The Effects of Age, Sex, and Hot Carcass Weight on Cooked Lamb Flavor and Off

Home , Lamb and mutton

Flavor in Four Muscle Cuts

Thesis

Presented in Partial Fulfillment of the Requirements for the Degree Master of Animal

Science in the Graduate School of The Ohio State University

Horacio Garza III

Graduate Program in Animal Science

The Ohio State University

2017

Thesis Committee

Lyda G. Garcia, Advisor

Francis L. Fluharty

Stephan Boyles

Steve J. Moeller

Copyrighted by

Horacio Garza III

2017

Abstract

Inconsistency of lamb quality in the United States has resulted from the wide- range of production management systems including variation of breed, diet, and animal age at time of harvest (slaughter endpoints), which in turn affects body composition. In the present study, a total of 48 lambs originating from three western U.S. regions (16 lambs; 8 ewes and 8 wethers from each location) and similar breed composition (Suffolk cross), were selected to represent different carcass weight compositions (heavy and light market weights), and age at time of harvest (5 and 12 mo.). Lambs were intended to represent different production systems and US market channels within the lamb category.

After harvest, carcass measurements were recorded, then fabricated into wholesale cuts

(inside boneless lamb leg, boneless sirloin, rack, and whole boneless shoulder), and the semimembranosus, gluteus medius, longissimus thoracis, and ground shoulder were derived from these cuts (14 d postmortem) to identify differences in meat quality and palatability characteristics.

Feeding lambs a high concentrate diet after weaning (HW12), and reaching heavier end weights clearly added excess fat (backfat and body wall), had a larger ribeye, and higher yield grades (P < 0.05), resulting in a lower lean to fat ratio compared with 5 and 12 month light weight lambs. Furthermore, intramuscular fat in the longissimus thoracis was influenced as 12 month lambs possessed higher lipid concentrations than 5

ii month lambs (P < 0.05), and a sex effect was noted as wether lambs had greater lipid concentrations than ewe lambs (P < 0.05). Additionally, shoulder patty samples possessed higher lipid concentrations in heavy weight lambs (P < 0.05). Color (L*, a*, and b*) values and pH were moderately influenced by age, sex, and body weight, but light weight 12 mo. lambs (LW12) had the lowest L* values (darkest) in the LT muscle

(P < 0.01). Lastly, a sex difference was noted as wether lambs possessed higher a* value

(redder) in the LT muscle (P < 0.05).

At an older age, lamb was considered tougher in both muscle cuts sampled (LT and SM) and possessed greater flavor and off flavor intensity in the longissimus thoracis muscle, but had similar flavor and off-flavor intensity in other sampled muscle cuts compared with meat derived from younger sheep. Of note, umami, browned, liver and metallic flavor profiles were most frequently observed by panelists across treatments.

However, livery off flavor was noted as being more consistent with light weight 12 mo. lambs in the semimembranosus and gluteus medius muscles (P < 0.05) in comparison to other treatments. Overall, lambs in the present study possessed mild flavor and low instances of off flavors in all muscle cuts sampled.

iii

Dedication

Dedicated in honor of my parents: Lacho and Nelda Garza

Acknowledgments

I would first like to send a big thank you to my advisor Dr. Lyda Garcia for the opportunity of being her first graduate student, and giving me a chance to experience a whole new state, venturing out of Texas. I really appreciated all the time you have spent with me over these last few years, getting exposure to different parts of the industry and learning all facets of meat science. I have cherished the times we’ve spent putting on various meat extension workshops and traveling all over the state to judge carcass shows.

Furthermore, I want to thank you for allowing me to coach the meats judging team these last two years, it has truly been an experience unlike any other. Your passion for teaching and mentoring students is second to none, and I hope to carry these qualities with me in the future.

Furthermore, I would like to thank my committee members Dr. Francis Fluharty,

Dr. Stephen Boyles, and Dr. Steve Moeller for sharing their knowledge and providing guidance when necessary. I would also like to thank to Jerad Jaborek for your help and support throughout this project.

I want to thank Mr. Ron Cramer for all his knowledge and background he has provided me during my time as a graduate student and meats judging coach. I appreciate all the time and patience you gave to plan for meat judging practices, and I thank you for

v your friendship. Also, I want to acknowledge the meat lab crew for their help cutting product.

I want to thank other meat science faculty that helped shape me into a more well- rounded meat scientist, Dr. Eric England, Dr. Wick, Dr. Mike Cressman, and Mr. Tom

Katen. Also, I would like to thank Dr. Luiz Moraes for his help with all my statistical questions. Moreover, I want to thank the graduate students and faculty which took part in our taste panel.

I want to thank all the friends I’ve made during my time in Ohio, especially Tori

Trbovich and Karli Feicht for all your help, support, and friendship during my graduate school career. I truly could not have done it without you all in my circle and will always cherish the time we’ve spent together. Thank you to the 2016 and 2017 meat judging members for all the hard work and dedication you all put during your time on your respective teams. I will always remember the times I spent with each of you and thank you for helping shape me into a better teacher and coach. 2017: Kristen Browne, Sierra

Jepsen, Nicole McKibben, Caroline Miller, April Rose, Ethan Scheffler, Zach Temple, and Emily Warnock. 2018: Bailie Corbin, Ethan Fink, Morgan Foster, Paige McAtee,

Randi Shaw, Hailey Shoemaker, and Kaitlyn Wiley.

Most importantly I would like to thank my parents, Lacho and Nelda Garza, for their continued love and support throughout my life. Thank you to both my sisters,

Jessica and Geovanna for their inspiration throughout this process. I would also like to thank my girlfriend, Keeley for your encouragement along the way. I know I wouldn’t be here if it wasn’t for all of you.

Vita

February 25, 1993………………………………………………...Born in Edinburg, Texas

May 2015…………………………………B.S. Animal Science, Texas Tech University

August 2015- Present………………………. Graduate Research Assistant, Department of

Animal Science, The Ohio State University

Fields of Study

Major Field: Animal Science

Discipline: Meat Science

vii

Table of Contents

Abstract ...... ii Dedication ...... iv Acknowledgments...... v Vita ...... vii List of Tables ...... ix List of Figures ...... x CHAPTER 1. Introduction...... 1 CHAPTER 2. Literature Review ...... 3 CHAPTER 3. Materials and Methods ...... 26 CHAPTER 4. Results and Discussion ...... 34

List of References……………………………………………………………………...... 54

viii

List of Tables

Table 2.1. U.S. commercial lamb, yearlings, and mature sheep slaughter by region in

2016...... 23

Table 4.1. Least-square means and standard deviations (± SD) of lamb carcass characteristics of different treatments and sex...... 43

Table 4.2. Least-square means of lamb carcass characteristics by treatment and sex...... 44

Table 4.3. Least square means of lipid concentrations (%) from the Longissimus thoracis and shoulder muscle cuts representing treatment and sex...... 45

Table 4.4. Least-square means of muscle (Longissimus thoracis) pH and instrumental color measurements by treatments and sex...... 46

Table 4.5. Least square means of slice shear force (kg) from the Longissimus thoracis and Semimembranosus muscles of lamb carcasses representing treatment and sex ...... 47

Table 4.6. Log of lamb flavor and off flavor intensity consumer panel scores in various lamb muscle cuts by treatment and sex...... 48

Table 4.7. Least square means, after data transformation, of lamb flavor and off flavor intensity consumer panel scores in various lamb muscle cuts by treatment and sex...... 49

Table 4.8. Percentage of consumer panel respondents identifying off- flavor characteristics in various muscle cuts by treatment...... 50

List of Figures

Figure 2.1. U.S. commercial sheep and lamb slaughter, average live and expected carcass weight by region in 2016 ...... 24

Figure 2.2. Illustration of degree flank fat streaking in lamb carcasses...... 25

Figure 4.1. Treatment by sex interaction for the longissimus thoracis lipid concentrations...... 51

Figure 4.2. Treatment by sex interaction for the shoulder patty flavor intensity score ... 52

Figure 4.3. Treatment by sex interaction for the shoulder patty off- flavor intensity score...... 53

CHAPTER 1

INTRODUCTION

The American lamb industry has continuously been faced with the challenging task of producing consistent lamb meat products. The inconsistency of lamb quality in the

United States has resulted from the wide-range of production management systems including variation of breed, diet, and animal age at time of harvest (slaughter endpoints), which in turn affects body composition. Lamb meat quality is a predictor of lamb meat palatability and consists of three factors: tenderness, flavor, and juiciness. These factors impact consumer eating satisfaction and influence the likelihood of purchasing lamb as consumer eating satisfaction has been reported by the 2015 National Lamb Quality Audit

(Hoffman et al., 2016) as the most important quality factor leading consumers to purchase lamb, while recognizing eating satisfaction/quality as taste or flavor. Sensory studies of lamb have also shown flavor to be very important in leading consumers to purchase lamb

(Oltra et al., 2015; Font I Furnols et al., 2009; Jamora et al., 1998).

Lamb meat possesses a distinctive, species-specific flavor profile occasionally being referred to as a “mutton-like” flavor, and classified as a strong in flavor. American consumers have commonly preferred a mild flavor when consuming lamb (Field, 1983).

Branched-chain fatty acids deposited in fat have been identified as primary contributors for the unique lamb flavor and has also been shown to increase with age (Wong et al., 1975a,

1 b). Consequently, a greater amount of fat would lead to an increase in flavor, which is disagreeable to many American consumers. However, sheep producers/feeders increase their profits by producing sheep with heavier live body weights, resulting in greater proportions of fat to lean, which can later affect consumer eating quality.

This research investigated the lamb meat quality characteristics due to age at time of harvest (within the young lamb category), body weight, and lamb sex. Therefore, we investigated the characteristics affecting eating quality of lamb meat, such as lamb flavor and tenderness. Lambs in the present study consisted of two carcass compositions, and two harvest ages simulating common production strategies to help determine a more beneficial production strategy, while investigating eating quality.

CHAPTER 2

LITERATURE REVIEW

Introduction

Consumer eating satisfaction has been reported by the 2015 National Lamb Quality

Audit (Hoffman et al., 2016) as the most important quality factor leading consumers to purchase lamb, while describing eating satisfaction/quality as taste or flavor, thus influencing the purchase decisions of consumers. The American lamb industry has continuously been faced with this issue due to the wide-range of lamb production management systems including variation in breed composition, diet, and animal age at time of harvest (slaughter endpoints); which result in differences in weight and body composition within and across age groups. Because of these factors, lamb products across the U.S. have been inconsistent regarding appearance, composition, and taste resulting in consumers being hesitant to purchase lamb meat relative to other meat protein alternatives.

The USDA lamb quality grading system currently models the quality grading system used for beef grading, and is greatly influenced by fat deposition (degrees of flank streaking) and animal maturity (break joint status) (USDA, 1992). The beef quality grading system attempts to quantify degree of marbling (intermuscular fat content) in the longissimus muscle as a predictor of the level of flavor in beef at the time of consumption.

In the lamb quality grading system, the degree of flank streaking serves as an indicator of

3 the expected level of intramuscular fat present and serves as a predictor of flavor intensity during the consumption of cooked lamb. However, unlike most beef flavor profiles, lamb meat possesses a distinctive, species-specific flavor profile occasionally being referred to as “mutton-like” or strong flavors. The presence of “mutton-like”, and/or stronger flavor intensity in lamb flavor profile can be disagreeable to many American consumers, particularly if they are not familiar with the flavor profile of lamb meat. The presence and relative abundance of specific branched-chain fatty acids within fat depots have been identified as primary contributors to the variation in lamb and mutton flavor, with intensity of mutton flavor being observed to increase with age (Wong et al., 1975a, b). Therefore, increased age and degree of fatness are likely to result in mutton and/or strong flavors; however, sheep producers and feeders are primarily paid on weight of the lamb resulting in the marketing of heavier sheep when profits margins are favorable, resulting in inconsistent lamb flavor profiles that can negatively affecting eating quality.

U.S. lamb production

According to the United States Department of Agriculture-Economic Research

Service (USDA-ERS), in 2016 an estimated total of 2.01 million head of lamb (< 14 mo. age), yearling (14 - 24 mo. age), and mutton (> 48 mo. age) sheep were slaughtered in the

U.S. Over 40% of the sheep marketed were derived from the mountain west region states of Colorado, Montana, North Dakota, South Dakota, Utah, and Wyoming. Whereas, the

West coast (California, Oregon, Washington etc.,) and Midwest (Michigan, Ohio, Illinois etc.,) accounted for 15 and 17% of the marketed supply, respectively (Table 1.). Live market weights of U.S. commercial sheep and slaughter lambs have been shown to be

4 region dependent in the United States (Figure 1.). For example, live weight data from the

Mountain West, Midwest, West Coast, and Southwest regions tend to be heavier (> 100 lbs.), while sheep marketed in the Eastern, Southern, and New England states tend to be lighter (< 100 lbs.). Feedlots predominantly located in Western states, taking lambs to heavier weights with access to readily available high concentrate feedstuffs may explain increases market weights, in comparison with Eastern and Southern states. Regional weight variation results in inconsistent carcass weights throughout the U.S., and may provide at least a partial explanation for the variation in lamb product consistency that has been reported throughout the U.S.

This production of live lamb translated to 150.2 million pounds of lamb and mutton meat in 2016 (USDA, 2016). However, this is a minimal amount when compared with other lamb producing countries on an annual basis (bone in retail cuts) such as China (4.4 billion lbs.), European Union (1.9 billion lbs.), Australia (1.4 billion lbs.), and New Zealand

(1 billion lbs.) (Brester, 2011). Over the past decade, U.S. lamb imports have increased, with 216 million pounds imported annually (predominantly from Australia), and imports account for nearly half of the total U.S. lamb at the retail level. Exports of U.S. raised lamb exports are minimal (about 3.2 % of domestic supplies) with 5 million pounds exported, comprised mostly of aged cull ewes and rams sent to Mexico as mutton (USDA, 2016;

Brester, 2011).

Consumption of lamb in the U.S.

In the United States, per capita disappearance (boneless retail weight) of lamb is very low (0.8 lb., 0.36 kg) when compared to pork (47.1 lb., 24.1 kg), beef (53.1 lb., 24.1

5 kg), and total poultry (76.8 lb., 34.9 kg) (USDA- ERS). The USDA-ERS recently published data between 2008 and 2016 describing the U.S. annual per capita consumption trends for lamb indicating lamb consumption has been steady and ranged from 0.4 to 0.5 kg/person/year, representing less than 0.5% of total U.S. meat and poultry consumption.

In 1995, Ward et al. conducted a survey in Tulsa, OK of 600 households comparing seven meats: beef, chicken, pork, lamb, fish, turkey, and veal. Of the meats surveyed, lamb was ranked last for: taste, economic value, convenience and overall preference. More than

20 years later, the average consumer has not deviated from these preferences, as evidenced by the relatively flat trend of consistently low per-capita consumption of lamb.

Nevertheless, a portion of lamb meat sold/consumed is channeled outside of the standard commercial American markets channels. Typically, these nontraditional markets are referred to as ethnic markets. These non-traditional markets stem from ethnic/religious customs and cultures whereby lamb meat is frequently consumed as a part of a regular meal or in instances revolving around the observance of an ethnic holiday (Shifilett et al., 2010).

It is estimated that for every lamb slaughtered in the commercial market, one-third is marketed separately for non-traditional (ethnic) marketing on the live basis (USDA/NASS,

2013). These non-traditional markets are generally referred to as direct marketing, as live lambs are sold directly from the producer to the consumer.

Due to cultural and religious significance, lamb is one of the most widely chosen forms of red meat consumed across the world demographic. In contrast, for example,

Islamic and Judaist religious beliefs prohibit the consumption of pork; while Hinduism beliefs prohibit consumption of beef as it is viewed as a “sacred” animal (Waibel, 2008).

With increased cultural diversity and ethnic population growth in the U.S., the opportunity to increase consumer acceptability and per capita consumption of lamb products is promising.

USDA lamb grading- quality grading

Currently, USDA utilizes quality and yield grading standards to segregate carcasses into similar market destinations and determine overall carcass value. Quality grading was developed to estimate eating experience and consumer satisfaction based on palatability traits (tenderness, flavor, and juiciness) and consumer acceptability. Lamb quality grades include: USDA Prime, Choice, Good, Utility, and Cull. Current estimates are that 98% of lamb marketed in the U.S. are graded as USDA Choice. Lamb quality grades are determined using a combination of two main factors: physiological maturity and degrees of flank streaking. Maturity categories in grading carcasses include: lamb (< 14 mo.), yearling mutton (14 - 24 mo.), and mutton (> 48 mo.) (USDA, 1992).

Physiological maturity

The presence of a break joint or spool joint is used to determine the youthfulness of a lamb carcass. According to USDA, a carcass with perfect break joints on both trotters

(metacarpals) at time of slaughter is classified as either lamb or yearling mutton, while a carcass with spool joints on both trotters is classified as yearling mutton or mutton based on secondary evidences of maturity (USDA, 1992). Secondary indicators of maturity include: width/shape of ribs, flank lean color and texture (rounder ribs with light red lean color indicate youthfulness). Carcasses grading into the mutton category are not eligible to grade USDA Prime, and can only grade as USDA Choice, Good, Utility and Cull. A carcass

7 having one trotter with a perfect break joint is classified as a lamb if secondary maturity characteristics fit the description of a young lamb (USDA, 1992). As lambs mature, the epiphyseal cartilage (growth plate) on the distal ends of the metacarpals (trotters) began to ossify, causing break joints to fuse, and forming the appearance of spool joints.

Classes of slaughter sheep include: ram (uncastrated male ovine), ewe (a female ovine), and wether (castrated prior to developing characteristics of a ram male). The age groups of live market classification include: lamb (immature ovine, < 14 mo. age, not cut its first permanent incisors), yearling (between 12-24 mo. of age, cut its first permanent incisors, but not second permanent incisors), and sheep (> 24 mo. of age, cut its second permanent incisors) (USDA, 1992). Therefore, in live markets, lamb is considered young by age and sheep are referred to as old by age.

Studies have shown sex of the sheep have been shown to influence maturity characteristics of lamb/sheep carcasses. Ho et al. (1989) studied the effect of sex (ewes, wethers, and ram lambs) and the maturity at five ages on carcass characteristics (271, 361,

459, 557, and 652 days of age (DOA)) and reported that ossification of break joint maturity occurred earliest in ewes, then rams, and followed by wethers. At 459 DOA (> 15 mo.), 40

% of ewes sampled had ossified epiphyseal plates indicative of spool joints and displayed yearling teeth, while wethers of the same age had zero instances of ossified epiphyseal plates or spool joints, but more than half exhibited yearling teeth. In contrast, 20% of rams at 459 DOA had spool joints, while 50 % showed yearling teeth. Essentially, this allows a greater proportion of wethers to be eligible to be graded in the U.S. as lamb carcasses,

8 although their chronological age is greater than 14 mo., giving wethers an advantage over ewe carcasses (Ho et al., 1989).

Degree of flank streakings and carcass conformation

Unlike beef carcasses, packing plants harvesting, and grading ovine carcasses do not expose the longissimus muscle to evaluate muscle quality characteristics (marbling, color, texture, and firmness). Rather, in ovine carcasses, quality is evaluated indirectly by giving considerations to the visual quantity of fat streaking beneath the epimysium of the flank muscle. The degree of flank streaking is an indirect estimate of expected eating experience of the middle meats (rib, rack, and loin). Degrees of flank streaking include nine scores/categories: abundant, moderately abundant, slightly abundant, moderate, modest, small, slight, traces and practically devoid (shown in Figure 2.). Despite a poor predictive relationship between flank fat streaking and intramuscular fat (marbling) within the loin, assessing flank streaking is a non-invasive means that is preferred when compared with exposing the 12th/13th rib interface. This method allows for the salvaging of the rib, rack, and loin, thus increased overall carcass value (Aberle et al., 2012). Lastly, carcass conformation score (relationship between carcass muscling, width and thickness relative to length, and ratio of muscle to bone) is also included in determining carcass quality grade.

Carcasses with a greater conformation score exhibit heavy muscled, thick, plump legs, loins and shoulders, and have a thin layer of fat, uniformly distributed over the carcasses.

Low conformation scores are indicative of thin muscled, narrow legs, loin, and shoulders, with higher proportions of fat over the carcass and lesser lean to bone ratio. While carcass

9 conformation is included in quality assessment, a greater relative importance is placed on maturity and flank streaking when establishing a quality grade.

USDA lamb grading- yield grading

The USDA lamb yield grade is an estimation of percent of the total carcass weight that is represented in the boneless trimmed, semi-boneless, and boneless leg, loin, rib, and shoulder cuts from the carcass. To determine yield grades, ovine carcasses are ribbed between the 12th/13th ribs to measure the backfat thickness over the longissimus muscle.

Five yield grades exist starting with the heaviest muscled, trimmest carcass, offering the greatest yield of sellable product (YG 1) and ending with a carcass that represents the fattest, lightest muscled carcass, offering the least amount of sellable product (YG 5). The yield grade equation includes: YG = 0.4 + (10  fat thickness (inches)). Therefore, carcasses with backfat measurements between 0.00 - 0.15 in., 0.16 - 0.25 in., 0.26 - 0.35 in., 0.36 - 0.45 in., and greater than 0.46 in. would classify as yield grades of 1, 2, 3, 4, and

5 respectively (USDA, 1992). The projected percentage of retail yield for lamb carcasses in relation to yield grade is 51.0% (YG 1), 49.7% (YG 2), 48.4% (YG 3), 47.1% (YG 4),

45.8% (YG 5) (Aberle et al., 2012). Garrett et al. (1992) compared lamb carcasses of different yield grades (YG: 2, 3, and 4) possessing the same weight and quality grade, and noted the increase in numerical yield is not only related to subcutaneous (external) fat but also results in an increased amount of intermuscular (seam) fat. Furthermore, this increase in intermuscular fat resulted in greater percentages of kidney and pelvic fat, which in turn explains the amount of profit or loss per carcass with greater amounts of fat.

Palatability and consumer concerns

In the United States, meat quality refers a combination of traits believed to be most important for consumer acceptability. Consumer acceptability includes both visual (color, texture, excess purge, etc.) and previous consumer eating experiences (Becker, 2000).

Meat palatability is often rated by how satisfied consumers are during their eating experience. Once cooked, meat palatability is broken up into three factors: tenderness, flavor, and juiciness. Historically in the beef industry, tenderness has been noted as the most important factor in determining consumer palatability (Huffman et al., 1996; Miller et al., 2001). However, O’Quinn et al., (2012) recently reported overall consumer acceptability was correlated with flavor in beef products (r = 0.88). In regard to lamb acceptability, American lamb consumers placed the greatest influence on flavor, followed by tenderness and last juiciness (Young et al., 1997; Pethick et al., 2006). Recent studies reported by the 2015 National Lamb Quality Audit acknowledge eating satisfaction as the most important factor leading consumers to purchase lamb, and described eating satisfaction/quality as taste or flavor (Hoffman et al., 2016).

Additionally, the combination of flavor and aroma have been shown to be critical components to purchasing decisions by American consumers. Flavor and aroma sensations are derived from a complex combination of compounds. Flavor molecules must interact with consumer sensory receptors primarily the sense of smell, so that perception of flavors can be detected. The five basic flavor sensations include: salty, sweet, sour, bitter, and umami (Aberle et al., 2012). Fresh, uncooked meat has very little flavor and does not develop its flavor until cooking. During cooking, lipid or water-soluble components in meat undergo two sets of chemical reactions. The first reaction is the Maillard reaction

11 occurring between meat sugars and amino acids which are the result of browning of meats when cooked, and the second reaction involves the oxidative breakdown of lipid components (Watkins et al., 2013). Flavors profiles of meat can be expressed differently based on the amount and proportion of volatile compounds found in lipids, and this variability contributes to species-specific flavors in meat. In relation to lamb flavor,

“mutton” and “pastoral” flavors have traditionally been used to describe the species- specific flavors in lamb (Young et al., 1997). Rhee and Ziprin (1996) addressed species- specific flavor profiles and conducted consumer (n = 71) and trained taste panels (n = 12) in Texas, serving lamb, beef, and pork patties containing 20% fat. The study indicated that lamb patties were assessed to have the most intense flavor, and panelists were able to correctly identify lamb patties with a highest degree of accuracy compared with all other species sampled.

Factors affecting flavor in lamb

Branched chain fatty acids (BCFA) in lamb have been attributed to the species- specific flavor and aroma in lamb. Three BCFA, 4-methyloctanoic acid (MOA), 4- methylnonanoic acid (MNA), and 4-ethyloctanoic acid (EOA) have specifically been identified as the primary contributors to the unique flavor of lamb (Young et al., 1997;

Watkins et al., 2010). Observed differences in the flavor of lamb have been attributed to the wide range of production systems present within the U.S. Production system approaches can be impacted by the age, diet, breed and body composition of sheep at time of slaughter, contributing to variation in BCFA and subsequent lamb flavor profiles. The concentration of specific branched chain fatty acids has been shown to increase as age

12 increases in sheep. Young et al. (2006) conducted fatty acid analysis looking at BCFA in lambs harvested at different ages in groups of six at 91 d intervals, ranging from 122 to 668 d of age. Young et al. (2006) determined that the concentrations of MOA and MNA in fat increased as animal age increased. Similarly, Watkins et al. (2010) studied three different age groups: lamb (less than 1 year old), hogget (1 – 2 years old, term used in Australian grading, yearlings by USDA classification standards), and mutton (more than 2 years old) and found the smallest amounts of BCFA found in lambs, followed by intermediate levels in hogget, and the greatest concentrations in mutton. Therefore, the younger lamb with less

BCFA will result in low lamb flavor intensity when compared with meat from an mutton

(old) with the greatest amount of BCFA.

Diet effects on flavor

Animal diets can vary among production systems in the lamb industry as a function of cost, availability of feedstuffs, and opportunity costs associated with a specific demographic region. Variation in nutrition can have a positive or negative effect on overall flavor of the meat causing differences in flavor and off flavor development in lamb.

Rousset-Akrim et al. (1997) identified significant flavors detected in the meat of ram lambs and mature ewes fed different diets and across differing planes of live weight growth.

Within this study, ten flavor attributes were selected for cooked meat flavor from the data collected in training and in panel discussions. Of the ten flavors (authors description of flavors) investigated, sheepmeat (cooked sheepmeat, irrespective of age), animal (flavor associated with confined livestock; “barnyard”), liver (cooked liver flavor), and poultry

(cooked poultry) were flavors panelists could significantly detect when consuming lamb

13 and mutton (Rousset-Akrim et al., 1997). Sheepmeat flavor was shown to have the most intense flavor across treatments, and was significantly greater in ram lambs slowly grown

(ADG 0.12 - 0.16 kg) on pasture. Liver was the most distinguishable flavor in meat from mature ewes grazing on pasture. Additionally, an olfactometer was used to evaluate cooked fat flavors. Flavors identified in slowly grown ram lambs at both heavy and light weights included sheepmeat, cabbage, roast, animal, and rancid odors. Slowly grown pasture raised lambs possessed the most intense sheepmeat, animal odors, and flavor in comparison to grain fed lamb. Grain fed lamb had lower intensities of sheepmeat flavor, sheepmeat odor, and animal odor, but not animal flavor.

Other studies have looked for indicators that contribute to off-odors and flavors by studying volatile compounds in subcutaneous fat of lambs (Young, 1997). Young et. al

(1997) used gas chromatography-mass spectrometry to distinguish individual volatile compounds in ram lambs finished with fast and slow growth rates from pasture and grain treatments, respectively. Long chain alkanes and alkenes were dominant in the slowly grown treatments for ram lambs and mature ewes. Branched chain fatty acids (BCFA) found in the highest concentrations of slow grown lambs on pasture were 2,3-ocanedione,

2-nitropyrazole and 3-methylindole (skatole) which, in turn, were shown to be good indicators of the pastoral diets. The presence of BCFA 4-methylphenol was reported in the light, slowly grown treatment, while BCFA 4-iso-propyphenol was reported in the fast- grown pasture treatment. Young et. al (1997) determined that medium BCFA were responsible for sheep odor, while skatole and 4-methylphenol were attributed to animal

14 odors (confined livestock odor), thus identifying volatiles associated with lamb off flavors and odors.

Impact of maturity on tenderness in lamb

Tenderness is a consumer concern with respect to eating quality, but not to the extent that flavor influences acceptability of lamb. Consumer measurements of palatability specific to tenderness, are often described by indicating if meat is tough or tender.

Perception of tenderness is defined by the softness of tongue and cheek, and resistance to tooth pressure by consumer. The amount of connective tissue, muscle fiber integrity, and adipose tissue level contributes to meat tenderness (Aberle et al., 2012). Objective measurements of cooked meat tenderness are commonly assessed using mechanical devices that simulate consumer bite, including Warner-Bratzler shear force testing

(WBSF), slice shear force assessment (SSF), and Instron Corporation (Norwood,

Massachussetts) compression approaches. Instrumental assessment of tenderness is commonly performed to reduce cost and the challenge to control variability of panelists during sensory testing.

Several studies have examined the effects of age at harvest on the tenderness of lamb muscles. Bouton et al. (1978) studied sheep of various ages ranging from 2-3 mo. to

6-8 yrs. of age, evaluating shear force values of four leg muscles (biceps femoris, semimembranosus, semitendinosus and gluteus medius). It was determined that the force

(peak force) required to cut through samples increased with animal age, resulting in the conclusion that tenderness decreases with age as the age at harvest increased. Hopkins et al. (2006) studied the relationship between animal age, intramuscular fat, cooking loss and

15 shear force of aged meat, and their effects on sensory characteristics using a consumer panel. Results indicated that as shear force and age increased, sensory characteristics

(tenderness, juiciness, flavor, and overall likability) scores decreased, indicating less acceptability by the panel. Hopkins et al. (2006) noted that, in the author’s opinion, shear force values were not an indication of juiciness and flavor, but rather attributed this result to the halo effect (Shorthose and Harris, 1991), where untrained consumer panelists tend to combine sensory traits and related them to each other. In a follow up study, Hopkins et al. (2007) evaluated lambs slaughtered at 4, 8, 14 and 22 mo. of age, representing five genotypes. They found that there was a significant interaction between animal age and tenderness, noting that meat cuts from the suckler lambs (4 mo. old) produced the lowest shear force values after five days of aging in the longissimus lumborum muscle in comparison to 14 and 22 mo. old lambs, but no significant (P > 0.05) differences were noted between 4 and 8 mo. old lambs. The age and tenderness interaction could be explained by an increased amount of collagen derived from meat in older animals, causing toughening of meat.

The impact of increased animal maturity on meat tenderness is directly linked to an increase in the amount of collagen crosslinking. Increased crosslinking causes a decrease in the solubility of collagen, and the structure of perimysial collagen remains intact after cooking (Bailey & Light, 1989), depending on cooking method (direct and indirect heat, and moist heat), time and temperature. In beef, Shorthose & Harris (1990) studied the tenderness of 12 beef muscles derived from cattle harvested at 8 different ages, ranging from 1 to 60 mo. of age. Authors found that the mean tenderness measurement significantly

16 decreased with an increase in animal age at harvest, and that the muscles found to be collagen rich, were less tender than muscles with low levels of connective tissue. Young

& Braggins (1992) studied the tenderness of the ovine semimembranosus and gluteus medius muscles at different harvest age endpoints to determine if collagen concentration, or the solubility of collagen impacted meat tenderness. Results of their study indicated that collagen concentration remained unchanged within the muscles studied across the age groups, while solubility of collagen decreased with increasing age between the range of 4 mo. and 5 yr. at harvest. Solubility in the semimembranosus (n = 36) ranged from 4.1-

16.9% solubility, while the solubility of the gluteus medius (n = 108) ranged from 13.5-

70.5%. However, sensory panel data findings showed collagen concentration was a more important factor in determining eating quality while, in contrast, shear force data indicated that solubility was a more important determinant.

Impact of diet on tenderness

Murphy et al. (2003) investigated the effects of dietary energy source on lamb tenderness by feeding either a high-concentrate, grain-based diet, a high-forage, alfalfa- based diet, or a diet that was equal parts grain and forage. Warner-Bratzler shear force values of the longissimus dorsi were lower (more tender) in the high forage and mixed diets than the high concentrate diet. The findings also concurred with sensory evaluation of the loin chop samples, whereby loin samples from lambs fed mixed forage diets were significantly more tender than samples from lambs fed the high concentrate diet (Murphy et al., 2003). In addition, Priolo et al. (2001) studied the effect of high and low growth rates on lamb carcass composition and meat quality. Lambs were divided into four feeding

17 systems, two groups were allowed to graze pasture to achieve high and low rate gains, while the other two groups were fed in stalls on different concentrate diets (high and low gain). Sensory evaluations on the longissimus were conducted with a trained panel to determine eating quality. In contrast to the report by Murphy et al. (2003), samples from stall fed animals were more tender than meat from grass fed animals, however, there was no difference in tenderness between animals fed to achieve high and low growth rates with concentrate diets (Priolo et al., 2001).

Impact of post-mortem aging on meat tenderness

Post-mortem tenderization of meat through an aging process is dependent on protein degradation (breakdown) within the structural units of muscle (sarcomere) and is defined as the breakdown of small proteins in muscle fibers (cytoskeletal proteins).

Cytoskeletal proteins are structural proteins responsible for the overall firmness of meat products. Common cytoskeletal proteins analyzed in protein degradation research include: desmin, nebulin, and titin, and the breakdown of these proteins can contribute to an increase in tenderness postmortem (Aberle et al., 2012). In the beef industry, it is a common practice to store meat in a vacuum packaged bag or dry age beef in controlled conditions over extended periods of time (14 - 30 days). Both methods result in an increased product tenderness due to the structural breakdown of muscle fiber post-mortem (Smith et al.,

2008). However, age at time of harvest can affect the degradation of post-mortem muscle.

Hopkins et al. (2007) investigated the effects of animal age, genotype, and post-mortem aging of meat on meat quality in lamb. Five genotypes and four animal ages were represented (4, 8, 14 and 22 mo. of age). Shear force testing results indicated a significant

18 interaction between animal age and post-mortem aging time, as the meat from 4 mo., 8 mo., and 14 mo. old lambs had the greatest decline (24.6, 21.2, and 24.6 N, respectfully) in shear force after ageing for five days, while meat from the 22 mo. old lambs displayed only a 11.6 N decline in shear force. However, the 8 and 14 mo. old lambs were not significantly different compared with 4 mo. old lamb, displaying decreased shear force values of 21.2 N and 24.6 N, respectfully.

In beef, Huff et al. (1994) studied the effects of postmortem aging time, animal age and sex on the degradation of the cytoskeletal proteins titin and nebulin in the longissimus muscle. They looked at three classifications of beef, steers and bulls (14 mo. old), and old

(44 - 108 mo. old), evaluating the change of protein degradation in postmortem longissimus samples. The rate of conversion of intact titin to degraded titin was slower in older animals

(cows) in comparison to the younger steers, while bull samples showed few differences when compared with samples from cows. While nebulin was mostly degraded in the loin of steers by day 3 (apart from less tender samples), all 3-d samples from bulls and cows remained intact, showing no differences (Huff et al., 1994).

Veiseth et al. (2004) studied factors affecting tenderness in the longissimus dorsi of lamb when slaughtered at 2, 4, 6, 8, and 10 mo. of age. They found that there was a notable change in the mechanism responsible for postmortem proteolysis (calpain system).

Calpastatin activity (increased activity results in tougher meat) declined from 4.18 to 1.91

µ/g in muscle when comparing the loin from lambs slaughtered a 10 and 2 mo. of age, respectfully. There was a steady decline in m-calpain activity within the loin as age at slaughter increased, with sheep at 10 months of age having reduced to 80% m-calpain

19 activity when compared with the loin from sheep slaughtered at 2 mo. In addition, μ-calpain activity within the muscle decreased as animal age at slaughter increased. Desmin degradation in the longissimus muscle was not different across sheep age classes at 2 d postmortem; however, at 10 d postmortem, desmin degradation increased in the loin when comparing slaughter from 2 to 6 mo. old sheep, indicating that as animals get older, the rate of postmortem proteolysis decreases over time.

Another postmortem factor that can affect tenderness is muscle shortening (causing toughening) in the cooler after harvest. Smith et al. (1976) showed lamb carcasses with an increased amount of subcutaneous fat and larger longissimus dorsi muscle (typically older animals) result in slower chilling times, thus having less of an effect on the shortening of sarcomeres and resulting in a more tender product.

Instrumental measurement of tenderness

Shear force analysis is designed to objectively capture the amount of force (N or kg) required to “bite” through a meat sample intended to replicate the human jaw. The two most common methods of tests used to measure tenderness of whole muscle samples are

Warner-Bratzler (WBSF) and Slice Shear Force (SSF). Both methods determine the peak force at time of shearing, which is defined as the greatest amount of force required to shear through the meat samples.

Warner-Bratzler Shear Force

In a review of the process, WBSF testing requires cooking of whole muscle samples

(2.54 cm thick) using an identified cooking method (flat top griddle; clam shell grill; impingement oven) to a desired temperature, indicative of a desired degree of doneness.

Cooked samples are held under refrigerated temperatures (2-5C) for over 24 hr. or until the internal temperature reaches 23C, to standardize internal temperatures across samples.

Whole muscle samples are then cored using a 1.22 cm diameter template with three to six cores commonly collected parallel to the muscle fiber orientation from each cut. Cores are placed on a shear device fitted with a V-type notch blade that shears perpendicular to the muscle fiber, with shear determined by recording individual peak shear values and recording the average peak force (Hopkins et al., 2007) across all cores collected from a sample. Warner-Bratzler shear force (WBSF) is the most common method of evaluating tenderness in lamb longissimus dorsi due to the relatively small anatomical muscle size in ovine.

Slice Shear Force

Slice shear force (SSF) is used for objective measurement of tenderness that also evaluates the force required to “bite” through a meat sample; however, in place of analyzing a cold sample, SSF analyzes a warm, cooked sample. This procedure calls for the use of a cut box (template used to produce uniform slices), and a double-bladed knife to obtain slices cut at 45 angles parallel to muscle fibers. A flat, blunt ended blade is then used to shear through slice samples, recording peak force. The SSF method has been evaluated in a scientific setting to compare the effectiveness of the measure for the prediction and assessment of the relationship with sensory panel tenderness scores and traditional Warner-Bratzler shear force (WBSF) in beef. Repeatability estimates for listed procedures with sensory panel tenderness were SSF = -0.82 and WBSF = -0.77

(Shackelford et al., 1999a). Results have indicated SSF can be a suitable or potentially more effective alternative to WBSF as a predictor of consumer perceptions of tenderness.

Warner-Bratzler Shear Force versus Slice Shear Force in Lamb

Since testing tenderness with WBSF was so common in lamb, there had been little research using SSF until Shackelford et al. (2004) studied the repeatability of SSF in lamb loins. Shackelford et al. (2004) conducted experiments comparing the effects of two cookery methods (belt grilling and electric broiler) and two shear force procedures (WBSF and SSF). Slice shear methods were modified from the SSF protocol of beef longissimus tenderness (Shackelford et al. 1999b) to better fit the size difference of lamb loin chops.

Shackelford et al. (2004) determined lamb chops (longissimus muscle) cooked using a belt grill and SSF techniques were found to have repeatability estimates of 0.95 (n = 87).

However, repeatability estimates were greater for WBSF (r = 0.93) than slice shear force

(r = 0.80) in smaller animal sample comparison (n =16) using the belt grill cooking method.

Nevertheless, due to the stronger relationship with sensory panel tenderness rating stated previously, using SSF protocol (Shackelford., 1999a) could provide to be more advantageous to reproduce tenderness scores on a hot basis, thus simulating results closer to consumer tenderness response.

Table 2.1. U.S. commercial lamb, yearlings, and mature sheep slaughter by region in 2016 Region1 Total Share Lamb & Mature Yearlings Slaughter Slaughter (1,000 head) (%) (1,000 head) (1,000 head) 1 34.1 1.7 32.8 1.3 2 195.2 9.7 176.9 18.3 3 122.5 6.1 109.0 13.5 4 77.4 3.9 64.9 12.5 5 344.7 17.2 317.4 27.3 6 39.0 1.9 34.0 5.0 7 9.2 0.5 8.8 0.4 8 849.7 42.3 829.9 19.8 9 297.6 14.8 290.5 7.1 10 40.3 2.0 37.5 2.8 Total U.S. 2009.7 100.0 1901.7 108.0 * Source: USDA/ NASS, 2017 1 Region 1: CT, ME, NH, VT, MA, RI Region 2: NY, NJ Region 3: DE-MD, PA, WV, VA Region 4: AL, FL, GA, KY, MS, NC, SC Region 5: IL, IN, MI, MN, OH, WI Region 6: AR, LA, NM, OK, TX Region 7 IA, KS, MO, NE Region 8: CO, MT, ND, SD, UT, WT Region 9: AZ, CA, HI, NV Region 10: AK, ID, OR, WA

Figure 2.1. U.S. commercial sheep and lamb slaughter, average live and expected carcass weight by region in 2016 Source: USDA/ NASS, 2017 * One value for New England data 1 Expected carcass weight = Average live weight x 0.49 (Average dressing percent of lamb) 2 Region 1: CT, ME, NH, VT, MA, RI Region 2: NY, NJ Region 3: DE-MD, PA, WV, VA Region 4: AL, FL, GA, KY, MS, NC, SC Region 5: IL, IN, MI, MN, OH, WI Region 6: AR, LA, NM, OK, TX Region 7 IA, KS, MO, NE Region 8: CO, MT, ND, SD, UT, WT Region 9: AZ, CA, HI, NV Region 10: AK, ID, OR, WA

Figure 2.2. Illustration of degree flank fat streaking in lamb carcasses *Source: Meat Evaluation Handbook, American Meat Science Association, 2001

CHAPTER 3

MATERIALS AND METHODS

Animal selection

Forty-eight lambs originating from three Western US regions (16 lambs; 8 ewes and 8 wethers from each location) and similar breed composition (Suffolk cross), were selected to represent different carcass weight classes (heavy and light market weights), and age at time of harvest (5 and 12 mo.). Lambs in the present study were intended to represent different production systems and US market channels within the lamb category. Lambs obtained from Montana (LW5) were 5 mo. of age at harvest and had light carcass weights of 31.8 kg (high concentrate diet after weaning). Utah (LW12) lambs were 12 mo. of age at harvest and had light carcass weights of 35.1 kg (forage based diet after weaning and finished with concentrate diet). Wyoming (HW12) lambs were 12 mo. of age at harvest and had heavy carcass weights of 57.9 kg (concentrate diet after weaning). Location of sourced lambs were considered treatments, with comparisons among and within lamb age and weight at harvest assessed in the present study. Lambs were harvested and carcasses were electrically stimulated (improve tenderness step) in a USDA inspected facility (JBS;

Greely, CO) and carcass measurements were recorded by trained personnel from Colorado

State University. Lamb carcasses were fabricated into wholesale cuts (rack, whole bone-in shoulder, and whole bone-in lamb leg), cuts were vacuum packaged with corresponding

ID tags. Wholesale cuts were shipped to The Ohio State University meat laboratory for fabrication and further research.

Fabrication

At 14 d post-harvest, fresh wholesale cuts were fabricated to yield individual muscle cuts. The Longissimus thoracis (LT) was removed from the lamb rack (IMPS 204,

NAMP) and all subcutaneous fat was removed. A 2.54 cm thick chop was cut from the posterior end of the LT, butterflied and allowed to bloom for 20 minutes prior to instrumental color assessment. A Model CR-410 Minolta Chroma Meter (Minolta Corp.

Ramsey, NJ) fitted with a 50-mm diameter orifice and using a D65 illuminant standardized against a white tile, was used to assess LT lightness (L*), redness (a*) and yellowness (b*).

The remaining LT samples and butterflied chop samples were labeled for carcass identification, vacuum packaged and frozen at -25°C for future lipid, pH, shear force, and sensory evaluation.

Fresh bone-in square cut shoulders (IMPS 207, NAMP) were deboned, trimmed of excess exterior fat, ground through a course plate, and frozen at -25 °C for ground shoulder patty sensory analysis and lipid extraction. The Gluteus medius (GM) was removed from the lamb leg sirloin boneless roast (IMPS 234G, NAMP) and the Semimembranosus (SM) was removed from the lamb leg top roast boneless (IMPS 234E, NAMP), and muscles were denuded of external fat and silver skin. Muscle cuts were kept frozen at -25 °C on 14 d post-harvest, until sensory analysis and shear force testing.

Lipid extraction

Total lipid extraction was conducted on the butterflied LT lamb chop and ground shoulder samples following the methods of Fisher et al. (2013). Samples were finely ground into powder with use of liquid nitrogen and a mortar and pestle. Two replicates of powered 2 g samples were placed inside of double sheeted filter paper for lipid extraction

(remaining LT samples were held in freezer (-25°C) until pH analysis was conducted).

Samples were labeled and weighed within 2 sheets of folded filter papers to obtain raw weights. Sample packets were freeze dried using a Labconco Freezedryer-6 (Labconco,

Kansas City, MO) for 22-24 h, removing any moisture present within ground samples.

After freeze drying, samples were weighed once again to determine moisture content (%).

Samples were then placed into glass cylinder terminals (Soxhlet glassware) for lipid extraction by conducting soxhlet extraction. Each cylinder contained one liter of 87:13 solution of chloroform:methanol and allowed to run for 12 hrs. Samples were allowed to vent under a fume hood for 30 min and placed into drying oven set at 100°C to ensure samples were free of moisture. Finally, samples were weighed again to determine total lipid percentage.

Determination of muscle pH

Ten grams of ground (powdered) LT muscle was put into a micro centrifuge tube, and a solution containing 5 mM sodium iodoacetate and 150 mM KCL (pH 7.0) at a 1:8

(wt/vol) ratio was added to each sample and homogenized (Bendall, 1973). Samples were centrifuged for 5 min at 10,000 rpm and placed in a heating block set at 25°C until pH was measured using a pH meter with a Accumet Basic semi-micro pH glass electrode

(Fisher Scientific, Waltham, MA). The pH probe was cleaned between samples and the meter was calibrated daily.

Cooking procedure

Longissimus thoracis (LT), ground shoulder, Gluteus medius (GM), and

Semimembranosus (SM) samples were randomly allocated by treatment to cooking and sampling day. The LT, GM and SM muscle samples were thawed overnight at 0-4oC. On cook day, whole muscle samples were removed from packages and trimmed of excess external fat and connective tissue. Ground shoulder samples were formed into a patty using a burger press template (Weber 6483 Original Burger Press) with patties weighing approximately 227 g prior to cooking. Samples were weighed to determine raw weight and cooked using clamshell cooking methods (George Foreman GRP99 Next Generation Grill with nonstick removable plates), preheated to a surface temperature of 190oC. The internal temperature of samples was monitored using a ThermaData Thermocouple® Logger KTC

(2 External K type mini-connector inputs) with thermocouple probe (mini needle probe,

8.5 cm, 1.016 mm diameter Type K). A thermocouple probe was inserted into the geometric center of each sample for internal temperature readings. Prior to cooking, the internal temperature was recorded to obtain the initial internal temperature. Whole muscle samples were removed once internal temperature reached 65oC, while shoulder patties were removed at an internal temperature of 71oC. Whole muscle cuts were cooked to 65oC to simulate the temperature cuts are most commonly cooked to among general consumers.

Pull and peak temperatures of samples were also recorded. Cooked weight was recorded

29 on each sample. Cooking loss percentage was calculated by: ((raw weight – cooked weight)/raw weight)) × 100.

Slice Shear Force procedure

Slice shear force (SSF) was determined using a modified version of the protocol developed by Shackelford et al. (2004). Due to the muscle size, shape, and thickness consistency, SSF determination was only performed on the Longissimus thoracis (LT) and

Semimembranosus (SM) muscles and not the Gluteus medius (GM) muscle. Cooked whole muscle LT and SM samples were cut into 2.5 cm chops using a cut box as a template. A minimum of two chops were obtained from each LT sample, while SM muscles were large enough to obtain multiple chops. Cuts were made on each lateral end of the chops to determine muscle fiber orientation, placed in a slice box and centered to match the 45o cutting slots of slice box. Using a double-bladed knife, a cut was made parallel to the muscle fibers to obtain a 1 cm thick by 2.54 cm width slice from LT samples and 1 cm thick by 5.0 cm width slice from the SM samples. Furthermore, to achieve the same slice length in that of beef (5 cm) in the LT (Shackelford et al., 1999a; 1999b), a second slice shear sample from the second chop was laid side by side, matching muscle fiber orientation of the first slice, in the shearing apprentice. SM slices matched the width (5 cm) of beef samples. Each sample was sheared with a flat, blunt-end blade using the electric testing machine (Model TA. XT 2plus), making sure the shearing blade cut perpendicular to the muscle fibers. The crosshead speed was set at 500 mm/min. Peak force was recorded and measured in kilograms.

Taste panel procedure

Muscle samples from the Longissimus thoracis (LT), shoulder patty, Gluteus medius (GM), and Semimembranosus (SM) were used for sensory testing. After cooking samples and obtaining shear force slices from only the LT and SM, samples were cut into

1.5 cm cubes for sensory analysis. Cubed sensory samples were placed into corresponding

ID Ziploc freezer bags and held in a water bath maintained at 65oC until consumed by panelists.

Taste panels for LT and shoulder patty samples were both conducted over 8 d period, while GM and SM were each conducted over a 2 d period. Taste panels were conducted 3 times a day (9am, 12pm, and 3pm) and panelists consisted of volunteer graduate students and office staff within the Department of Animal Sciences at The Ohio

State University. Before testing, panelists were provided instructions for the sensory evaluation process and were asked to not eat for 1 h prior to testing sessions. Panelists were provided a record sheet for each sample, and a set of palate cleansers including: unsalted crackers, distilled water, and apple juice. Eight cooked samples were consumed in a session, and order of dissemination was random. Panelists were instructed to chew samples for at least 5 to 10 s to fully obtain all flavor characteristics present within samples. Record sheets asked panelists to rate lamb flavor and off-flavor on a scale from 0 to 100 (mild = 0 and very intense = 100) for each sample. In addition, panelists were asked to record the presence of specific flavor attributes, choosing from a pre-selected array of flavors. Flavor characteristics listed on recorded sheets included the five basic flavors: sweet, sour, salty, bitter, umami/meaty; meat based flavors: browned, metallic, livery, bloody; off flavors:

31 grassy, fecal/barnyard, urine/ammonia; and asked to rate the intensity of the observed flavor(s) on a (mild = 0 and very intense = 100).

Statistical analysis

Statistical analysis was performed using the MIXED procedure in SAS (SAS Inst.

Inc., Cary, NC) to analyze carcass characteristics, muscle quality characteristics, and SSF of muscle cuts. The statistical model used was: Yij = μ + Li + Sj + LSij + eij, where Li = treatment, Si = sex, LSij = the interaction of treatment and sex, and eij = the error.

Location (lamb sourced from Montana (LW5), Utah (LW12), and Wyoming (HW12)) possessed confounding factors of age (5 mo. and 12 mo.) and body weight (light and heavy weight). The FREQ procedure in SAS was used to summarize quality grade, yield grade, and break joint percentage by treatment (location) and sex. Significance of quality grade and break joint was determined using the GLIMMIX procedure using a binary distribution model with the logit function, estimating the probability samples would classify into quality grade (USDA Choice and Prime) and break joint status (break and

휋푖푗 spool joints). The statistical model used was: log ( ) = μ + Li + Sj + LSij with Yij ~ 1− 휋푖푗

0 푖푓 퐶ℎ표푖푐푒 표푟 푏푟푒푎푘 푗표푖푛푡 ij ij { ; i i ij Binary (π ); Y = 1 푖푓 푃푟푖푚푒 표푟 푠푝표표푙 푗표푖푛푡 where L = treatment, S = sex, LS = the interaction of treatment and sex.

Taste panel data was analyzed using the GLIMMIX procedure in SAS (SAS Inst.

Inc., Cary, NC). The statistical model used was: Yij = μ + Li + Sj + LSij + Pk + eijk, where

Li = treatment, Si = sex, LSij = the interaction of treatment and sex, Pk = panelist listed as a random effect, and eijk = the error. For sensory traits, a data log transformation was conducted (if necessary) when a non-normal distribution residual of variance was 32 observed. The FREQ procedure in SAS was used to detect the frequency of flavor characteristics identified by panelists. Significance was determined using the GLIMMIX procedure using a binary distribution model with the logit function, estimating the probability a given sample would express the flavor characteristics. The statistical model

휋푖푗푘 used was: log ( ) = μ + Li + Sj + LSij + Pk with Yijk ~ Binary (πijk); Yijk = 1− 휋푖푗푘

0 푖푓 푝푎푛푒푙푖푠푡 푟푎푡푒푑 푓푙푎푣표푟 푐ℎ푎푟푎푐푡푒푟푖푠푡푖푐 0 { ; where Li = treatment, Si = sex, LSij = the 1 푖푓 푝푎푙푒푙푖푠푡 푟푎푡푒푑 푓푙푎푣표푟 푐ℎ푎푟푎푐푡푒푟푖푠푡푖푐 > 0 interaction of treatment and sex, Pk = panelist listed as a random effect (if flavor characteristic was detected the sample would be rated as “1”, whereas no detection would be rated as “0”).

The LSMEANS and DIFF statements were used to record treatment means, standard errors, and differentiate differences between treatment levels. At P < 0.05, differences between treatments were considered significant.

CHAPTER 4

RESULTS AND DISCUSSION

Carcass characteristics

Simple means of lamb carcass characteristics are presented in Table 4.1, while least square means of lamb carcass characteristics are presented in Table 4.2. Heavy weight 12 mo. lamb carcasses (HW12) possessed a larger ribeye area (P < 0.01); and had greater backfat and body wall thickness measurements (P < 0.01), when compared with both 5

(LW5) and 12 mo. light weight (LW12) lamb carcasses. Published research (Crouse et al.,1981; Borton et al., 2005a) has also noted greater carcass measurements from heavy weight lambs when compared with lighter weight lambs. However, the excess fat thickness opposite the ribeye (BF) and body wall of heavy weight carcasses would lead to a decreased percentage of retail cuts, resulting in a less profitable margins for meat packers. Treatment by sex effects were shown to be significant for ribeye area (P = 0.0138) and back fat (P =

0.0132). Heavy weight 12 mo. ewes had significantly larger ribeye area (P = 0.003) and less backfat (P = 0.002) when compared with heavy weight wethers. These results indicate

12 mo. wethers, when fed to heavier weights, were shown to be fatter and lighter muscled in comparison to contemporary ewe lambs. In contrast, Wylie et al. (1997) reported ewe lambs fed to heavier slaughter weights (40 - 48 kg carcass weight) were fatter over the 12th rib, and possessed a smaller ribeye area when compared with contemporary wether

34 carcasses. However, in the present study, no treatment by sex effects were observed for carcass ribeye area or backfat when comparing sexes within light weight lambs treatment groups (LW5 and LW12). Between all 3 treatments, HW12 carcasses showed the highest yield grades (YG 4), LW12 had intermediate (YG 3) yield grade, and LW5 possessed the lowest yield grade score (YG 2) (P < 0.01), indicating that carcass lean percentage was greater, as expected, when lambs were marketed at a lighter end weight. The LW5 lambs carcasses possessed lower flank streaking scores (P < 0.01) than LW12 and HW12 carcasses (Slight vs. Modest), indicating age at slaughter appears to influence degree of flank streaking regardless of weight end point. Similarly, Field et al. (1990) noted degree of fat streaking increased as age at time of slaughter increased in ewe and wether lambs.

However, no differences were noted in carcass characteristics between ewe and wether lambs, similar to the present study (P > 0.05).

Quality and yield grade frequency distributions within each treatment group are presented in Table 4.1, while least square means are shown in Table 4.2. A higher frequency of LW5 carcasses quality graded Choice (13/16), while carcasses from LW12 lambs were more balanced between Choice (9/16) and Prime (6/16) grades, and over one- half (9/16) of HW12 carcasses graded in the Prime category. As expected based on the age at harvest alone, all LW5 lambs possessed break joints (indicative of a more youthful carcass), whereas LW12 lambs had one of sixteen carcasses with a spool joint(s), and three of sixteen HW12 carcasses had identified spool joint(s). Across treatments, wether lamb carcasses had higher frequencies of break joints (23/24 break joints), compared with ewe lambs (21/24 break joints). Though, not significant (P > 0.05), his finding can be expected

35 as ewe lambs have been reported to mature at a faster rate and reach puberty at an earlier age when compared with wether lambs, resulting in higher instances of spool joints at the time of puberty (~ 12 mo. of age) (Ho et al.,1989). Over half of the light weight carcasses

(LW5 and LW12) were primarily classified into the yield grade 3 category, while a majority (84%) of HW12 carcasses were designated as a yield grade 5. Lastly, there were no notable differences in the yield grade when comparing sex across treatments (P > 0.05).

Lipid concentration

Lipid concentrations from the Longissimus thoracis (LT) and ground shoulder patties are presented in Table 4.3. In the LT, 12 mo. lambs possessed a greater lipid concentrations (P < 0.01) than LW5 lambs, indicating greater intramuscular fat (marbling) deposition in the LT as age at slaughter increased. In the present study, intramuscular fat deposition was shown to be more dependent on age rather than weight, even with the large differences in backfat and body wall thickness noted earlier in the manuscript for heavy weight lambs. The increase in intramuscular fat with an increase in age is expected, as reported in previous studies (Hopkins et al. 2007; Pethick et al. 2005b). Furthermore, across all treatments wether lambs had greater (P < 0.05) LT lipid concentrations than ewe lambs.

In contrast, Okeudo et al. (2007) noted no differences between ewe and wether lamb intramuscular fat content. A treatment by sex interaction (Fig. 4.1) was noted in the present study, whereby 12 mo. wethers had a significantly greater LT lipid concentrations when compared with contemporary ewe samples, while no sex difference in LT lipid concentration was observed in the LW5 treatment group.

Lipid concentrations of the shoulder patty were greatest (P < 0.05) in heavy weight lambs, as expected due to overall greater fat content, indicative of a greater amount of intermuscular (seam fat) and subcutaneous fat in the shoulder regions. This is in conjunction with greater backfat and body wall thickness measurements as noted earlier.

However, there were no significant differences in shoulder patty lipid concentrations due to sex (P > 0.05) within each treatment.

Longissimus pH and Color

Muscle pH and color measurements from Longissimus thoracis (LT) chops are shown in

Table 4.4. The ultimate pH of LT chops was not significantly influenced by age (P = 0.25), weight (P = 0.22), or sex (P = 0.55). Similarly, Hopkins et al. (2007) reported no significant differences in pH at 24hrs in the longissimus lumborum muscle of lambs of 4, 8, 14, and

22 mo. of age. In the present study, Minolta L* values from the LT were greater indicating a lighter appearance (P < 0.01) to the LT in LW5 and HW12 lambs when compared with the LT of LW12 lambs, while no differences (P > 0.05) were detected between sexes within treatment groups. Furthermore, Minolta a* (redness) and b* (yellowness) values for the LT were shown not to be different between treatments. Minolta a* values were greater (P <

0.05) or redder in color in the LT from wethers when compared with the LT of ewe lambs.

However, while statistically significant in the present study, these differences (< 2-unit change) would be difficult to detect by the human eye. Consumer acceptability threshold values for lightness (L*) and redness (a*) have been reported by Hopkins et al. (1996) as lower limits of 34 and 19 units, respectively. Values in the present study average from

38.16 – 41.29 (L*, lightness) and 25.48- 26.76 (a*, redness) well above the reported lower

37 acceptability thresholds and reflect lamb meat that meets satisfactory visual consumer acceptability. Nevertheless, Khiliji et al. (2010) noted L* and a* values in fresh lamb must be much greater (44 units for L* and 14.5 units for a*) to have 95% confidence a random consumer would consider a sample acceptable. In the Present study, LT color parameters meet the a* criteria noted by Khilijii et al. (2010) but are reported to be darker than their expectations for L*.

Slice Shear force

Slice Shear force (SSF) measurements for the Longissimus thoracis (LT) and

Semimembranosus (SM) muscles are presented in Table 4.5. The SSF values were shown to be greater (P < 0.01) or tougher in older 12 mo. lambs (LW12 and HW12) when compared with the young, 5 mo. lambs in both the LT and SM muscles. Hopkins et al.,

(2007) reported significant effects between animal age and tenderness, reporting meat cuts from young lambs (4 and 8 mo. old) produced lower shear force values after five days of aging in the longissimus lumborum muscle compared with 14 mo. lambs. Similarly, Bouton et al. (1978) evaluated shear force values of four leg muscles (biceps femoris, semimembranosus, semitendinosus and gluteus medius) and determined peak force required to cut through samples increased with animal age (2 mo. – 6 yrs.). No significant differences (P > 0.05) in shear force measurements due to sex of the lamb were across treatments. Lastly, Shackelford et al. (2012) reported mean SSF values in lamb longissimus muscle (aged 7 d postmortem), with an average age of 7 mo. at time of slaughter, to range between 19.8 to 26.3 kg. Additionally, Jaborek et al. (2017) reported SSF values of 16.7,

13.1, and 18.74 kg in lamb, yearling, and mature sheep, respectively, in the LT muscle

(aged for 14 d postmortem). In the present study, SSF values were shown to be lower in comparison with previous studies, in the range of 10.0 to 13.3 kg; this could be due to the combination of a longer aging period (14 d postmortem) of LT samples, with the use of electrical stunning and electrical stimulation during harvest, thus increasing tenderness.

Captive-bolt stunning without use of electrical stimulation methods were reported in both

Shackelford et al. (2012) and Jaborek et al. (2017) studies. Furthermore, SSF values in the present study would be considered “very tender” in conjunction with previous studies reporting SSF values below 14.5 kg in the beef longissimus muscle (Igo et al., 2015).

Taste panel

Table 4.7. illustrates consumer lamb flavor and off flavor intensity response scores, after data transformation (shown in table 4.6), for treatment lambs in sampled muscle cuts from the Longissimus thoracis (LT), Semimembranosus (SM) Gluteus medius (GM), and ground shoulder patty. Lamb flavor intensity scores in the LT were found to be lower in the LW5 group when compared with both LW12 and HW12 groups

(P < 0.05). Furthermore, panelists rated LW5 lambs lower in off-flavor intensity in the

LT than LW12 and HW12 lambs. No differences (P > 0.05) in flavor and off flavor intensities were noted for SM, GM, and shoulder patty samples between treatment groups. However, treatment by sex interactions were present for the ground shoulder patty flavor and off-flavor intensities (Figs. 4.2 & 4.3). Light weight 12 mo. old wethers and heavy weight 5 mo. old ewes possessed lower flavor intensity scores (P < 0.05) when compared with all other weight/age and sex treatment combinations (Fig.4.2). While light weight 12 mo. ewes possessed significantly (P < 0.05) higher off flavor scores

39 inconsistent to all other treatment and sex combinations in the present study (Fig. 4.3). In contrast, Batcher et al. (1969) noted higher intensity of flavor due to increased age in lamb broth (cooked lamb in water) created from muscles derived from 7 to 8 mo. old lambs when compared with broth derived from muscles from 15 to 16 mo. old yearling mutton; however, they reported no differences in flavor of sliced meat (braised shoulder, broiled lamb chops, and roasted leg) for the two age groups evaluated. Jeremiah et al.

(1998) studied flavor and texture profiles in boneless shoulder roasts, comparing age and slaughter weights. Jeremiah et al. (1998) noted increasing slaughter weights (over 50 kg) produced more balanced/beneficial effects in flavor in comparison to lighter slaughter weights. Overall, in the present study, lamb intensity scores across muscles averaged towards a relatively mild flavor (> 41), and low off- flavor intensity scores (> 3), considering the scale from 0-100.

Flavor characteristics noted by panelists for the sampled muscle cuts are presented in Table 4.8. Of the flavor characteristics listed, umami, browned, metallic, and livery were the most frequently reported flavors noted by panelists across treatments.

Urine, fecal, and grassy (pastoral flavors) were less frequently noted by panelists as these are more commonly produced by pasture-fed lambs (Schreurs et al., 2008). In the LT, umami was used more frequently to describe LW12 lambs, while browned flavor was more frequently used to describe HW12 lambs, and LW5 lambs had the lowest instances of browned and livery flavors. In the SM, metallic, and livery were used most frequently used to describe LW12 lambs, while LW5 lambs had the least occurrences of browned and livery notes. It was noted LW12 lambs possessed a higher probability of displaying a

40 more livery flavor characteristic in the SM, with comparison to LW5 (P = 0.001) and

HW12 (P = 0.041) lambs. In the GM, metallic, livery, and bloody flavors were most commonly used to describe LW12 lambs, and showed a higher probability (P = 0.019) of possessing a more livery flavor characteristic in comparison to LW5. In shoulder patty samples, grassy flavors were used more frequently to describe LW12 lambs. On the other hand, LW5 lambs possessed the lowest instances of bitter and umami flavors. Differences between ewe and wether lambs across treatments were only shown in the GM muscle, as ewe lambs showed a higher probability of sour flavor notes in comparison to wether lambs (P = 0.019). No treatment by sex interactions were noted in describing flavor characteristics across muscles (P > 0.05).

Conclusion

The present study outlined the influences of three common sheep production practices within the U.S. lamb industry on lamb composition, quality and flavor profiles.

Feeding lambs a high concentrate diet after weaning (HW12), and reaching heavier end weights clearly added excess fat and proportionately less muscle, resulting in a substantially lower percentage of closely trimmed retail cuts. Twelve month lambs fed a forage based diet after weaning and finished on concentrate diet (LW12) did not improve composition relative to lambs fed a concentrate diet and reaching market weight at a younger age (LW5). Muscle pH, color, and intramuscular fat levels were moderately influenced when sheep were fed to an older age in comparison to younger lamb. At an older age, lamb was considered tougher in both muscle cuts sampled (LT and SM) and possessed greater flavor and off flavor intensity in the longissimus thoracis muscle, but

41 had similar flavor and off-flavor intensity in other sampled muscle cuts compared with meat derived from younger sheep. Of note, umami, browned, liver and metallic flavor profiles were most frequently observed by panelists across treatments. However, livery flavor characteristics was noted as being more consistent with light weight 12 mo. lambs in the semimembranosus and gluteus medius muscles in comparison to other treatments.

Overall, lambs in the present study possessed mild flavor and low instances of off flavors in all muscle cuts sampled.

Table 4.1. Simple means and standard deviations (± SD) of lamb carcass characteristics and frequency of carcass grades by treatment and sex.

Treatment1 Sex LW5 LW12 HW12 Ewe Wether Item (n=16) (n=16) (n=16) (n=24) (n=24) Hot carcass weight, kg 31.8 ± 4.2 35.1 ± 9.8 57.9 ±10.4 40.7 ± 12.9 42.5 ±12.0 Ribeye area, cm2 13.4 ± 1.9 14.9 ± 2.9 20.1 ± 4.8 16.6 ± 5.4 15.6 ± 3.1 Carcass Conformation2 12.6 ± 0.9 12.8 ± 0.7 13.1 ± 0.8 12.8 ± 0.8 12.9 ± 0.8 Back fat, cm 0.72 ± 0.23 0.83± 0.28 1.04 ± 0.47 1.04 ± 0.50 1.19 ± 0.67 Body wall thickness, cm 2.26 ± 0.34 2.46± 0.54 4.43 ± 0.62 3.03 ± 1.04 3.08 ± 1.20 Yield grade3 2.67 ± 0.49 3.26± 0.59 4.75 ± 0.60 3.45 ± 0.99 3.67 ± 1.08 Flank streaking score4 349 ± 90 490 ± 96 558 ± 145 466 ± 153 464 ± 132 Quality grade* Choice %, count/total 81.3, 13/16 56.3, 9/16 30.8, 4/16 59.1, 13/24 56.5, 13/24 Prime %, count/total 12.5, 2/16 37.5, 6/16 69.2, 9/16 31.8, 6/24 43.5, 9/24 Break joint %, count/total 100.0, 16/16 93.6, 15/16 81.3, 13/16 87.5, 21/24 95.8, 23/24

43 Yield grade

2, % 33.3 6.7 - 15.0 13.0 3, % 66.7 60.0 7.7 50.0 44.0 4, % - 33.3 7.7 15.0 13.0 5, % - - 84.6 20.0 30.0 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 2 Carcass conformation are based on a numeric scale: 11= - Choice, 12= +Choice, and 13= Prime 3 Yield grades are based on a numeric scale: 1 (most muscular, lean) – 5 (excess fat, light muscle) 4 Flank streaking score is subjective and based on numeric scale: 300-399 = slight, 400-499= small, 500-599= modest * Lamb carcass indicating two spool joints (advanced maturity) were not eligible to grade; therefore % Choice, % Prime does not total 100%

Table 4.2. Least-square means of lamb carcass characteristics within treatment and sex. 1 Treatment Sex 2 2 Item LW5 LW12 HW12 SEM Ewe Wether SEM (n=16) (n=16) (n=16) (n=24) (n=24) b b a Hot carcass weight, kg 31.8 35.1 57.9 0.9 40.7 42.5 0.8 2 b b a Ribeye area, cm 13.4 14.9 20.1 0.8 16.6 15.6 0.6 3 Carcass Conformation 12.6 12.8 13.1 0.2 12.8 12.9 0.2 b b a Back fat, cm 0.70 0.83 1.04 0.1 1.04 1.19 0.06 Body wall thickness, b b a cm 2.26 2.46 4.43 0.13 3.03 3.08 0.11 4 c b a Yield grade 2.67 3.26 4.75 0.16 3.45 3.67 0.13 Quality grade5 0.99 0.61 0.28 0.13 0.99 0.57 1.48 Flank streaking score6 349b 490a 558a 29 466 464 23 Break joint7 1.00 0.99 0.82 0.10 0.99 1.00 0.57 a,b,c Treatment means within a row without a common superscript letter significantly differ (P < 0.01) 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 44 2 SEM= standard error of the mean. SEM reported is the greatest between treatments 3 Carcass conformation are based on a numeric scale: 11= - Choice, 12= +Choice, and 13= Prime 4 Yield grades are based on a numeric scale: 1 (most muscular, lean) – 5 (excess fat, light muscle) 5 Quality grades were based on probability estimates of lambs grading USDA Choice; higher value = greater number of Choice lambs 6 Flank streaking score is subjective and based on numeric scale: 300-399 = slight, 400-499= small, 500-599= modest 7 Break joints were based on probability estimates of lambs having break joints; higher value = greater number of break joints

Table 4.3. Least square means of lipid concentrations (%) from the Longissimus thoracis and shoulder muscle cuts within treatment and sex. Treatment1 Sex Muscle Cut LW5 LW12 HW12 SEM2 Ewe Wether SEM2 (n=16) (n=16) (n=16) (n=24) (n=24) Longissimus 2.92b 5.10a 4.78a 0.37 3.82e 4.71d 0.31 thoracis Shoulder patty 17.7e 18.90e 23.47d 1.42 19.95 20.14 1.16 a,b Treatment means within a row without a common superscript letter differ (P < 0.01) d,e Treatment and sex means within a row without a common superscript letter differ (P < 0.05) 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 2 SEM= standard error of the mean. SEM reported is the greatest between treatments

Table 4.4. Least square means for Longissimus thoracis pH and instrumental color measurements within treatments and sex.

Treatment1 Sex Item LW5 LW12 HW12 SEM2 Ewe Wether SEM2 (n=16) (n=16) (n=16) (n=24) (n=24) pH 5.69 5.73 5.67 0.03 5.68 5.70 0.02 L* 3 41.19a 38.16b 40.45a 0.53 39.64 40.23 0.43 a* 4 26.38 25.48 26.76 0.41 25.65e 26.76d 0.33 b* 5 7.89 7.53 8.74 0.35 7.66 8.45 0.28 a,b Treatment means within a row without a common superscript letter significantly differ (P < 0.01) d,e Sex means within a row without a common superscript letter significantly differ (P < 0.05) 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 2 SEM= standard error of the mean. SEM reported is the greatest between treatments 3 L* = lightness 4 a* = redness 46 5 b* = yellowness

Table 4.5. Least square means of slice shear force (kg) from the Longissimus thoracis and Semimembranosus muscles of lamb carcasses within treatment and sex.

Treatment1 Sex Muscle Cut LW5 LW12 HW12 SEM2 Ewe Wether SEM2 (n=16) (n=16) (n=16) (n=24) (n=24) Longissimus thoracis 10.0b 13.3a 12.8a 0.7 11.8 12.3 0.6 Semimembranosus 17.3b 20.0a 19.4a 0.7 19.0 18.9 0.5 a,b Treatment means within a row without a common superscript letter differ (P < 0.01) 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 2 SEM= standard error of the mean. SEM reported is the greatest between treatments

Table 4.6. Log of lamb flavor and off flavor intensity consumer panel scores in within lamb muscle cuts within treatment and sex.

Treatment1 Sex Taste Muscle Cut LW5 LW12 HW12 SEM2 Ewe Wether SEM2 Intensity (n=16) (n=16) (n=16) (n=24) (n=24) Longissimus Lamb Flavor 3.32e 3.52d 3.46de 0.13 3.42 3.44 0.13 thoracis Off Flavor -0.61b 0.15a 0.18a 0.42 -0.24 0.05 0.40 Lamb Flavor 3.63 3.70 3.66 3.68 3.65 Semimembranosus Off Flavor 0.73 0.91 0.52 0.44 0.86 0.58 0.42 Lamb Flavor 3.63 3.70 3.66 3.68 3.65 Gluteus medius Off Flavor 0.72 0.91 0.53 0.44 0.86 0.58 0.42 Lamb Flavor 3.56 3.48 3.52 0.10 3.52 3.52 0.10 Shoulder patty Off Flavor 0.01 0.50 0.21 0.52 0.39 0.08 0.51 a,b

48 Treatment means within a row without a common superscript letter differ (P < 0.01) d,e

Treatment means within a row without a common superscript letter differ (P < 0.05) 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 2 SEM= standard error of the mean. SEM reported is the greatest between treatments

Table 4.7. Least squares means, after data transformation, of lamb flavor and off flavor intensity consumer panel scores in various lamb muscle cuts by treatment and sex.

Treatment1 Sex Taste Muscle Cut LW5 LW12 HW12 SEM3 Ewe Wether SEM3 Intensity2 (n=16) (n=16) (n=16) (n=24) (n=24) Lamb Flavor 27.7e 33.7d 31.7de 30.7 31.2 Longissimus thoracis Off Flavor 0.5b 1.2a 1.2a 0.8 1.1

Lamb Flavor* 37.9 40.5 38.9 4.5 39.7 38.5 4.4 Semimembranosus Off Flavor 2.1 2.5 1.7 2.4 1.8 Lamb Flavor* 37.8 40.5 38.9 4.5 39.7 38.5 4.4 Gluteus medius Off Flavor 2.1 2.5 1.7 2.4 1.8

Lamb Flavor 35.3 32.5 33.6 33.8 33.7 Shoulder patty 49 Off Flavor 1.0 1.6 1.2 1.5 1.1

a,b Treatment means within a row without a common superscript letter differ (P < 0.01) d,e Treatment means within a row without a common superscript letter differ (P < 0.05) 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age 2 Taste intensity score numeric scale: 0 (mild) – 100 (intense) 3 SEM= standard error of the mean. SEM reported is the greatest between treatments * Data transformation not needed

Table 4.8. Percentage of consumer panel respondents identifying flavor characteristics in various muscle cuts by treatment.

Treatment1

Flavor Longissimus thoracis Semimembranosus Gluteus medius Shoulder patty Characteristics LW5 LW12 HW12 LW5 LW12 HW12 LW5 LW12 HW12 LW5 LW12 HW12 % % % % % % % % % % % % Sweet 1.9 2.5 2.3 9.4 6.9 9.1 5.2 3.2 7.2 9.0 7.4 7.2 Sour 1.9 2.5 1.7 10.4 9.2 7.3 6.0 4.0 4.8 3.3 3.3 3.2 Salty 2.5 1.3 1.8 1.9 6.1 5.4 8.2 3.2 5.6 8.2 9.1 8.0 Bitter 6.9 7.6 6.4 1.9 4.3 1.8 2.2 5.6 3.2 0.8 6.6 5.6 Umami 12.6 17.1 12.9 20.7 20.0 20.9 21.6 16.9 20.0 13.1 20.7 19.2 Browned 8.2 10.1 15.8 10.4 13.1 14.6 5.2 8.1 8.8 10.7 16.5 13.6 Metallic 5.7 8.2 5.8 13.2 23.8 14.6 23.9 26.6 18.4 12.3 9.9 12.8

Livery 5.7 10.3 11.1 9.4 28.5 20.0 20.9 33.9 22.4 12.3 9.9 12.8 Bloody 1.3 3.8 0.6 9.4 17.7 6.4 7.5 13.7 6.4 2.5 3.3 2.4 Grassy 4.4 5.7 6.4 7.5 6.9 10.0 10.4 6.4 8.0 7.4 13.2 8.0 Fecal 2.5 7.0 6.4 4.7 5.4 5.4 6.0 8.1 4.0 7.4 13.2 9.6 Urine 0.6 4.4 3.5 4.7 3.1 7.3 2.2 4.8 1.6 6.6 5.0 6.4 Other* 3.1 0.6 1.7 0.9 1.5 3.6 0.0 1.6 0.8 2.5 3.3 3.2 1 Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age * Other includes, but not limited to: Fatty; Musty; Burned; etc.

a a

b b b b

Figure 4.1. Treatment by sex interaction for longissimus thoracis lipid concentrations. Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age. Treatment least square means without a common letter (a, b) are significantly different (P ≤ 0.05)

a a a a

b b

Figure 4.2. Treatment by sex interaction for the shoulder patty flavor intensity score. Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age. Treatment least square means without a common letter (a, b) are significantly different (P ≤ 0.05)

b b b b b

Figure 4.3. Treatment by sex interaction for the shoulder patty off- flavor intensity score. Treatment: LW5= light weight lambs at 5 mo. of age; LW12= light weight lambs at 12 mo. of age; HW12= heavy weight lambs at 12 mo. of age. Treatment least square means without a common letter (a, b) are significantly different (P ≤ 0.05)

List of References

Aberle, J., Forrest, J., Gerrard, D., Mills, E. (2012). Principles of Meat Science 5th Edition.

Dubuque: Kendall.

Bailey, A.J. & Light, N.D., (1989). Connective tissue in meat and meat products. Elsevier

Applied Science.

Becker, T. (2000). Consumer perception of fresh meat quality: a framework for analysis.

British Food Journal, 102, 158-176.

Bendall, J. R. (1973). Postmortem changes in muscle. The structure and function of

muscle, 2, (pp. 243-309).

Borton, R. J., Loerch, S. C., McClure, K. E., & Wulf, D. M. (2005a). Characteristics of

lambs fed concentrates or grazed on ryegrass to traditional or heavy slaughter

weights. II. Wholesale cuts and tissue accretion. Journal of Animal Science, 83,

1345-1352.

Bouton, P.E., Harris, P.V., Ratcliff, D. & Roberts, D.W. (1978). Shear force measurements

on cooked meat from sheep of various ages. Journal of Food Science, 43, 1038-

1039.

Crouse, J. D., Busboom, J. R., Field, R. A., & Ferrell, C. L. (1981). The effects of breed,

diet, sex, location and slaughter weight on lamb growth, carcass composition and

meat flavor. Journal of Animal Science, 53, 376-386.

Field, R. A., Williams, J. C., & Miller, G. J. (1983). The effect of diet on lamb flavor.

Food Technology, 37, 258-263.

Field, R.A., Maiorano, G., McCormick, R.J., Riley, M.L., Russell, W.C., Williams, F.L.

& Crouse, J.D. (1990). Effect of plane of nutrition and age on carcass maturity of

sheep. Journal of Animal Science, 68, 1616-1623.

Fisher, K. D., Scheffler, T. L., Kasten, S. C., Reinholt, B. M., Van Eyk, G. R., Escobar,

J., & Gerrard, D. E. (2013). Energy dense, protein restricted diet increases

adiposity and perturbs metabolism in young, genetically lean pigs. Plos One, 8,

e72320.

Font I Furnols, M., Realini, C. E., Guerrero, L., Oliver, M. A., Sañudo, C., Campo, M.

M., Nute, G. R., Caneque, V., Alvarez, I., San Julian, R., Luzardo, S., Brito, G.,

and Montossi, F. (2009). Acceptability of lamb fed on pasture, concentrate or

combinations of both systems by European consumers. Meat Science, 81, 196-

202.

Garrett, R. P., Savell, J. W., Cross, H. R., & Johnson, H. K. (1992). Yield grade and

carcass weight effects on the cutability of lamb carcasses fabricated into

innovative style subprimals. Journal of Animal Science, 70, 1829-1839.

Ho, L., Field, R.A., Russell, W.C., Riley, M.L., Ercanbrack, S.K. & Williams, F.L. (1989).

Influence of gender, breed and age on maturity characteristics of sheep. Journal of

Animal Science, 67, 2460-2470.

Hoffman, T. W., Belk, K. E., Woerner, D. R., Tatum, J. D., Delmore, R. J., Peel, R. K.,

LeValley, S.B., Pendell, D.L., Zerby, H.N., English, L.F., Moeller, S. J., &

Fluharty, F.L., (2016). Preferences associated with American lamb quality in retail

& foodservice markets. Meat Science, 112, 138.

Hopkins, D.L., Hegarty, R.S., Walker, P.J. & Pethick, D.W. (2006). Relationship between

animal age, intramuscular fat, cooking loss, pH, shear force and eating quality of

aged meat from sheep. Animal Production Science, 46, 879-884.

Hopkins, D.L., Stanley, D.F., Martin, L.C., Toohey, E.S. & Gilmour, A.R. (2007).

Genotype and age effects on sheep meat production. 3. Meat quality. Animal

Production Science, 47, 1155-1164.

Huff-Lonergan, E., Parrish, F.C. & Robson, R.M. (1995). Effects of postmortem aging

time, animal age, and sex on degradation of titin and nebulin in bovine longissimus

muscle. Journal of Animal Science, 73, 1064-1073.

Huffman, K. L., Miller, M. F., Hoover, L. C., Wu, C. K., Brittin, H. C., Ramsey, C. B.

(1996). Effect of beef tenderness on consumer satisfaction with steaks consumed

in the home and restaurant. Journal of Animal Science, 74, 91-97.

Igo, M. W., Arnold, A. N., Miller, R. K., Gehring, K. B., Mehall, L. N., Lorenzen, C. L.,

Delmore JR., R.J., Woerner, D.R., Wasser, B.E., & Savell, J. W. (2015).

Tenderness assessments of top loin steaks from retail markets in four US cities.

Journal of Animal Science, 93, 4610-4616.

Jaborek, J.R., Zerby, H.N., Moeller, S.J., Wick, M.P., Fluharty, F.L., Garza III, H.,

Garcia, L.G., & England, E.M. (2017). Effect of energy source and Level, and

animal age and sex on meat characteristics of sheep. Meat Science. Submitted.

Jamora, J. J., & Rhee, K. S. (1998). The uniqueness of lamb: nutritional and sensory

properties. Sheep and Goat Research Journal (USA).

Jeremiah, L. E., Tong, A. K. W., & Gibson, L. L. (1998). The influence of lamb

chronological age, slaughter weight, and gender. Flavor and texture profiles. Food

Research International, 31, 227-242.

Khliji, S., Van de Ven, R., Lamb, T. A., Lanza, M., & Hopkins, D. L. (2010).

Relationship between consumer ranking of lamb colour and objective measures of

colour. Meat Science, 85, 224-229.

Miller, M. F., Carr, M. A., Ramsey, C. B., Crockett, K. L., & Hoover, L. C. (2001).

Consumer thresholds for establishing the value of beef tenderness. Journal of

Animal Science, 79, 3062-3068.

Murphy, M.A., Zerby, H.N. and Fluharty, F.L. (2003). The effects of energy source and

ionophore supplementation on lamb growth, carcass characteristics and tenderness.

Sheep and Goat Research Journal, 18, 89-95.

Okeudo, N. J., Moss, B. W. (2007). Intramuscular lipid and fatty acid profile of sheep

comprising four sex-types and seven slaughter weights produced following

commercial procedure. Meat Science, 76, 195-200.

Oltra, O. R., Farmer, L. J., Gordon, A. W., Moss, B. W., Birnie, J., Devlin, D. J., Tolland,

E. L. C., Tollerton, I. J., Beattie, A. M., Kennedy, & J.T., Farrell, D. (2015).

Identification of sensory attributes, instrumental and chemical measurements

important for consumer acceptability of grilled lamb Longissimus lumborum.

Meat Science, 100, 97-109.

O’Quinn, T. G., Brooks, J. C., Polkinghorne, R. J., Garmyn, A. J., Johnson, B. J.,

Starkey, J. D., Miller, M. F. (2012). Consumer assessment of beef strip loin steaks

of varying fat levels. Journal of Animal Science, 90, 626-634.

Pethick, D. W., Pleasants, A. B., Gee, A. M., Hopkins, D. L., & Ross, I. R. (2006). Eating

quality of commercial meat cuts from Australian lambs and sheep. Proceedings of

the New Zealand Society of Animal Production, 66, 363−367.

Priolo, A., Micol, D., Agabriel, J., Prache, S. & Dransfield, E. (2001). Effect of grass or

concentrate feeding systems on lamb carcass and meat quality. Meat Science, 62,

179-185.

Rhee, K. S., Ziprin, Y. A. (1996). Identification and acceptance of lamb versus beef and

pork by consumers and experienced sensory panelists. Journal of Muscle Foods, 7,

243-253.

Rousset-Akrim, S., Young, O. A., & Berdague, J.L. (1997). Diet and growth effects in

panel assessment of sheep odour and flavor. Meat Science. 45, 169-181.

Shackelford, S.D., Wheeler, T.L. & Koohmaraie, M., (1999). Tenderness classification of

beef: II. Design and analysis of a system to measure beef longissimus shear force

under commercial processing conditions. Journal of Animal Science, 77, 1474-

1481.

Shackelford, S.D., Wheeler, T.L. & Koohmaraie, M., (2004). Evaluation of sampling,

cookery, and shear force protocols for objective evaluation of lamb longissimus

tenderness. Journal of Animal Science, 82, 802-807.

Shackelford, S.D., Leymaster, K.A., Wheeler, T.L. & Koohmaraie, M. (2012). Effects of

breed of sire on carcass composition and sensory traits of lamb. Journal of Animal

Science, 90, 4131-4139.

Shiflett, J. S., Williams, G., & Rodgers, P. (2010). Nontraditional lamb market in the

United States: characteristics and marketing strategies. Nontraditional lamb market

in the United States: characteristics and marketing strategies.

Shorthose, W.R. & Harris, P.V. (1990). Effect of animal age on the tenderness of selected

beef muscles. Journal of Food Science, 55, 1-8.

Smith, G.C., Dutson, T.R., Hostetler, R.L. & Carpenter, Z.L., (1976). Fatness, rate of

chilling and tenderness of lamb. Journal of Food Science, 41, 748-756.

Smith, R. D., Nicholson, K. L., Nicholson, J. D. W., Harris, K. B., Miller, R. K., Griffin,

D. B., & Savell, J. W. (2008). Dry versus wet aging of beef: Retail cutting yields

and consumer palatability evaluations of steaks from US Choice and US Select

short loins. Meat Science, 79, 631-639.

USDA. (1992). United States standards for grades of slaughter lambs, yearlings, and

sheep. Agricultural Marketing Service, USDA, Washington, DC.

USDA-Economic Research Service. (2016).

http://usda.mannlib.cornell.edu/usda/current/LiveSlau/LiveSlau-08-24-2017.pdf.

National Agriculture Statistics Service (NASS), USDA, Washington, DC.

Veiseth, E., Shackelford, S.D., Wheeler, T.L. & Koohmaraie, M. (2004). Factors regulating

lamb longissimus tenderness are affected by age at slaughter. Meat Science, 68,

635-640.

Waibel, R. A, F. A. Q. (2008). Religion and dietary practices. Available from:

http://www.faqs.org/nutrition/Pre-Sma/Religion-and-Dietary-

Practices.html#ixzz1LsxWnBf6.

Ward, C. E., Trent, A., & Hildebrand, J. L. (1995). Consumer perceptions of lamb

compared with other meats. Sheep and Goat Research Journal (USA).

Watkins, P. J., Frank, D., Singh, T. K., Young, O. A., & Warner, R. D. (2014). Sheepmeat

flavor and the effect of different feeding systems: a review. Journal of Agricultural

and Food Chemistry, 61, 3561-3579.

Watkins, P. J., Rose, G., Salvatore, L., Allen, D., Tucman, D., Warner, R. D., & Pethick,

D. W. (2010). Age and nutrition influence the concentrations of three branched

chain fatty acids in sheep fat from Australian abattoirs. Meat Science, 86, 594-599.

Wong, E., Johnson, C. B., & Nixon, L. N. (1975a). The contribution of 4-methyloctanoic

(hircinoic) acid to mutton and goat meat flavour. New Zealand Journal of

Agricultural Research, 18, 261-266.

Wong, E., Nixon, L. N., & Johnson, C. B. (1975b). Volatile medium chain fatty acids and

mutton flavor. Journal of Agricultural and Food Chemistry, 23, 495-498.

Wylie, A. R. G., Chestnutt, D. M. B., & Kilpatrick, D. J. (1997). Growth and carcass

characteristics of heavy slaughter weight lambs: effects of sire breed and sex of

lamb and relationships to serum metabolites and IGF-1. Animal Science, 64, 309-

318.

Young, O. A., Berdagué, J. L., Viallon, C., Rousset-Akrim, S., & Theriez, M. 1997. Fat-

borne Volatiles and Sheepmeat Odour. Meat Science, 45, 183-200.

Young, O.A. & Braggins, T.J. (1992). Tenderness of ovine semimembranosus: Is collagen

concentration or solubility the critical factor?. Meat Science, 35, 213-222.

Young, O. A., Lane, G. A., Podmore, C., Fraser, K., Agnew, M. J., Cummings, T. L., &

Cox, N. R. (2006). Changes in composition and quality characteristics of ovine

meat and fat from castrates and rams aged to 2 years. New Zealand Journal of

Agricultural Research, 49, 419-430.