness, Proc. 34th International Congress of Meat Science and Technology, 16
Harris, P.V. and Shorthose, W.R. (1988) Meat Texture - Chapter 5, In : Developments in Meat
Science Vol. 4, Lawrie, R.A ed, Elsevier Science Publishers, pp 245 - 296 Harrison, A.R., Smith, M.E., Allen, D.M., Hunt, M.C., Kastner, C.L. and Kropf, D.H. (1978)
Nutritional Regime Effects on Quality and Yield Characteristics of Beef. Journal of Animal Scien
Hedrick, H.B., Stringer, W.C., Epley, R.J., Alexander, M.A. and Krause, G.F. (1968)
Comparison of Factors Affecting Warner-Bratzler Shear Values of Beef Steaks, Journal of Animal Science, 27 (3) : 628 - 631
Hinnergardt, L.C. and Tuomy, J.M. (1970) A Penetrometer Test to Measure Meat Tenderness,
Journal of Food Science, 35 : 312 - 315
H
Hood, D.E. (1980) Summary, In : The Problem of Dark Cutting Beef, Hood, D.E. and Tarrant,
P.V. (Eds), Martinuus Nijhoff, The Hague, 498 - 500
Biometrics, 9 : 2 - 21
Hof Collagen, Journal of Food Science, 56 (5) : 1203 - 1204
HJournal of Agricultural and Food Chemistry, 8 (6) : 494 - 498
Temperature is Similar to Degradation in Postmortem Bovine Muscle, Journal of Animal Science, 74 : 993 - 1008
nd
206
Jeremiah, L.E. (1980) Effect of Frozen Storage and Protective Wrap Upon the Cooking Losses, Palatability, and Rancidity of Fresh and Cured Pork Cuts, Journal of Food Science, 45 : 187 - 196
Je ght, R.W. (1993) Effects of Freezing and Thawing on Moisture Losses and Sensory Traits of Shabu-
ang, J.O., Oh, S.J. and Kim, C.J. (1997) The Identification of Korean Cattle Beef, Domestic
309
emp, J.D., Montgomery, R.E. and Fox, J.D. (1976) Chemical, Palatability and Cooking
im, C. (1997) Beef in the Healthy Korean Diet, AMLC Trade Seminar - Beef and Human
oohmaraie, M. (1992) Effect of pH, Temperature, and Inhibitors on Autolysis and Catalytic
regel, K.K., Kenneth, J.P. and Hughes, K.V. (1986) Cholesterol Content and Sensory Analysis
ruk, Z.A., Malau-Aduli, A.E.O., Thompson, A.M., Siebert, B.D., Pitchford, W.S. and Bottema,
of Muscle Phospholipid in Beef? Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand,
uentzel, H. and Bahri, D. (1991) Synthetic Ingredients of Food Flavourings, Food Flavourings,
arick, D.K. and Turner, B.E. (1989) Influence of Finishing Diet on the Phospholipid
f Animal Science, 67 : 2282 - 2293
Oxford, New York, pp 43 - 73.
remiah, L.E., Busboom, J.R., Gibson, L.L., Johnson, K.A., Reeves, J.J. and Wri
Shabu Beef, Journal of Muscle Foods, 4 : 161 - 170
KHolstein Beef and Imported Beef by Random Amplified Polymorphic DNA’s (RAPD) Method, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 308 -
Kapsalis, J.G. and Szczesniak, A.S. (1976) Instrumental Testing of Meat Texture - Comments on
the Past, Present and Future, Journal of Texture Studies, 7 : 109 - 113
KCharacteristics of Normal and Low Quality Pork Loins as Affected by Freezer Storage, Journal of Food Science, 41 : 1 - 3
KNutrition, pp 25
KActivity of Bovine Skeletal Muscle µ-Calpain, Journal of Animal Science, 70 : 3071 - 3080
Kof Ground Beef as Influenced by Fat Level, Heating and Storage, Journal of Food Science, 51 (5) : 1162 – 1165,1190
KC.D.K. (1997) Do Breed and Season Affect β-carotene and the Fatty Acid Composition
pp. 314 - 315
KAshurst, P.R (ed), Blackie, Glasgow and London, pp. 115 - 157
LComposition and Fatty Acid Profile of Individual Phospholipids in Lean Muscle of Beef Cattle, Journal o
Larmond, E. (1977) Laboratory methods for sensory evaluation of food, Pub 1637, Canadian
Govt. Publishing centre, Ottawa, Canada. Lawrie, R.A. (1985a) Chemical and Biochemical Constitution of Muscle - Chapter 4, In : Meat
Science - 4th Ed, Lawrie, R.A (ed), Pergamon Press,
207
Lawrie, R.A. (1985b) The Conversion of Muscle to Meat - Chapter 5, In : Meat Science - 4th Ed, Lawrie, R.A (ed), Pergamon Press, Oxford, New York, pp 74 - 91.
.A., Johnston, D.E. and Knight, M.K. (Eds), Royal Society of Chemistry, Cambridge, pp 43 - 61
Lsimus and Semitendinosus Muscles During Cooking. Journal of
Food Science, 45 : 1 - 6
L osition of Adipose Tissue of Jersey Cattle During Growth and Development, Journal of Agricultural Science, 85 : 551 - 558
LGroup, pp 41 - 45
L
ic Properties of Meat, Outlook on Agriculture, 25 (4) 263 - 266
Lugsdin, A. (2000) Market Briefs – Korea, Meat and Livestock Australia Market Information
Services
M
aland, 128 - 129
acleod, G. (1986) The Scientific and Technological Basis of Meat Flavours, In : Developments
Lawrie, R.A. (1985c) The Eating Quality of Meat - Chapter 10, In : Meat Science - 4th Ed,
Lawrie, R.A (ed), Pergamon Press, Oxford, New York, pp 169 - 207. Lawrie, R.A. (1992) Conversion of Muscle into Meat : Biochemistry, In : The Chemistry of
Muscle Based Foods, Ledward, D
eander, R.C., Hedrick, H.B., Brown, M.F. and White, J.A. (1980) Comparison of Structural Changes in Bovine Longis
eat, W.M.F. (1975) Fatty Acid Comp
eeds, W. and Lugsdin, A. (1997) Korea – 1997 Market Trends, In : Korean Beef Market Extracts, Meat and Livestock Review 1996, AMLC Beef Marketing
Ledward, D.A. (1992) Colour of Raw and Cooked Meat, In : The Chemistry of Muscle Based
Foods, Ledward, D.A., Johnston, D.E. and Knight, M.K. (Eds), Royal Society of Chemistry, Cambridge, pp 128 - 143
Lepetit, J. and Culioli, J. (1993) Mechanical Properties of Meat, Meat Science, 36 : 203 - 237
ister, D. (1996) The Meat We Eat : Notions of Quality for Today and Tomorrow. 2. The Appeal and Organolept
Lowry, S.R. (1979) Statistical Planning and Designing of Experiments to Detect Differences in
Sensory Evaluation of Beef Loin Steaks, Journal of Food Science, 44 (2) : 488 - 491
acFarlane, J.J. McKenzie, I.J. and Turner, R.N. (1984) Binding of Comminuted Meat: Effect of High Pressure, Meat Science, 10 : 307 - 320
MacFie, H. (1997) Developing Issues in Consumer and Sensory Science, Proc. 43rd ICOMST -
Vitality of Meat, Auckland, New Ze Macleod, A.J. (1994) Instrumental Methods of Meat Flavour Analysis, In : Flavor of Meat and
Meat Products, Shahidi, F (Ed), Blackie, London, pp 231 - 246
Min Food Flavours, Birch, G. G. and Lindley, M.G. (Eds), Elsevier, London, pp 191 - 223
208
acleod, G. (1994) The Flavour of Beef, In : Flavor of Meat and Meat Products, Shahidi, F
adruga, M.S. and Mottram, D.S. (1995) The Effect of pH on the Formation of Maillard-derived
aga, J.A. (1994) Umami flavour of meat, In : Flavor of Meat and Meat Products, Shahidi, F
M Pitchford, W.S. (1998a) A Comparison
of the Fatty Acid Composition of Triacylglycerols in Adipose Tissue from Limousin and Jersey Cattle, Australian Journal of Agricultural Research, 48 : 715 - 722,
Malau-Aduli, A.E.O., Siebert, B.D., Bottema, C.D.K. and Pitchford, W.S. (1998b) Breed
Comparison of the Fatty Acid Composition of Muscle Phospholipids in Jersey and Limousin
M
pose Tissue of Beef Cattle at Weaning and Slaughter, Proc. 6 World Congress of Genetics applied to Livestock Production,
105 Mandell, I.B., Buchanan-Smith, J.G., Holub, B.J. and Campbell, C.P. (1997) Effects of Fish Meal
in Beef Cattle Diets on Growth Performance, Carcass Characteristics, and Fatty Acid
Mason, D.D. and Koch, E.J. (1953) Some Problems in the Design and Statistical Analysis of
Taste Tests, Biometrics, 9 : 39 - 46
ay, S.G., Sturdivant, C.A., Lunt, D.K., Miller, R.K. and Smith, B. (1993) Comparison of
be
M(Ed), Blackie, London, pp 4 -37
MAroma Volatiles using a Cooked Meat System, Journal of the Science of Food and Agriculture, 68 : 305 - 310
M(Ed), Blackie, London, pp 98 - 115
alau-Aduli, A.E.O., Siebert, B.D., Bottema, C.D.K. and
Cattle, Journal of Animal Science, 76 : 766 - 733
alau-Aduli, A.E.O., Siebert, B.D., Bottema, C.D.K., Deland, M.P.B. and Pitchford, W.S. (1998c) Heritabilities of Triacylglycerol Fatty Acids form the Adi
th
25 : 181 - 184 Malau-Aduli, A.E.O., Edriss, M.A., Siebert, B.D., Bottema, C.D.K. and Pitchford, W.S. (2000)
Breed Differences and Genetic Parameters for Melting Point, Marbling Score and Fatty Acid Composition of Lot-fed Cattle, Journal of Animal Physiology and Animal Nutrition, 83 : 95 -
Composition of Longissimus Muscle, Journal of Animal Science, 79 : 910 - 919
MSensory Characteristics and Fatty Acid Composition Between Wagyu Crossbred and Angus Steers, Meat Science, 35 : 289 - 298
Meat Research Corporation (1993) Final Report - Defining the strategic options for Australian ef in Korea, Booz Allen and Hamilton, Australia
Meilgaard, M., Civille, V. and Carr, B.T. (1991) Sensory Evaluation Techniques, 2nd Ed, CRC
Press Inc., USA.
209
rain-finished Steers, Journal of Animal Science, 55 (1) : 77 - 87
Melton, S.L., Black, J.M., Davis, G.W. and Backus, W.R. (1982b) Flavour and Selected
iller, R.K., Rockwell, L.C., Lunt, D.K. and Carstens, G.E. (1996) Determination of the Flavor
of Crossbred Beef Cattle with and without Wagyu Genetics, Canadian Journal of Animal Science, 77 (4) : 655 - 662
M
orton, J.D., Bickerstaffe, R., Le Couteur, C.E. and Keeley, G.M. (1997) Comparison of
Vitality of Meat, Auckland, New Zealand, 672 - 673
ottram, D.S. (1991) “Meat” In : Hank Maarse (Ed). Volatile Compounds in Foods and
M 94) Some Aspects of the Chemistry of Meat Flavour, In : Flavor of Meat and
Meat Products, Shahidi, F (Ed), Blackie, London, pp 210 - 230
Ma of Cooked Beef, Journal of Food Science and Agriculture, 34 : 517 - 522
ilsen, B.N., Aaby, K., Kvaal, K., Arnkvaern, E., Blumlein, L., Risvik, E. and Ellekjaer, M.R.
’Mahony, B., McKenna, B.M. and Joseph, R.L.J. (1997) Very Fast Chilling of Beef, Proc. 43rd
O Statistical Methods and Procedures, Marcel
Deckker Inc, New York, pp. 1 - 474
Melton, S.L., Amiri, M., Davis, G.W. and Backus, W.R. (1982a) Flavour and Chemical Characteristics of Ground Beef from Grass, Forage-grain, and G
Chemical Components of Ground Beef from Steers Backgrounded on Pasture and Fed Corn up to 140 Days, Journal of Food Science, 47 : 699 - 704
MAttributes of Cooked Beef from Cross-bred Angus Steers Fed Corn or Barley-Based Diets, Meat Science, 44 (4) : 235 - 243
Mir, P.S., Bailey, D.R.C., Mir, Z., Jones, S.D.M., Entz, T., Husar, S.D., Shannon, N.H. and
Robertson, W.M. (1997) Effect of Feeding Barley Based Diets on Animal Performance, Carcass Characteristics and Meat Quality
oody, W.G. (1983) Beef Flavour – a review, Food Technology, 37 (5) : 227 - 232
MDifferent Types of Low Voltage Electrical Stimulation on Tenderisation : Interaction with Stress and Calpains, Proc. 43rd ICOMST -
Mottram, D. (1994) Meat Flavour,Understanding Natural Food Flavours, Piggott and Paterson
(Eds), 140 - 163
MBeverages, Marcel Dekker Inc., New York. pp 107 - 177
ottram, D.S. (19
ottram, D.S., and Edwards, R.A. (1983) The Role of Triglycerides and Phospholipids in the Arom
Naes, T., Baardseth, P., Helgesen, H. and Isaksson, T. (1996) Multivariate Techniques in the
Analysis of Meat Quality, Meat Science, 43 : S135 - S149
N(1997) Electronic Nose as a Potential Method for Measuring Flavour Characteristics of Pork, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 584 - 585
OICOMST - Vitality of Meat, Auckland, New Zealand, 560 - 561
’Mahony, M. (1986), Sensory Evaluation of Food :
210
age, G. (1997) The Role of Science in the Meat Industry of the Future, Proc. 43rd ICOMST -
Cutting Beef, Hood, D.E. and Tarrant, P.V. (Eds), Martinuus Nijhoff, The Hague, 454 - 460
Pend Fatty Acid Composition of Subcutaneous Fat in Steers, Journal of Animal Science,
76 : 87 - 95
Pr ST - Vitality f Meat, Auckland, New Zealand, 134 - 137
Posearch
Corporation
Ruality of Hamburger Patties, Journal of Food Science, 42
(3) : 728 - 730
R Differentiation, Meat Science, 35 : 171 -
182
Rpid in Beef and Pork, Journal of Animal Science,
61 (4) : 876 - 881
Roked Beef Longissimus Muscles with Different Degrees of Marbling, Journal
of Food Science, 47 : 716 - 719
Rng on Postmortem Changes in Glycolysis and Calpains Activities of Koran Native
Cattle (Hanwoo), Meat Science, 58(3) : 231 - 237
R 7) Effect of Temperature Conditioning and Electrical Stimulation on Postmortem Metabolism and Tenderness of
Offer, G. and Knight, P. (1988) The Structural Basis of Water-holding in Meat, Part 1 and Part 2, In : Developments in Meat Science Vol. 4, Lawrie, R.A ed, Elsevier Science Publishers, pp 152 - 235
PVitality of Meat, Auckland, New Zealand, 130 - 133
Patterson, J.T. and Bolton, G. (1980) Some Odours Produced by Bacteria on High pH and
Normal Beef, In : The Problem of Dark
rry. D., Nicholls, P.J. and Thompson, J.M. (1998) The Effect of Sire Breed on the Melting Point a
escott, J. (1997) Consumer Evaluations : Cross Cultural issues, Proc. 43rd ICOM
o
well, V. (1993) The Optimisation of the Thawing and Processing of Australian Frozen Quarter Beef in Korea, Final Report for Research Work on Project CS193C, Meat Re
andall, C.J. and Larmond, E. (1977) Effect of Method of Comminution (Flake-Cutting and Grinding) on the Acceptability and Q
eid, D.H., Young, O.A. and Braggins, T.J. (1993) The Effects of Antioxidative Treatments on Mutton Flavour/Odour Intensity and Species Flavour
enk, B.Z., Kauffman, R.G. and Schaefer, D.M. (1985) Effect of Temperature and Method of Cookery on the Retention of Intramuscular Li
hee, K.S., Dutson, T.R., Smith, G.C., Hostetler, R.L. and Reiser, R. (1982) Cholesterol Content of Raw and Co
hee, M.S. and Kim, B.C. (2001) Effect of Low Voltage Electrical Stimulation and Temperature Conditioni
hee, M.S., Ryu, Y.C., Joo, S.T., Koh, K.C., and Kim, B.C. (199
Hanwoo (Korean Native Cattle), Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 600 - 601
211
Rodin, J. (1976) Relationship Between Obesity, Weight Loss, and Taste Responsiveness,
Physiology and Behaviour, 17 : 591 - 597
Rmistry, Cambridge, pp 193
- 202 Rutley, D.L., Deland, M.P.B. and Pitchford, W.S. (1995) Cross Breeding Beef Cattle in Southern
Australia, Proc. Australian Assoc. Animal Breeding and Genetics, 11 : 151 - 154
Rs to Two Degrees
of Doneness, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand, 212 - 213
Rity Grading System and Imported Beef, Korean Journal
of Animal Science, 36 (3) : 340 - 346
Saty During Ageing, Proc. 43rd ICOMST -
Vitality of Meat, Auckland, New Zealand, 214 - 215
Sat-frozen or Frozen Tempered Beef Subprimals,
Journal of Food Science, 45 : 107 - 110, 133
Scerences Between Forage and Grain-finished Beef, Journal of Animal Science, 50
(5) : 852 - 859
Sc Young, O.A. (1997) Consumer Perceptions of the Doneness of Cooked Steak, Proc. 43rd ICOMST - Vitality of Meat, Auckland, New Zealand,
Seirer, R.C., Gaunt, G.M. and Thatcher, L.P. (1992) Fat Colour and Odour from Carcasses of
Pasture Fed Cattle Supplemented on Grain, MRC Report DAV 078
Sh vor of Meat and Meat Products, Shahidi, F (Ed), Blackie, London, pp 1 -3
Shcts, In : Flavor of Meat and Meat Products, Shahidi, F (Ed), Blackie, London, pp
247 - 266
Risvik, E. (1992) Sensory Properties and Preferences, Proc. 38th ICoMPST, 43 - 48
ossell, J.B. (1992) Chemistry of Lipids, In : The Chemistry of Muscle Based Foods, Ledward, D.A., Johnston, D.E. and Knight, M.K. (Eds), Royal Society of Che
ymill, S.R., Thompson, J.M. and Ferguson, D.M. (1997) The Effect of Intramuscular Fat Percentage on the Sensory Evaluation of Beef Cooked by Different Method
yu, Y.S., Lee, M. and Ko, K.C. (1994) A Study on the Quality Comparison of Korean Native Cattle Beef in Relation to Korean Qual
nudo, C., Nute, G.R., Campo, M.M., Alberti, P., Baker, A. and Olleta, J.L. (1997) The Infuluence of Cattle Breed on the Sensory Meat Quali
vell, J.W., Smith, G.C. and Huffman, D.L. (1980) Cutting Yields and Palatability Traits of Hand-cut or Pressed-cleaved, and Fresh, Crus
hroeder, J.W., Cramer, D.A., Bowling, R.A. and Cook, C.W. (1980) Palatability, Shelflife and Chemical Diff
ott, S.M., Devine, C.E., Cummings, T.L. and
218 - 219
ahidi, F. (1994a) Flavor of Meat and Meat Products - An Overview, In : Fla
ahidi, F. (1994b) Assessment of Lipid Oxidation and Off Flavour Development in Meat and Meat Produ
212
Shrness of Cooked Meat from Measurements Made on the Raw Meat, Proc. 34th
International Congress of Meat Science and Technology, 169 - 172
Siebert, B.D (2000) Unpublished data, Adelaide University, pers comm. [email protected]
orthose, W.R., Harris, P.V., Eadie, L.M. and Dickinson, R.F. (1988) Prediction of the Tende
ebert, B.D., Deland, M.P.B. and Pitchford, W.S. (1996) Breed Differences in the Fatty Acid Composition
Siof Subcutaneous and Intramuscular Lipid of Early and Late Maturing, Grain-
Finished Cattle. Australian Journal of Agricultural Research, 47 : 943 - 952
Si.A., Johnston, D.E. and Knight, M.K. (Eds), Royal Society of
Chemistry, Cambridge, pp 106 - 127
Sming, Frozen Storage Conditions and Degree of Doneness on Lamb Palatability
Characteristics. Journal of Food Science, 33 : 19 - 24
So ed Meat Products, In : Flavours and Off-Flavours, Proc. 6th International Flavour Conference, Charalambous (ed), Elsevier
St.Angelo, A.J., Vinyard, B.T. and Bett, K.L. ( 994) Sensory and Statistical Analyses in Meat
tone, H. (1963) Determination of Odor Difference Limens for three Compounds, Journal of
St ith, S.B. (1992) Fatty Acid Composition of
Subcutaneous and Intramuscular Adipose Tissues and M. longissimus dorsi of Wagyu Cattle,
zczesniak, A.S. (1987) Correlating Sensory with Instrumental Texture Measurements - an
akahashi, K. (1996) Structural Weakening of Skeletal Muscle Tissue During Post-mortem
y Acids of the Bovine Longissimus Dorsi, Journal of Food Science, 33 : 562 - 565
ms, T.J. and Bailey, A.J. (1992) Structural Aspects of Cooked Meat, In : The Chemistry of Muscle Based Foods, Ledward, D
ith, G.C., Spaeth, C.W., Carpenter, Z.L., King, G.T. and Hoke, K.E. (1968) The Effects of
Freez
fos, J.N. and Raharjo, S. (1989) Lipid Oxidation in Restructur
Science Publishers, Amsterdam, pp. 907 - 924
1Flavour Research, In : Flavor of Meat and Meat Products, Shahidi, F (Ed), Blackie, London, pp 267 - 290
SExperimental Psychology, 66 (5) : 466 -473
urdivant, C.A., Lunt, D.K., Smith, G.C. and Sm
Meat Science, 32 : 449 - 458
SOverview of Recent Developments, Journal of Texture Studies, 18 (1) : 1 - 15
TAgeing of Meat: the Non-Enzymatic Mechanism of Meat Tenderization, Meat Science, 43 (S) : S67 - S80
Terrel, R.N., Suess, G.C., Cassens,R.G. and Bray, R.W. (1968) Broiling, Sex and
Interrelationships with Carcass and Growth Characteristics and their Effect on the Neutral and Phospholipid Fatt
213
ressler, D.K. and Evers, C.F. (1957) Preparation and Freezing of Meat, In : The Freezing
T R. (1995) Fat Quality. Proceedings of a Seminar : Australian Export Beef – Better
arnam, A.H. and Sutherland, J.P. (1995) Frozen Meat and Meat Products, In : Meat and Meat
48
and, New Zealand, 350 - 351
W ition of Bovine Lipids as Influenced by Diet, Sex and Anatomical Location and Relationship to Sensory Characteristics,
Wheeler, T.L., Cundiff, L.V. and Koch, R.M. (1994) Effect of Marbling Degree on Beef
Palatability in Bos taurus and Bos indicus Cattle, Journal of Animal Science, 72 : 3145 - 3151 Wheeler, T.L., Miller, R.K., Savell, J.W. and Cross, H.R. (1990) Palatability of Chilled and
Frozen Beef Steaks. Journal of Food Science, 55 (2) : 301 - 304 Whipple, G. and Koohmaraie, M. (1992) Freezing and CaCl Marination Effects on Beef
Tenderness and Calpastatin Activity, Journal of Animal Science, 70 : 3081 - 3085 Winger, R.J. and Fennema, O. (1976) Tenderness and Water Holding Properties of Beef Muscle
as Influenced by Freezing and Subsequent Storage at -3°C or 15 °C, Journal of Food Science, 41 : 1433 - 1438
Wulf, D.M., Tatum, J.D., Green, R.D., Morgan, J.B., Golden, B.L. and Smith, G.C. (1996)
Genetic Influences on Beef Longissimus Palatability in Charolais- and Limousin- Sired Steers and Heifers, Journal of Animal Science, 74 : 2394 - 2405
Tilley, L. (2001) Market Briefs : Lotfeeding, Market Information Services, Meat and Livestock Australia
TPreservation of Foods, vol 1 - Freezing of Fresh Foods, 3rd ed, The Avi Publishing Co. Inc, pp 679 - 706
ume, Performance Through Improved Technology, Beautiful Virginia Palms Motel and Convention Centre, MRC and AMT, 77 -85
VProducts : Technology, Chemistry and Microbiology, Chapman and Hall, London, pp 355 - 385
Voisey, P.W. (1976) Engineering Assessment and Critique of Instruments used for Meat
Tenderness Evaluation, Journal of Texture Studies, 7 : 11 - Wasserman, A.E. and Talley, F. (1968) Organoleptic Identification of Roasted Beef, Veal, Lamb
and Pork as Affected by Fat, Journal of Food Science, 33 : 219 – 223
West, J., Young, O.A., Agnew, M.P. and Knight, T. (1997) Levels of α-tocopherol in Beef Raised on New Zealand Pastures, Proc. 43rd ICOMST - Vitality of Meat, Auckl
esterling, D.B. and Hedrick, H.B. (1979) Fatty Acid Compos
Journal of Animal Science, 48 (6) : 1343 - 1348
214
Xie, Y.R., Busboom, J.R., Cornforth, D.P., Shenton, H.T, Gaskins, C.T., Johnson., K.A., Reeves, J.J., Wright, R.W. and Cronrath, J.D. (1996a) Effects of Time on Feed and Post-Mortem Aging on Palatability and Lipid Composition of Crossbred Wagyu Beef, Meat Science, 43 (2) : 157 - 166
Xie, Y.R., Busboom, J.R., Gaskins, C.T., Johnson., K.A., Reeves, J.J., Wright, R.W. and Cronrath, J.D. (1996b) Effects of Breed and Sire on Carcass Characteristics and Fatty Acid Profiles of Crossbred Wagyu and Angus Steers, Meat Science, 43 (2) : 167-177
Yang, A., Larsen, T.W., Powell, V.H. and Tume, R.K. (1999) A Comparison of Fat Composition
of Japanese and long-term grain-fed Australian Steers, Meat Science, 51 (1) : 1 - 9 Young, O.A. and Braggins, T.J, (1996) Variation in Sheepmeat Odour and Flavour, Proc. of the
New Zealand Society of Animal Production, 56 : 167 - 172 Young, O.A., Reid, D.H. and Scales, G.H. (1993) Effect of Breed and Ultimate pH on the Odour
and Flavour of Sheep Meat, New Zealand Journal of Agricultural Research, 36 : 363 - 370 Young, O.A., Reid, D.H., Smith, M.E. and Braggins, T.J. (1994) Sheepmeat Odour and Flavour,
In : Flavor of Meat and Meat Products, Shahidi, F (Ed), Blackie, London, pp 71 - 97 Young, O.A., Berdague, J.L., Viallon, C., Rousset-Akrim, S. and Theriez, M. (1997) Fat-borne
Volatiles and Sheepmeat Odour, Meat Science, 45 (2) : 183 - 200 Yueh, M.H. and Strong, F.M. (1960) Beef Aroma, SomeVolatile Constituents of Cooked Beef,
Journal of Agricultural and Food Chemistry, 8 (6) : 491 - 494 Zembayashi, M. and Nishimura, K. (1996) Genetic and Nutritonal Effects on the Fatty Acid
Composition of Subcutaneous and Intramuscular Lipids of Steers, Meat Science, 43 (2) : 83 - 92