Lamb growth impacts muscle

oxidative capacity and meat quality

This thesis is presented for the degree of

Doctor of Philosophy of Murdoch University

Khama Kelman

B.Sc. (Veterinary Biology)

B.V.M.S. (Veterinary Medicine and Surgery)

M.A.C.V.Sc. (Small Animal Surgery)

B.Sc. (Mathematics and Statistics)

December 2014

II

Declaration

I declare that this thesis is my own account of my research and contains as its main content

work which has not previously been submitted for a degree at any tertiary education institution

……………………………

Khama Rhian Kelman

III

Summary

Genetic selection to increase lamb growth is advantageous in reducing the time it takes

for lambs to reach market weight. However the effects of rapid growth on muscle fibre

type, as indicated by oxidative capacity, is not clear. Muscle fibre type is intrinsically

linked with attributes of meat quality, which are becoming increasing important to

consumers. Therefore, changes to muscle fibre type through selection for growth may

impact meat quality.

Producers can impact lamb growth at any time during the growth period including the

weaning (100 days) and post weaning time points (150 days). As not all lambs were

weighed at these time points a growth curve was fitted to the live weight data of each

individual animal to allow weight to be estimated at any time point during the growth

period. Once estimates of weight and growth rate at the weaning and post weaning time

points were obtained, the impact of growth on muscle fibre type and meat quality traits

could be investigated.

In the first experiment growth curves were fitted to live weight data for 18,185 lambs

with weight data between birth and 300 days. Linear (cubic polynomial), nonlinear

(Brody) and random effects models were fitted and the characteristics of the residual

values examined to determine the best curve fit. The random effects model was

hypothesised to have the lowest residual values as it is based on a population, rather

than an individual fit, and is less biased when fitting growth curves to weight recordings

from animals with different durations of data collection. However, the cubic polynomial

fit had the lowest average residual values and the Brody fit had the highest. The

difference between the residual values for the fit of each function was relatively small

therefore the decision on which function to use was based on a variety of additional

IV

factors. As it is a population based fit, a random effects model accommodates missing

weight recordings and different durations of data collection, which are common with

industry data recording, without biasing the fit. A random effects model also allows

environmental effects specific to the time of recording to be accounted for and can

accommodate genetic differences in the shape of each animal’s growth curve. Based on

these characteristics the estimates of lamb weights and growth rates from the random

effects model were used in this thesis.

In the second experiment the impact of production factors on lamb weight at key time

points in production were analysed. While several studies have published information

on these effects the studies were generally small. Therefore, this analysis acted as a

benchmarking for the magnitude of the effects, based on large data. The full expression

of the post weaning weight breeding value is depressed with poor nutrition so an

analysis was performed to determine the extent to which other forms of nutritional

restriction, including lamb and dam factors such as birth type, dam breed and dam age,

moderate the expression of the growth breeding value. As expected, factors such as

multiple births, which act as a source of nutritional restriction, reduced the full

expression of the growth breeding value and hindered lambs reaching expected weights.

Muscle oxidative capacity is intrinsically linked to lamb growth and meat quality so in

the third experiment the influence of genetic selection for growth on muscle oxidative

capacity was examined. Oxidative capacity was indicated by the activity of the

oxidative enzyme isocitrate dehydrogenase as well as the concentration of myoglobin in

the muscle. As expected, genetic selection for growth reduced isocitrate dehydrogenase

activity and myoglobin concentration. The reduction in oxidative capacity with

increased growth was due to lambs tending toward a larger matures size and therefore

being physiologically less mature at a given weight

V

The impact of lamb growth on oxidative capacity and meat quality has previously been

estimated by the impact of the sire breeding value for growth on these traits or by using

hot carcass weight as a proxy for whole of life growth. However the accuracy of these

measures of growth is untested. The lamb weights and growth rates derived in the

second experiment were used to determine the real impact of growth on oxidative

capacity and meat quality. The impact of genetic selection on growth, or the use of hot

carcass weight as a proxy for the impact of growth on meat traits, were both shown to

be inadequate in quantifying the magnitude of the effect of growth on meat quality traits

and indicating how the effect of growth varied with time.

In the fourth experiment the impact of lamb growth on intramuscular fat, shear force

and consumer overall liking was examined. Intramuscular fat was increased and shear

force was reduced with lamb growth. Intramuscular fat and consumer overall liking

have a positive association however contrary to expectations growth had no impact on

overall liking. This is likely to be due to influences on overall liking other than

intramuscular fat.

In the final experiment the impact of lamb growth in isocitrate dehydrogenase activity,

myoglobin concentration, iron and zinc concentration and meat browning were

examined. As hypothesised, myoglobin, iron and zinc concentrations increased with

lamb growth. Contrary to expectations isocitrate dehydrogenase activity and meat

browning were reduced with lamb growth. Meat browning was thought to reduce due to

its strong association with oxidative capacity and thus the oxidative indicator enzyme

inverse isocitrate dehydrogenase.

These experiments have contributed to the sheep industry at a highly practical and

adoptable level showing why the full potential of genetic selection for growth may not

be realised and demonstrating the impact of growth on meat eating quality. Currently,

VI

increased growth in lambs, particularly in the early stages between birth and weaning

has positive effects on meat quality however lambs had few weights recorded between

these times so work needs to be done to characterise these associations.

VII

Acknowledgments

This journey has been filled with surprises, the greatest of which was meeting my

husband Richard. I did not know it was possible to be this happy. I am grateful for

meeting Meghan Cornelius when I started my thesis at Murdoch and for her introducing

the vegetarian city girl to the sheep farming country boy. We didn’t know this would

happen either, quite a lovely surprise.

I have had two great academic supervisors during my thesis, Associate Professor

Graham Gardner and Professor David Pethick. Their largest contribution has been to

teach me how to design every experiment with practical outcomes in mind for industry

adoption. They have provided both encouragement and opportunity to deliver these

outcomes in industry and scientific forums, allowing diverse and practical feedback.

These skills will set me in good stead as I continue my career.

I am especially grateful to Graham for his attention to detail and willingness to debate

the statistical nuances so integral to a thesis based on modelling. Our incongruous

opinions have largely been limited to less serious issues, such as colour schemes for

presentations and the variation in our desired amounts of personal space.

Dave has been incredibly helpful in seeing the broader picture, distilling practical

messages from a range of complex results. It was a difficult couple of years avoiding

telling Dave I’m actually a vegetarian, although he took the news gracefully.

The research would not have been possible without the support of the Sheep CRC in

terms of both funding and support. The annual Postgraduate Conference run by the

Sheep CRC has been a highlight of my candidature providing specific feedback from

industry experts which would not have been available in any other setting. Murdoch

VIII

University has also provided welcome support and it has been great to return home to

my alma mater. I cannot thank the Murdoch University Postgraduate Student

Association enough for their help in supporting all Murdoch postgraduate students with

in-house social events and conferences, and myself personally with travel and

completion scholarships.

The technical support provided by Mal Boyce, Ken Chong, Janice Stigwood, Rini

Margwani and Di Pethick have helped keep the whole project running smoothly. The

lab work would not have been possible on my own and I am so grateful for your help.

The statistical support from Clair Alton has given me further confidence in our analysis

and introduced me to the world of Bayes. With your help I am no longer bound only to

P values.

I have particularly loved the office I have shared with Fiona Anderson, Andrew

Williams, Linda Dawson and Sarah Stewart. Fiona has been a reassuring counterbalance

to the joys and frustrations of research as she underwent her own thesis journey and

Willo revolutionised our capacity for analysis with the automatic regression macro.

In addition to office friends it has been great to work with a large and supportive meat

science team of postgrads and postdocs. I am thankful to have also shared this journey

with Sarah Blumer, Sandra Corbett, Sarah Bonny, Peter McGilchrist, Liselotte Pannier,

Honor Calnan, who has always been ready to brunch, and Cam Jose who first

introduced me to Graham.

I am grateful to my family for their constant support. They mention that I’ve been

studying for a while. I know.

IX

Publications

Journal publications:

Kelman, K.R., Pannier, L., Pethick, D.W. and Gardner, G.E. (2014). Selection for lean

meat yield in lambs reduces indicators of oxidative metabolism in the longissimus

muscle. Meat Science. 96: 1058-1067.

Journal publications to be submitted:

Kelman, K.R., Alston, C., Pethick, D. W. and Gardner, G. E. (2016). Lamb birth and

rearing type influence growth and weight potential of progeny from high growth sires.

Kelman, K.R., Pannier, L., Pethick, D. W. and Gardner, G. E. (2016). Lamb growth

increases intramuscular fat levels and reduces shear force in the longissimus muscle.

Kelman, K.R., Pannier, L., Calnan, H., Pethick, D. W. and Gardner, G. E. (2016).

Lamb growth increases oxidative capacity and mineral content in the longissimus

muscle.

Kelman, K.R., Alston, C., Pethick, D. W. and Gardner, G. E. (2016). Comparison of

linear and nonlinear functions to estimate lamb growth.

Conference proceedings:

de Hollander, C.A, Moghaddar, N., Kelman, K.R., Gardner, G.E., van der Werf, J.

(2014) Is variation in growth trajectories genetically correlated with meat quality traits

in Australian terminal lambs? In ‘Proceedings of the 10th

World Congress on Genetics

Applied to Livestock Production’, Vancouver, Canada.

Kelman, K.R., Alston, C.L., Pethick, D.W. and Gardner, G.E. (2013) Multiple births

limit the advantage of using high growth sires – Dealing with large data sets. In

‘Proceedings of the 64th Annual Meeting of the European Federation of Animal

Production’. Nantes, France. Book of Abstracts No.19. p 620. (Wageningen Academic

Publishers).

X

Kelman, K.R., Alston, C.L., Pethick, D.W. and Gardner, G.E. (2013) Multiple births

limit the advantage of using high growth sires. In ‘Proceedings of the 59th

International

Congress of Meat Science and Technology’, Izmir, Turkey. (Eds M Serdaroglu, B

Oxturk, T Akcan.) Proceedings S3-2. (Ege University, Food Engineering

Development).

Kelman, K.R., Pethick, D.W. and Gardner, G.E. (2012) Multiple births and young

ewes limit the advantage of using high PWWT sires. In ‘Proceedings of Lambex 2012’.

Bendigo, Victoria. Proceedings p 38.

Kelman, K.R., Pannier, L., Pethick, D.W. and Gardner, G.E. (2012) Selection for

leanness decreases meat aerobicity. In ‘Proceedings of Sheep CRC Postgraduate

Student Conference’. Coffs Harbour, New South Wales.

Gardner, G.E., Anderson, F., Williams, A., Kelman, K.R., Pannier, L. and Pethick,

D.W. (2012) Growth breeding value redistributes weight to the saddle region of lamb

carcasses. In ‘Proceedings of the 63rd Annual Meeting of the European Federation of

Animal Production’, Bratislava, Slovakia. Book of Abstracts No.18 p 302. (Wageningen

Academic Publishers).

Kelman, K.R., Pethick, D.W. and Gardner, G.E. (2011) Selection for leanness

decreases isocitrate dehydrogenase activity. In ‘Proceedings of Sheep CRC

Postgraduate Student Conference’. Coffs Harbour, New South Wales.

Kelman, K.R., Pannier, L., Pethick, D.W. and Gardner, G.E. (2011) Selection for

reduced PFAT decreases Isocitrate Dehydrogenase activity. In ‘Proceedings of the 62rd

Annual Meeting of the European Federation of Animal Production’ Stavanger, Norway.

Book of Abstracts No.17, p 299. (Wageningen Academic Publishers).

Gardner, G.E., Pannier, L., Anderson, F., Kelman, K.R., Williams, A., Jacob, R.H.,

Ball, A.J. and Pethick, D.W. (2011) The impact of selection for leanness on lamb

carcase composition, intramuscular fat, and muscle metabolic type, is not influenced by

nutritional variation within Australian production systems. In ‘Proceedings of the 8th

International Symposium on the Nutrition of Herbivores Advances in Animal

Biosciences’. Aberystwyth, Wales. p 406.

XI

Kelman, K.R., Pethick, D.W. and Gardner, G.E. (2010) Determining lamb growth rate

during development. In ‘Proceedings of Cooperative Research Centre for Sheep

Industry Innovation Conference’. Adelaide, South Australia.

Kelman, K.R., Pethick, D.W. and Gardner, G.E. (2010) The impact of lamb growth

rate on Information Nucleus Flock slaughter traits. In ‘Proceedings of Sheep CRC

Postgraduate Student Conference’. Coffs Harbour, New South Wales.

XII

Abbreviations

AMSA American Meat Science Association

ASBV Australian sheep breeding value

ATP Adenosine triphosphate

CoA Co-enzyme A

HCWT Hot carcass weight

ICDH Isocitrate dehydrogenase

MSA Meat Standards Australia

NADH Nicotinamide adenine dinucleotide

NIR Near infrared

TAG Triacylglycerol

TCA Citric acid cycle

XIII

Enzymes

ATPase EC 3.6.1.3

Citrate synthase EC 2.3.3.1

Cytochrome-c oxidase EC 1.9.3.1

Glycogen phosphorylase EC 2.4.1.1

Glycogen synthase EC 2.4.1.11

Hexokinase EC 2.7.1.1

Hydroxyacyl-CoA dehydrogenase EC 1.1.1.35

Isocitrate dehydrogenase EC 1.1.1.41

Lactate dehydrogenase EC 1.1.1.27

Phosphofructokinase EC 2.7.1.11

XIV

XV

Table of contents

Declaration ........................................................................................................................................... II

Summary .............................................................................................................................................. III

Acknowledgments .............................................................................................................................. VII

Publications ......................................................................................................................................... IX

Abbreviations...................................................................................................................................... XII

Enzymes ............................................................................................................................................. XIII

Table of contents ................................................................................................................................ XV

List of figures...................................................................................................................................... XIX

List of tables ....................................................................................................................................... XXI

Chapter 1. Introduction ................................................................................................................... 1

Chapter 2. Literature Review ........................................................................................................... 5

2.1 Animal growth ................................................................................................................................... 5

2.2 Growth curves .................................................................................................................................... 6 2.2.1 Linear functions ......................................................................................................................... 7 2.2.2 Nonlinear functions ................................................................................................................... 8 2.2.3 Random Effects Model ....................................................................................................... 12 2.2.4 Residual weight ........................................................................................................................ 13

2.3 Growth and tissue differentiation ................................................................................................... 14

2.4 Production factors influencing growth ............................................................................................ 16 2.4.1 Sex ............................................................................................................................................ 16 2.4.2 Genetics ................................................................................................................................... 17 2.4.3 Nutritional restriction .............................................................................................................. 22 2.4.4 Birth type and rear type........................................................................................................... 24 2.4.5 Dam factors.............................................................................................................................. 24

2.5 Growth and fibre type ..................................................................................................................... 27 2.5.1 Muscle fibre classification ....................................................................................................... 27 2.5.2 Oxidative fibres ........................................................................................................................ 30 2.5.3 Glycolytic fibres ....................................................................................................................... 30 2.5.4 Muscle fibre development ....................................................................................................... 30 2.5.5 Production factors impacting fibre type .................................................................................. 32 2.5.6 Fibre type and meat quality ..................................................................................................... 35

2.6 Growth and meat quality ................................................................................................................ 35 2.6.1 Intramuscular fat ..................................................................................................................... 36 2.6.2 Shear force ............................................................................................................................... 43 2.6.3 Overall liking ............................................................................................................................ 47 2.6.4 Isocitrate dehydrogenase activity ............................................................................................ 50 2.6.5 Myoglobin concentration ........................................................................................................ 52 2.6.6 Iron and zinc concentration ..................................................................................................... 57 2.6.7 Meat browning ........................................................................................................................ 61

2.7 Hypothesis .................................................................................................................................... 65

XVI

Chapter 3. Comparison of linear and nonlinear functions to estimate lamb growth ...................... 69

3.1 Introduction ......................................................................................................................................69

3.2 Materials and Methods ....................................................................................................................71 3.2.1 Experimental Design ................................................................................................................71

3.3 Statistical Analysis ............................................................................................................................72 3.3.1 Data available ...........................................................................................................................72 3.3.2 Nonlinear curve fit....................................................................................................................73 3.3.3 Linear curve fit..........................................................................................................................74 3.3.4 Random effects model fit using Bayesian analysis ...................................................................74 3.3.5 Curve fitting comparisons ........................................................................................................74

3.4 Results ..............................................................................................................................................76 3.4.1 Nonlinear curve fit....................................................................................................................76 3.4.2 Linear curve fit..........................................................................................................................79 3.4.3 Bayesian random effects model fit ..........................................................................................81 3.4.4 Comparison of residual graphs over time ................................................................................83

3.5 Discussion .........................................................................................................................................84 3.5.1 Mean square error ...................................................................................................................84 3.5.2 Curve attributes .......................................................................................................................84 3.5.3 Residual weights .......................................................................................................................86 3.5.4 Random effects model .............................................................................................................87

3.6 Conclusion ........................................................................................................................................88

Chapter 4. Lamb birth and rearing type influence growth and weight potential of progeny from high growth sires ................................................................................................................................. 89

4.1 Introduction ......................................................................................................................................89

4.2 Materials and Methods ....................................................................................................................91 4.2.1 Experimental Design ................................................................................................................91 4.2.2 Lamb weights ...........................................................................................................................92 4.2.3 Data available ...........................................................................................................................92 4.2.4 Statistical analysis.....................................................................................................................93

4.3 Results ..............................................................................................................................................96 4.3.1 Actual birth weights .................................................................................................................96 4.3.2 Estimated live weights and growth rates at 100, 150 and 240 days ..................................... 103

4.4 Discussion ...................................................................................................................................... 111 4.4.1 Association between carcass breeding values and lamb weight and growth rate ............... 111 4.4.2 Production effects ................................................................................................................. 114

4.5 Conclusion ..................................................................................................................................... 120

Chapter 5. Selection for lean meat yield in lambs reduces indicators of oxidative metabolism in the longissimus muscle ............................................................................................................................ 121

5.1 Introduction ................................................................................................................................... 121

5.2 Materials and Methods ................................................................................................................. 123 5.2.1 Experimental Design ............................................................................................................. 123 5.2.2 Collection of phenotypic measurements .............................................................................. 124 5.2.3 Data available ........................................................................................................................ 125 5.2.4 Statistical analysis.................................................................................................................. 125

5.3 Results ........................................................................................................................................... 128 5.3.1 Isocitrate Dehydrogenase ..................................................................................................... 128 5.3.2 Myoglobin Concentration ..................................................................................................... 137

5.4 Discussion ...................................................................................................................................... 147 5.4.1 Association between carcass breeding values and oxidative capacity ................................. 147

XVII

5.4.2 Associations with carcass phenotypic measurements .......................................................... 149 5.4.3 Production effects ................................................................................................................. 151

5.5 Conclusion ...................................................................................................................................... 154

Chapter 6. Lamb growth increases intramuscular fat levels and reduces shear force in the longissimus muscle ............................................................................................................................ 155

6.1 Introduction ................................................................................................................................... 155

6.2 Materials and Methods ................................................................................................................. 156 6.2.1 Experimental Design .............................................................................................................. 156 6.2.2 Collection of phenotypic measurements ............................................................................... 157 6.2.3 Data available ........................................................................................................................ 158 6.2.4 Statistical analysis .................................................................................................................. 161

6.3 Results ........................................................................................................................................... 165 6.3.1 Intramuscular fat ................................................................................................................... 165 6.3.2 Shear force ............................................................................................................................. 167 6.3.3 Overall liking .......................................................................................................................... 170 6.3.4 Phenotypic correlations ......................................................................................................... 170

6.4 Discussion ...................................................................................................................................... 172 6.4.1 Intramuscular fat ................................................................................................................... 172 6.4.2 Shear force ............................................................................................................................. 174 6.4.3 Overall liking .......................................................................................................................... 176

6.5 Conclusion ...................................................................................................................................... 178

Chapter 7. Lamb growth increases oxidative capacity and mineral content in the longissimus muscle..............................................................................................................................................179

7.1 Introduction ................................................................................................................................... 179

7.2 Materials and Methods ................................................................................................................. 181 7.2.1 Experimental Design .............................................................................................................. 181 7.2.2 Collection of phenotypic measurements ............................................................................... 181 7.2.3 Data available ........................................................................................................................ 183 7.2.4 Statistical analysis .................................................................................................................. 186

7.3 Results ........................................................................................................................................... 188 7.3.1 Isocitrate Dehydrogenase ...................................................................................................... 188 7.3.2 Myoglobin Concentration ...................................................................................................... 192 7.3.3 Iron Concentration................................................................................................................. 195 7.3.4 Zinc Concentration ................................................................................................................. 197 7.3.5 Meat Browning ...................................................................................................................... 199 7.3.6 Phenotypic correlations ......................................................................................................... 200

7.4 Discussion ...................................................................................................................................... 202 7.4.1 Isocitrate dehydrogenase activity .......................................................................................... 202 7.4.2 Myoglobin concentration ...................................................................................................... 204 7.4.3 Iron and zinc concentration ................................................................................................... 206 7.4.4 Meat browning ...................................................................................................................... 208

7.5 Conclusion ...................................................................................................................................... 211

Chapter 8. General Discussion ...................................................................................................... 213

8.1 Benefits of lamb growth to industry .............................................................................................. 213

8.2 Brody function ............................................................................................................................... 214

8.3 Cubic polynomial function ............................................................................................................. 215

8.4 Random effects model ................................................................................................................... 216

8.5 Impact of environment on growth ................................................................................................. 218

XVIII

8.6 Impact of genetics on growth ....................................................................................................... 220

8.7 Fibre type and oxidative capacity ................................................................................................. 222

8.8 Impact of genetics on oxidative capacity ...................................................................................... 223

8.9 Impact of growth on oxidative capacity ........................................................................................ 224

8.10 Impact of growth on objective and subjective measures of eating quality ................................. 225

8.11 Impact of growth on mineral content of meat ............................................................................ 226

8.12 Impact of growth on meat browning .......................................................................................... 228

8.13 Impact of growth on meat traits ................................................................................................. 229

8.14 Adjustments to experimental design........................................................................................... 230

8.15 Future studies .............................................................................................................................. 232

References ......................................................................................................................................... 234

XIX

List of figures

Figure 2-1: A growth curve demonstrating the phases of growth with time: lag, exponential growth,

transitional diminishing growth and the stationary plateau at maturity ........................................................ 5

Figure 2-2 Graph of a linear (···) and cubic polynomial function (—) fitted to lamb live weight data (o) in

kilograms against time in days. .................................................................................................................... 7

Figure 2-3 A typical graphical representation of lamb live weight data with a Brody function fitted to

lamb live weight data (o) in kilograms against time in days. ..................................................................... 11

Figure 2-4 Change in carcass fat, muscle and bone as a portion of mature live weight during allometric

growth: Source (Butterfield et al., 1984) ................................................................................................... 14

Figure 2-5 Interpretation of an Australian Sheep Breeding Value (ASBV) for post weaning weight based

on value, trait, age and accuracy. Source: Sheep Genetics ......................................................................... 19

Figure 2-6 Selection index for rams (circled) based on Australian Sheep Breeding Value for weight

(PWT), muscle (PEMD) and fat (PFAT). Source: Sheep Genetics ............................................................ 20

Figure 2-7 Interconversions of myoglobin redox forms in fresh meat. Source: AMSA Color Measurement

Guidelines (AMSA, 2012) ......................................................................................................................... 54

Figure 3-1 Graph of the number of weight recordings against lamb age (in days) between 1 and 300 days.

.................................................................................................................................................................... 73

Figure 3-2 Graph of average residual weight (kg) per 10 day period with a Brody curve fit (-- -- -- --), a

cubic curve fit (- - - -) and a random effects model fit (····). The solid line at residual weight zero

represents the line the residuals would fall on if the random effects model were a perfect fit to the weight

data. ............................................................................................................................................................ 77

Figure 3-3 Histogram of the distribution of residual weight values (kg) against the percent frequency for

the Brody function fit. ................................................................................................................................ 79

Figure 3-4 Histogram of the distribution of residual weight values (kg) against the percent frequency for

the cubic function fit. ................................................................................................................................. 81

Figure 3-5 Histogram of the residual weight values (kg) against the percent frequency for the random

effects model fit.......................................................................................................................................... 83

Figure 4-1 Comparison of sire estimates for post weaning weight (PWWT) Australian Sheep Breeding

Values for Maternal, Merino and Terminal sires with predicted means for weight at 150 days. Lines

represent predicted means for weight (± s.e.) for each sire type. Symbols (∆,+,o) represent individual sire

estimates. .................................................................................................................................................. 107

XX

Figure 4-2 Comparison of sire estimates for post weaning fat depth (PFAT) Australian Sheep Breeding

Values for Maternal, Merino and Terminal sires with predicted means for weight at 150 days. Lines

represent predicted means for weight (± s.e.) for each sire type. Symbols (∆,+,o) represent individual sire

estimates. .................................................................................................................................................. 110

Figure 5-1 Linear relationship between muscle sample protein concentration (mg/g tissue) and isocitrate

dehydrogenase activity (± s.e.) in μmol/min/g tissue per gramme of protein. Each icon (o) represents an

individual lamb. ........................................................................................................................................ 127

Figure 5-2 The effect of site by year of birth on isocitrate dehydrogenase activity (μmol/min/g tissue) (±

s.e.) from the base model. ........................................................................................................................ 131

Figure 5-3 Linear relationship between isocitrate dehydrogenase activity (μmol/min/g tissue) (± s.e.) and

slaughter age in days. Symbols represent predicted means for kill groups. Lines represent predicted means

for age at slaughter (± s.e.) from base model with no kill group term. ..................................................... 132

Figure 5-4 Curvilinear relationship between intramuscular fat percentage and isocitrate dehydrogenase

activity (± s.e.) in μmol/min/g tissue. Symbols (o) represent individual lamb residuals from the

intramuscular fat model. ........................................................................................................................... 134

Figure 5-5 Relationship between sire estimates for isocitrate dehydrogenase activity (μmol/min/g tissue)

and subcutaneous post weaning c-site fat depth (PFAT) for Maternal, Merino and Terminal sires. Lines

represent predicted means for isocitrate dehydrogenase activity (± s.e.) for all sire types combined.

Symbols represent individual sire estimates. ............................................................................................ 135

Figure 5-6 The effect of site by year of birth on myoglobin concentration (mg/g tissue) (± s.e.) from the

base model. ............................................................................................................................................... 140

Figure 5-7 Linear relationship between myoglobin concentration (mg/g tissue) and slaughter age in days.

Symbols represent predicted means for kill groups. Lines represent predicted means for age at slaughter

(± s.e.) from base model with no kill group term. .................................................................................... 141

Figure 5-8 Curvilinear relationship between intramuscular fat percentage and myoglobin concentration (±

s.e.) in mg/g tissue. Symbols (o) represent individual lamb residuals from the intramuscular fat model. 143

Figure 5-9 Relationship between sire estimates for myoglobin concentration (mg/g tissue) and post

weaning eye muscle depth (PEMD) for Maternal, Merino and Terminal sires. Lines represent predicted

means for myoglobin concentration (± s.e.) for each sire type. Symbols (o) represent individual sire

estimates. .................................................................................................................................................. 146

XXI

List of tables

Table 2-1 Contractile and metabolic characteristics of individual fibre types. Adapted from Lefaucheur

(2010). ........................................................................................................................................................ 29

Table 3-1 Mean square error (MSE) ± standard error (kg2) for curve fits between 0 and 300 days and 90

and 160 days for Brody, Cubic, Random Effects Model and Random Effects Model with all data

(including animals with less than 4 live weight observations) curve fits. .................................................. 76

Table 3-2 Results of tests for the location of mu, normality, skewedness and kurtosis of the average

residual values from a Brody function fit to 17,377 animals. .................................................................... 78

Table 3-3 Results of test for the location of mu, skewedness, kurtosis and normality of the average

residual values from a cubic function fit to 17,382 animals. ...................................................................... 81

Table 3-4 Results of tests for the location of mu, skewedness, kurtosis and normality of the average

residual values from a random effects model fit to 17,382 animals. .......................................................... 83

Table 4-1 Actual birth weight (kg) ranges and estimated weight (kg) and growth rate (g/day) ranges at

100, 150 and 240 days. ............................................................................................................................... 92

Table 4-2 Phenotypic correlations (standard errors) of post weaning sire breeding values for weight, fat

depth and eye muscle depth among Maternal, Merino and Terminal sires. ............................................... 93

Table 4-3 Numerator and denominator degrees of freedom and F-values for the base models for birth

weight and estimated weights and growth rates at 100, 150 and 240 days. ............................................... 97

Table 4-4 Predicted means for birth weight (kg) and estimated weight at 100, 150 and 240 days for

different sexes, birth type-rear types, dam ages, sire types and dam breed within sire type combinations.99

Table 4-5 Magnitude of effect (difference between the highest and lowest values) for production or for

genetic effects on weight and growth rate. ............................................................................................... 100

Table 4-6 95% Confidence intervals of sire estimates for birth weight (kg) and estimated weights (kg) and

growth rates (g/day) at 100, 150 and 240 days. ........................................................................................ 101

Table 4-7 Regression coefficients for the relationship between sire ASBVs for growth and actual birth

weight (kg) or estimated weight and growth rate at 100, 150 and 240 days. ........................................... 102

Table 4-8 Predicted means for estimated growth rate at 100, 150 and 240 days for different sexes, birth

type-rear types, dam ages, sire types and dam breed within sire type combinations. .............................. 104

Table 4-9 Regression coefficients for the relationship between sire ASBVs for muscling and leanness and

estimated weight and growth rate at 100, 150 and 240 days. ................................................................... 109

XXII

Table 5-1 Numerator degrees of freedom, denominator degrees of freedom and F-values for the base

models for isocitrate dehydrogenase activity (μmol/min/g tissue) and myoglobin concentration (mg/g

tissue). ....................................................................................................................................................... 128

Table 5-2 Number of progeny at each site for year of birth, sex, sire type and dam breed within sire type

from the base ICDH activity model. ......................................................................................................... 129

Table 5-3 Raw mean ± standard deviation, and minimum to maximum value, for isocitrate dehydrogenase

activity and the phenotypic covariates tested within the base model. ....................................................... 130

Table 5-4 Number of sires and Australian Sheep Breeding Values mean (min, max) for each sire type

from the isocitrate dehydrogenase (ICDH) activity and myoglobin concentration data. .......................... 136

Table 5-5 Number of progeny at each site for year of birth, sex and sire type from the base myoglobin

concentration model.................................................................................................................................. 138

Table 5-6 Raw mean ± standard deviation and minimum to maximum values for myoglobin concentration

and the phenotypic covariates tested within the base model. .................................................................... 139

Table 6-1 Weight and growth rate ranges for intramuscular fat, shear force and overall liking. .............. 160

Table 6-2 Number of animals, years of data and number of sires available for intramuscular fat, shear

force and overall liking. ............................................................................................................................ 161

Table 6-3 Phenotypic correlations (standard errors) of post weaning sire breeding values for weight, fat

depth and eye muscle depth among Maternal, Merino and Terminal sires. .............................................. 161

Table 6-4 The magnitude of effect (difference between the highest and lowest values) of lamb weight and

growth rate on intramuscular fat or on shear force across the entire weight or growth rate range. .......... 165

Table 6-5 Variation in slopes between sire types for the impact of birth weight, weight at 100 days and

weight at 150 days on intramuscular fat. .................................................................................................. 166

Table 6-6 Phenotypic correlations of intramuscular fat, shear force and overall liking with lamb weight

and growth rate. Standard error values in brackets. Significant correlations are in bold. ......................... 171

Table 7-1 Weight and growth rate ranges for isocitrate dehydrogenase (ICDH) activity, myoglobin

concentration, iron concentration, zinc concentration and meat browning. .............................................. 184

Table 7-2 Number of animals, years of data and number of sires available for isocitrate dehydrogenase

(ICDH) activity, myoglobin concentration, iron concentration, zinc concentration and meat browning. 185

Table 7-3 Phenotypic correlations (standard errors) of post weaning sire breeding values for weight, fat

depth and eye muscle depth among Maternal, Merino and Terminal sires. .............................................. 186

XXIII

Table 7-4 The magnitude of effect (difference between the highest and lowest values) of lamb weight and

growth rate on isocitrate dehydrogenase activity, myoglobin concentration, iron concentration, zinc

concentration and meat browning. ........................................................................................................... 190

Table 7-5 Variation in slopes between sire types for the impact of birth weight, weight at 100 and 150

days, weight change between 100 and 150 days and growth rate at 100 and 150 days on myoglobin

concentration, iron concentration and zinc concentration. ....................................................................... 193

Table 7-6 Phenotypic correlations of isocitrate dehydrogenase (ICDH) activity, myoglobin concentration,

iron concentration, zinc concentration and meat browning with lamb weight and growth rate. Standard

error values in brackets. Significant correlations are in bold. .................................................................. 201

1

Chapter 1. Introduction

Lamb growth is a key profit driver for producers in the Australian Sheep Industry,

particularly with the fall in wool prices causing a shift toward meat production (Curtis

et al., 2006). Lambs with fast growth rates will reach market weights earlier, resulting

in an overall lower cost of production per lamb and an increase in farm profitability.

Australians have among the highest sheep meat consumption in the world with each

person consuming around 10 kg of lamb per year (MLA, 2013). This resulted in a

domestic expenditure on lamb of around $2 billion in the 2012-2013 financial year

(MLA, 2013). As demand for protein in the form of red meat is increasing worldwide,

increasing efficient lamb production is desirable.

Productivity in lamb production can be increased by optimising nutrition, and genetic

selection for improvement is becoming more widely adopted. Historically, increased

growth in progeny was achieved by cross breeding, or selecting animals with higher

weaning, post weaning or hogget weights. This caused some improvement due to the

heritability of weight (Safari and Fogarty, 2003). However, in recent times genetic

evaluation systems such as LAMBPLAN™ and MERINOSELECT™ have been

developed by Sheep Genetics to provide an objective measurement of important

production traits, including lamb growth. Producers are able to use the genetic estimates

to select sires with the desirable trait characteristic and based on the accuracies provided

have confidence that the progeny of that sire will inherit the benefits of that trait. More

recently, genomic testing has become available, which further aids producers when

selecting prospective sires.

2

However, selection for growth has raised a number of questions concerning its broader

impact on the physiology of the animal including the impact of selection for increased

growth on muscle fibre type. This is important as muscle fibre type is intrinsically

linked with both subjective and objective measures of meat quality (Calkins et al., 1981;

Calnan et al., 2014; Hocquette et al., 2010; Pearce et al., 2009). Subjective measures of

eating quality include customer rated tenderness, flavour, juiciness and overall liking

(Polkinghorne et al., 1999) while objective measures include intramuscular fat, shear

force, meat colour and mineral content. Therefore any impact of growth on muscle fibre

type could have correlated impacts on meat quality and consumer acceptability. The

meat quality traits of interests selected at the outset of this study were indicators of

eating quality (intramuscular fat, shear force and consumer overall liking), indicators of

meat nutrition (iron and zinc concentration) and indicators of retail shelf life (rate of

meat browning).

Production factors include the site of lamb rearing, year of birth, sex, age, dam breed,

dam age and sire type. These factors have been shown in a number of studies to impact

growth, however the results vary in magnitude due to small sample sizes. Therefore a

study needs to be undertaken to quantify the magnitude of the impact of production

factors on lamb growth, based on a large data set.

In addition to production factors influencing growth, genetic selection for growth using

Australian Sheep Breeding Values (ASBVs) also influence growth with lamb weight

increasing with selection for an increased post weaning weight (PWWT). However the

full impact is reduced by suboptimal ewe nutrition (Hegarty et al., 2006c). As

production factors such as dam breed, dam age and lamb birth type (single or multiple)

can act as a source of nutritional restriction, the impact of PWWT may vary with these

3

production factors. This is important to investigate so producers can temper their

expectations of lamb weights based on production factors when using ASBVs.

Impacts of genetic selection for increased growth using ASBVs on measures of meat

quality have been published. Selection for increased growth via an increased PWWT

has no impact on intramuscular fat (Pannier et al., 2014c), consumer overall liking

(Pannier et al., 2014a), iron and zinc concentration (Pannier et al., 2014b) or rate of

meat browning (Calnan et al., 2014). While these meat quality traits not impacted by

selection for increased growth via the PWWT ASBV, what remains to be investigated is

the impact of PWWT on indicators of oxidative capacity and thus fibre type.

Meat Standards Australia (MSA) is an eating quality program for beef and sheep meat,

based on extensive consumer research (MSA, 2012). All sheep meat which falls under

the banner of the MSA program has to meet strict criteria including a minimum weight

gain of 100-150 g/day, depending on breed, for 2 weeks prior to consignment. Thus

animals need to be fed to potential with appropriate nutrition to reach these growth

rates. As nutrition is an effect of environment, and the magnitude of the impact of

environment on meat traits in the studies above was larger than the magnitude of the

impact of genetic selection, growth itself may impact on meat traits. Hot carcass weight

has been used as a proxy for whole of life growth rate to indicate the association

between meat traits and growth (Calnan et al., 2014; Pannier et al., 2014a; Pannier et

al., 2014b; Pannier et al., 2014c). However, how carcass weight is dependent on age at

slaughter and is unlikely to be an accurate measure of whole of life growth rate. Rather

weight at a given age becomes less representative of growth and more representative of

environmental influences as animals get older. Furthermore, lamb growth is sigmoidal

while the association between hot carcass weight and growth indicates a linear

4

relationship. Therefore, the hot carcass weight effect cannot illustrate how the

association between growth and meat traits may vary across time.

The impact of lamb growth on meat traits needs to be investigated at key time points in

production such as birth, weaning and post weaning to determine whether growth

impacts meat traits and these time points and to quantify the magnitude of the effect.

This approach will also demonstrate how the association varies with time and show the

window of opportunity for the traits to be manipulated through growth.

Estimation of lamb growth at birth, weaning and post weaning time points is essential to

the implementation of appropriate management and feeding strategies to allow lambs to

grow to their full potential, increasing farm profitability. To sustain long term

profitability the potential impact of this growth on muscle fibre type and meat quality

needs to carefully examined as consumers are increasingly demanding, and willing to

pay for, increased eating quality (Pethick et al., 2006a). As lamb growth impacts the

economic viability of a production system the focus of these experiments was at a

highly practical and adoptable level for industry.

The Australian Cooperative Research Centre (CRC) for Sheep Industry Innovation

established an Information Nucleus Flock (INF) in 2007. It produced approximately

2,000 lambs for slaughter each year over a 5 year period. Objectives of the INF include

the measurement of a diverse range of phenotypic traits, including lamb weight,

estimation of the heritability of traits and estimation of genetic correlations between

traits (Mortimer et al., 2014). Information pertaining to genetic factors, such as sure

type and sire breeding values, and non-genetic factors such as site of rearing, year of

birth, sex, birth type-rear type combination, dam age and kill group were recorded or

calculated for each animal. The analysis and findings presented in this thesis are based

on date collected during the INF project.

5

Chapter 2. Literature Review

2.1 Animal growth

Animal growth is characterised by a change in live weight with time as animals follow a

trajectory to a final or mature size (Brody, 1945; Parks, 1982). This growth trajectory is

sigmoidal in nature (Figure 2-1) with an initial lag phase followed by the greatest

growth rate immediately after birth in the exponential phase. Growth rate then declines

as the animal nears its mature weight in the transitional phase which is followed by a

stable mature weight in the plateau or stationary phase after which further growth is

retarded (Hendrick et al., 1989).

Figure 2-1: A growth curve demonstrating the phases of growth with time: lag,

exponential growth, transitional diminishing growth and the stationary plateau at

maturity

Time

Wei

gh

t

6

Weight change measured over time results in longitudinal time series data.

Longitudinal data can be difficult to analyse, particularly if not all animals are weighed

at the time of interest. One option is to discard the weight information from animals not

weighed at the time of interest (Fitzhugh, 1976). However, incomplete weight data

recordings are common in industry therefore utilisation of the data that is available is

essential.

To overcome this, linear and nonlinear mathematical functions have been used to

estimate this sigmoidal shape based on available weight recordings. This results in the

ability to estimate weight for animals at times of interest, regardless of whether they

have been weighed at this time.

2.2 Growth curves

In an industry setting, animals are rarely weighed on a daily basis. In addition to the

labour and equipment required, weighing animals close to birth may increase the risk of

mis-mothering. Australian production systems are based on extensive grazing where

animals cannot practically be herded for weighing on a daily basis. Technologies such

as electronic identification and walk over weighing provide opportunities for the

collection of real time live weight measurements, however these technologies are still

being refined for use in extensive grazing systems (Brown et al., 2015; Scott, 2014).

A practical solution is the fitting of a growth function to available live weight data so

that weight can be estimated on a given day or across a time period of interest. This also

7

allows for the estimation of instantaneous growth rates which give an indication of how

the shape of the growth curve is changing.

Growth functions can be fitted on an individual or population basis using linear or

nonlinear functions. Each methodology has advantages and limitations, as outlined

below in Sections 2.2.1, 2.2.2 and 2.2.3.

2.2.1 Linear functions

A linear function is fitted on an individual animal basis and describes a simple

relationship where weight is a function of time only. This is true for polynomial

functions. Thus, linear, quadratic, cubic and quartic polynomials are all linear functions.

An example of first and third order polynomial functions fitted to lamb live weight data

is shown in Figure 2-2.

Figure 2-2 Graph of a linear (···) and cubic polynomial function (—) fitted to lamb live weight data (o) in

kilograms against time in days.

0

5

10

15

20

25

30

35

40

45

50

0 50 100 150 200 250 300 350

Wei

gh

t (k

g)

Age (days)

8

Initially, all animals have simple linear growth. However lambs often experience a

setback immediately after weaning (Atkins and Thompson, 1979) resulting in weight

loss, which is difficult to capture and quantify with a linear function (Emmans, 1997).

This results in an general lack of ability to describe growth over the entire range of life

from conception to adulthood (Taylor, 1978).

Linear functions are useful for the ability fit live weight data with ease, however there

are notable limitations to their use. When a linear function is used to model growth, the

curve parameters do not have biological interpretations, limiting the usefulness of the

information. Some linear functions also have unrealistic asymptotes, where weight

approaches but never reaches an extremely high or low weight value. For a cubic

polynomial, this can result in extreme overestimation of weight close to adulthood. This

effect is minimised when the function is fitted over a narrow range of weights where the

function has good fitting characteristics. For these reasons, individually fitted linear

functions are not used as commonly as nonlinear functions to model growth.

2.2.2 Nonlinear functions

Nonlinear functions are fitted on an individual basis and describe a relationship where

weight is a function of several parameters, one of which is time. Other parameters can

include birth weight, mature weight and growth rate, depending on the nonlinear

function used. Nonlinear functions are favoured as their parameters have simple

biological interpretations.

A number of nonlinear models are available with Parks (1982) demonstrating ten

nonlinear models that have been used to describe size-age relationships in animals. The

most commonly used functions are the Brody (1945), von Bertalanffy (1957), Richards

(1959), Logistic (Nelder, 1961), and Gompertz (Laird, 1965). These are based on

9

exponential functions, are sigmoidal in nature and reach an asymptotic value, equivalent

to an adult or mature weight, with time.

When the fit of nonlinear functions are compared, there is generally no one function

which fits best in all circumstances. Rather, the function that best fits live weight data

varies with both species and breed, as seen in Sections 2.2.2.1 and 2.2.2.2.

2.2.2.1 Nonlinear functions in sheep

The Gompertz function is empirically used to model growth in sheep in some studies,

without comparison with other functions. This is demonstrated in the Friggens et al.

(1997), Zygoyiannis et al. (1997) and Lewis et al. (2002) experiments.

However, each of the five functions mentioned in Section 2.2.2 has been assessed as the

best fit for sheep live weight data in at least one comparative study. The Brody function

demonstrated the best fit in the Gbangboche et al. (2008) experiment on West African

Dwarf sheep while the Logistic function had a superior fit for Brazilian sheep breeds in

both the McManus et al. (2003) and Malhado et al. (2009) experiments.

Richards and Gompertz functions were considered to have equally good curve fits for

Texel and Scottish Blackface lambs in the Lambe et al. (2006) experiment while in the

Topal et al. (2004) experiment the Gompertz function resulted in the best fit for the

Morkaraman breed, while the von Bertalanffy function was the best for the Awassi

breed.

The lack of a single nonlinear function which best fits live weight data is not unique to

sheep, with similar findings in cattle. With a lack of continuity in the best fitting

nonlinear function to use, factors such as ease of computation and proportions of

convergence become important considerations.

10

2.2.2.2 Nonlinear functions in cattle

A study in cattle showed that for two varying breeds, purebred Hereford and a

composite of Charolais, Angus and Galloway, the Richards function provided the best

overall fit to both sets of data (Goonewardene et al., 1981). The Richards function also

demonstrated the best fit in the Brown et al. (1976) and Denise and Brinks (1985)

experiments although it was noted in both experiments that the Brody function fit was

faster, computationally easier and resulted in similar estimated weights to the Richards

function. When Malhado et al. (2009) used the Richards function in sheep it had

difficulty converging limiting the usefulness of the function.

The Richards function, which is based on plant growth, is highly flexible however it has

no fixed inflection point which hinders the iterative process for the adjustment in the

function. Nonetheless, it has been used to describe growth in multiple species including

chickens (Gous et al., 1999; Hancock et al., 1995), pigs (Ferguson and Gous, 1993a,

1993b; Knap, 2000), quail (Akbas and Yaylak, 2000) and whales (Clark et al., 2000).

Despite the prevalence of the use of the Richards function it is not always the best

fitting function for live weight data with the Brody function (1945) having the best fit

for Angus cattle in the Beltrán et al. (1992) experiment and the von Bertalanffy function

(1957) function having the best fit in Retinta beef cattle in the López de Torre et al.

(1992) experiment.

2.2.2.3 Brody function

The Brody function (1945) has been used to model growth in sheep due to ease of

estimation, simplicity of interpretation (Bathaei and Leroy, 1996; Bathaei and Leroy,

1998; Topal et al., 2004) and its versatility (de Freitas, 2005).

11

Originally Brody (1945) intended the equation to be used after the point of inflection in

the self-inhibiting phase of growth, however no restrictions were placed on the values

for time (Doren et al., 1989). This may account of the tendency for the function for

overestimate early life growth. An example of the Brody function is shown in Figure

2-3.

Figure 2-3 A typical graphical representation of lamb live weight data with a Brody function fitted to

lamb live weight data (o) in kilograms against time in days.

In cattle the Brody function is also used to model weight (Doren et al., 1989) providing

the best fit between 18 months and maturity of Angus cattle in the Beltrán et al. (1992)

study. Before 18 months of age the Richards function gave the best fit due to the slight

overestimation of early weights with the Brody model, a fitting characteristic alluded to

previously.

A main disadvantage of the Brody function, and all functions fitted on an individual

basis, is the inability to accommodate different durations of data collection without

introducing bias estimating weights close to the edge of the available data. Population

0

5

10

15

20

25

30

35

40

45

0 50 100 150 200 250 300 350

Wei

gh

t (k

g)

Age (days)

12

based functions are less biased, accommodating different lengths as well as frequencies

of data.

2.2.3 Random Effects Model

A random effects model is a curve fitting methodology based on a population fit, with

the curve of each individual animal varying from the population curve. It is based on

polynomial functions with random coefficients, with orthogonal Legendre polynomials

chosen for their convergence properties (van der Werf and Schaeffer, 1997). A random

effects model is favoured for its ability to allow for variation in the covariance of weight

measurements with time.

Population based models have been widely used for the analysis of longitudinal data in

animals including live weight gain in pigs (Andersen and Pedersen, 1996), cattle

(Meyer, 1999; Meyer, 2000) and sheep (Fischer et al., 2004; Lewis and Brotherstone,

2002).

There are a number of important advantages, with the models able to be used for

animals with relatively few weight recordings or animals with records that are unevenly

distributed across ages of measurement (Fischer and Van der Werf, 2002). Therefore

random effects models can easily accommodate missing weight data from individual

animals, with the curve being informed by animals with similar genetic and

environmental effects (for example; sex, dam age and sire type). The ability to

accommodate different durations of data collection results in less biased estimates of

weight. The methodology also allows environmental effects specific to the time of

recording to be accounted for, and differences in the shape of each animals growth

curve can be accommodated (Lewis and Brotherstone, 2002).

13

A noted limitation is that nonlinear models cannot currently be accommodated (Lewis

and Brotherstone, 2002) so the coefficients do not have a clear biological interpretation.

Furthermore, when polynomials are used there is instability at the edges of the

trajectory (Fischer and Van der Werf, 2002; Meyer, 2002; van der Werf and Schaeffer,

1997) potentially leading to the overestimation of weight close to maturity.

The instability due to the use of polynomials is currently unresolved. Lewis and

Brotherstone (2002) proposed transforming nonlinear function into their linear analogue

(Portolano and Todaro, 1997) however this may require several transformations of the

nonlinear functions. Attempts to fit such a model have been problematic with difficulty

defining starting values and achieving convergence.

2.2.4 Residual weight

As the fit of a growth function appears to vary between breeds and experiments, it is

important to be able to compare different fits. Residual weight, the difference between

the estimated weight and the live weight recording, is commonly used as a measure of

curve fit as it is applicable to both linear and nonlinear curves. Average, squared or

summed residual weights have all been have been used to determine curve fit in sheep

(Topal et al., 2004) and cattle (Brown, 1970; Brown et al., 1976; Goonewardene et al.,

1981; Parks, 1982).

An advantage of the use of residual weights to determine curve fit is that residual

weight values are produced for each weight recording for each animal. Thus, curve fit

during different periods of growth can be determined by assessing residual weights

across the entire curve fit, as well as during the time period of interest.

14

2.3 Growth and tissue differentiation

The body is composed of bone, muscle, fat and viscera. Carcass composition refers to

the proportion of these tissues in the body (Berg and Butterfield, 1968). Bone, muscle

and fat vary in their rates of maturation (Berg and Butterfield, 1968; Butterfield, 1988).

During early growth muscle and bone growth predominate over fat deposition (Perry

and Thompson, 2005). As sheep mature, the proportion of bone declines, muscle

remains stable and the proportion of fat increases (Lewis et al., 2004; Thompson et al.,

1985a). Thus muscle and bone have been described as early maturing tissues and fat as

a late maturing tissue (Butterfield et al., 1983; Butterfield et al., 1984), as seen in Figure

2-4.

Figure 2-4 Change in carcass fat, muscle and bone as a portion of mature live weight during allometric

growth: Source (Butterfield et al., 1984)

15

An animal of large mature size takes a longer time to reach its mature weight than the

animal of small mature size. Differences in composition are evident when comparing

lambs tending toward different mature sizes. At a given age, the lamb tending toward a

large mature size will be less mature and therefore have a greater proportion of bone

and muscle than a lamb tending toward a smaller mature size which will be more

mature and have a greater proportion of fat (Berg and Butterfield, 1968; Butterfield et

al., 1984).

Mature size can vary with genotype, environment and production factors such as sex. In

the Australian lamb industry, selection pressure has been place on Terminal sire breeds

(Brown et al., 2007) for increased growth (Hall et al., 2002). This has resulted in leaner

carcasses at slaughter as the progeny of Terminal sires are reaching target market

weight when they are younger and less physiologically mature.

Despite selection pressures, when lambs do reach maturity they have similar

proportions of body tissues regardless of mature size (Thompson et al., 1985a).

Therefore there is no discernible difference between breeds of sheep in the rate of

maturation of bone, muscle and fat and breed differences in composition largely reflect

differences in maturity.

Mature size can vary with season. Between weaning and maturity, body composition is

subject to seasonal changes with Merino rams having increased fat weight during

summer and muscle weight during winter and Merino ewes having increased fat weight

during autumn in ewes (Ball et al., 1996).

Mature size also varies between sexes leading to variations in carcass composition at

maturity between rams, wethers and ewes despite similar patterns of maturation of

bone, muscle and fat. This was confirmed by Butterfield et al. (1984) who demonstrated

16

a greater proportion of muscle in rams and a greater proportion of carcass fat in wethers

and ewes.

2.4 Production factors influencing growth

Lamb growth is influenced by a variety of genetic and production factors. These factors

include sex, mature size, breed, nutrition, birth type and dam factors. An understanding

of the impact of these production factors on growth has specific advantages for

producers.

Producers are able to manipulate some production factors, such as breed and nutrition,

to achieve specific goals in lamb production including accelerated lamb growth. Other

factors, such as sex and birth type, are more difficult to manipulate, although an

understanding of how their impact on growth varies allows producers to moderate

expectations.

2.4.1 Sex

It is well established that male lambs have a heavier mature weight and are heavier than

female lambs at any given age (Gbangboche et al., 2008; Gbangboche et al., 2006b;

London and Weniger, 1994; Thompson et al., 1985a). Weight varies between male

lambs with entire male lambs heavier than castrate lambs. This is largely due to higher

levels of testosterone, a potent muscle growth stimulant, in entire male lambs

(Lawrence and Fowler, 2002). Thus, entire male lambs are heavier than castrate lambs

which are heavier than ewe lambs (Afolayan et al., 2007).

To achieve heavier weights, entire male lambs grow significantly faster than both

wether or ewe lambs (Lee et al., 1990). Likewise, Lee et al. (1990) found that wethers

17

grew faster than ewe lambs during early life growth, achieving heavier weights. Post

weaning, ewe lambs were found to grow faster than wether lambs although wether lamb

weights were not reached before lambs were slaughtered at six months of age.

While entire male lambs are faster growing and heavier than wether lambs it may not be

practical in a production setting to keep all male lambs entire. Ewe lambs are even

slower growing than wether lambs and will take longer than wether lambs to reach

target weights.

2.4.2 Genetics

2.4.2.1 Breed

In Australian sheep production sire types are classified by Sheep Genetics as Terminal,

Maternal or Merino (Brown et al., 2007). Terminal breed characteristics include rapid

lean growth of lambs, such as for prime lamb production. There are many Terminal

breeds, some of which include Suffolk, Texel and Poll Dorset. Maternal breeds such as

the Border Leicester and Dohne Merino have well developed reproductive and

mothering traits. Merino breed characteristics include superior wool production with

Merino breeds including Merino and Poll Merinos. Composite breeding has become

more common in the Australian sheep industry and when coupled with heterosis can

lead to increased fertility and litter size compared to the parental mean (Fogarty et al.,

1984).

Breeds differ in their mature weights and in the length of time taken to reach maturity

(Schreurs et al., 2008). The impact of sire and dam breeds on growth potential in their

progeny is largely a reflection of the mature size of the breed. Aligning with this,

progeny of Terminal breeds have been shown to be heavier than the progeny of

Maternal breeds at all ages (van der Werf and Wheaton, 1999). The progeny of breeds

18

with a larger mature size also have higher growth rates during all phases of growth than

the progeny of breeds with a smaller mature size (Perry and Thompson, 2005). This was

demonstrated by Thompson and Parks (1983) who showed food intake, efficiency and

consequent growth were largely functions of the mature size of the animal.

Ewe breed is a significant predictor of lamb birth weight (Afolayan et al., 2007; Holst et

al., 2002) which is positively correlated with adult or mature weight (Huisman and

Brown, 2008). A group of studies evaluating the impact of ewe breed on the weight and

growth rate of the progeny demonstrated that lambs from Border Leicester-Merino ewes

were both heavier than lambs from Merino ewes at birth (Holst et al., 2002) and had

higher pre weaning growth rates (Atkins, 1980; Atkins and Gilmour, 1981; Atkins and

Thompson, 1979). Specifically, Atkins and Thompson (1979) demonstrated a 40 gram

per day higher pre weaning growth rates in the progeny of Border Leicester-Merino

dams compared to the progeny of Merino dams.

Sire and dam breed both contribute to lamb growth and a change in genotype through

the introduction of a new breed or cross can impact lamb weights and growth rates in a

production system focused on high lamb growth.

2.4.2.2 ASBVs

Australian Sheep Breeding Values (ASBVs), developed by Sheep Genetics, are used to

make targeted genetic change within individual production systems within the

Australian sheep industry. Sheep Genetics analyses performance values, pedigree

information and relevant environmental and management information to determine

ASBVs for economically significant growth, wool, fertility, fitness and carcass traits

(Brown et al., 2007). There are several main components including LAMBPLAN™ for

Terminal and Maternal breeds and MERINOSELECT™ for Merino and Merino-based

breeds.

19

ASBVs are essentially an estimate of how that animals progeny will perform for a range

of traits (Brown et al., 2006) and contain information about the name of the trait, the

stage at which the effect is expected in progeny, an ASBV value and an accuracy, as

seen in Figure 2-5. Using ASBVs producers can select for increased growth in lambs.

This is achieved by selecting for a higher ASBV value for weight at the desired time

such as at birth, weaning or post weaning.

Figure 2-5 Interpretation of an Australian Sheep Breeding Value (ASBV) for post weaning weight based

on value, trait, age and accuracy. Source: Sheep Genetics

Producers are also able to select for changes in multiple traits at once using a selection

index. A selection index combines the values of ASBVs for several different traits into

one number. As can been seen in Figure 2-6 rams with differing ASBVs for weight,

muscle and fat may have a similar index value based on the weighting applied to each

individual ASBV. Indexes vary based on the breeding objective of the particular index.

20

For example, the Carcase Plus index has a 60% emphasis on increasing growth, 20% on

decreasing fat and 20% on increased eye muscle depth.

Figure 2-6 Selection index for rams (circled) based on Australian Sheep Breeding Value for weight

(PWT), muscle (PEMD) and fat (PFAT). Source: Sheep Genetics

With the fall in wool prices there is an increasing shift toward meat production (Banks,

2002; Curtis et al., 2006) and thus breeding for improved growth and carcass

characteristics. In Merinos, characteristics of live weight, fat depth and eye muscle

depth have been shown to be moderate to highly heritable. Therefore there is merit in

selection for these characteristics using breeding values (Clarke et al., 2003) with

ASBVs reliably predicting yearling and hogget weights in Merinos (Brown et al.,

2005).

Industry focus in on large, lean lambs with increased average slaughter weights (Hall et

al., 2002; Pethick et al., 2006a). This is achieved through multi trait selection to provide

simultaneous improvements in growth, fat depth and muscle depth (Banks et al., 2002).

21

Post weaning weight

Australian lamb producers select for lamb growth via an increased sire post weaning

weight (PWWT) ASBV. Selection for growth produces heavier lambs (Fogarty et al.,

1997; Hall et al., 2002) with increased mature size (Huisman and Brown, 2008), growth

rate (Thompson et al., 1985b), weight at slaughter and hot carcass weight (Gardner et

al., 2015; Gardner et al., 2010). Thus when lambs from high PWWT sires are compared

to lambs from low PWWT sires of the same age they are less physiologically mature.

The expression of the growth ASBVs in lambs, such as PWWT, is negatively impacted

by low nutrition of both the lamb (Hall et al., 2002; Hegarty et al., 2006b) and the ewe

(Hegarty et al., 2006c). In addition to environmental factors, production factors such as

multiple births, dam breed and dam age may restrict lamb nutrition.

Post weaning fat depth

Selection for a reduced post weaning fat depth (PFAT) in lambs reduces fat depth at the

c-site (5cm from the midline over the 12th

rib) in the loin (Hall et al., 2002; Hegarty et

al., 2006b), GR tissue depth (110 mm from the midline over the 12th

rib) (Hegarty et al.,

2006b) and loin fat weight (Gardner et al., 2010). The full expression of the PFAT

ASBV has been shown to be reduced with suboptimal lamb nutrition (Hegarty et al.,

2006b).

Post weaning eye muscle depth

Selection for increased post weaning eye muscle depth (PEMD) in lambs increases

muscle at the c-site in the loin (Hall et al., 2002; Hegarty et al., 2006a; Hegarty et al.,

2006b) by around 0.61 mm for each mm increase in sire PEMD ASBV irrespective of

plane of nutrition (Hegarty et al., 2006b). Selection for an increased PEMD and reduced

PFAT has been shown to increase lean meat yield in lambs (Gardner et al., 2010).

22

2.4.3 Nutritional restriction

Adequate nutrition is essential to maximise lamb growth potential. Lambs fed optimal

nutrition grow in a sigmoidal fashion whereas lambs with restricted nutrition have

fluctuations in their growth rate based on nutritional quality and availability. Producers

are able to mitigate some forms of nutritional restriction, such as restriction due to

seasonal pasture availability, through supplementary dam and lamb feeding to minimise

the impact on lamb growth where possible.

Environmental factors cause variation in lamb weights (Gbangboche et al., 2006a;

Gbangboche et al., 2008; Gbangboche et al., 2006b) with seasonal variation influencing

feed quality and availability (Perry and Thompson, 2005). Lambs born during high

rainfall seasons are less likely to experience nutritional restriction and are heavier and

faster growing than lambs born in dry seasons (Gbangboche et al., 2008; London and

Weniger, 1995). Nutritional restriction due to environmental and production factors can

also take place in-utero and pre weaning. When nutritional restriction is lifted post

weaning growth is accelerated, however additional nutritional and economic inputs may

be required to reach target weights after slow pre weaning growth (Greenwood et al.,

2009)

2.4.3.1 In-utero restriction

Animals have a predetermined mature size which is programmed during the early

embryonic stages of life however optimal nutrition is necessary for maximal expression

of genetic characteristics (Butterfield, 1988). In line with these findings Cafe et al.

(2006) found that plane of nutrition accounted for approximately 20% of the variation in

birth weight when comparing progeny of poorly fed and well fed Hereford cows.

23

The number of myofibres in each muscle is mostly determined prior to birth

(Greenwood et al., 2000a; Nissen et al., 2003). Malnutrition during the foetal period

may decrease the number of secondary fibres leading to a permanent reduction in

postnatal muscle growth potential (Hegarty and Allen, 1978; Wigmore and Stickland,

1983).

In line with these findings Fahey et al. (2005) found that restricted maternal nutrition

resulted in lambs developing fewer fast (glycolytic) fibres and more slow (oxidative)

fibres. This effect was seen when maternal nutrition was restricted before fibre

formation in lambs (days 30-70) but not with restriction after this period, highlighting

the importance of optimal maternal nutrition during early lamb growth in-utero.

2.4.3.3 Pre weaning restriction

Pre weaning, the dam provides the majority of the nutrition for the lamb through its

milk. However, when lactation declines the intake of pasture by the lamb increases

(Langlands, 1972). The main source of nutritional restriction during lactation occurs in

multiple birth lambs which experience sibling competition for available nutrition. Dam

milk volume increases by only 27% at peak yield in ewes with twin born lambs,

restricting milk volume on a per lamb basis (Morgan et al., 2006).

2.4.3.4 Post weaning restriction

In the post weaning phase, growing animals which have undergone any form of

restricted nutrition often experience accelerated growth when the restriction is lifted

(Berg and Butterfield, 1968; Hendrick et al., 1989). This has been demonstrated in

sheep (Thornton et al., 1979) and cattle (Brandstetter et al., 1998b) where growth was

depressed during a period of nutritional restriction followed by compensatory growth

upon restoration of nutrition.

24

2.4.4 Birth type and rear type

Number of lambs born per ewe, also defined as birth type, significantly impacts birth

weight at it acts as a form of in-utero nutritional restriction for multiple births. This

results in single born lambs being heavier than multiples (Yapi-Gnaorè et al., 1997)

such as twin born lambs (Gbangboche et al., 2008) and triplet born lambs (Afolayan et

al., 2007).

Gbangboche et al. (2008) hypothesised that the differential in weight in single versus

multiple birth types was due to a limited capacity for the dam to provide further in-utero

nourishment for the development of multiple foetuses and colostrum for new born

lambs. This pattern continues through to weaning where weight in multiple birth lambs

is further attenuated by increased rearing type (Afolayan et al., 2007). Hence, a lamb

born as a twin but raised as a single lamb is heavier than a lamb born and raised as a

twin.

Birth type is difficult to control in a production setting as many breeds of sheep have

high ovulation rates (Mulsant et al., 2001; Scaramuzzi and Radford, 1983). It is possible

to increase the ovulation rate (Juengel et al., 2004) however reducing the ovulation rate

is not easily achieved. Increased ewe and lamb nutrition is currently the most effective

and easily adoptable tool for optimising growth rates for multiple birth lambs.

2.4.5 Dam factors

The dam plays a significant role in lamb nutrition and growth being the sole source of

nutrition during early life growth. Multiple birth lambs are particularly susceptible to

factors which influence the ability of the dam to provide adequate nutrition to lambs.

These factors include dam nutrition, milk production and dam age.

25

2.4.5.1 Nutrition and feed restriction

While ewe nutrition is integral for the provision of lambs with adequate nutrition, the

significance of the nutritional restriction on progeny varies with the timing and duration

of the restriction. Short periods of nutritional restriction in Merino ewes during early

pregnancy did not significantly impact postnatal lamb growth if nutrition was adequate

in late pregnancy (Krausgrill et al., 1999). However, when ewe nutritional restriction

was sustained during lactation, subsequent underfeeding of lambs reduced the

expression of the genetic potential for growth in lambs (Hegarty et al., 2006b; Hegarty

et al., 2006c).

The sensitivity of lambs to dam nutrition varies when lambs are in-utero with triplet

born lambs more sensitive than twin born lambs. This was demonstrated by Everett-

Hincks et al. (2005) who showed that doubling pasture availability for ewes had no

effect on the birth weight of twin lambs while the birth weight and subsequently

survival of triplet lambs significantly increased.

2.4.5.2 Milk production

The maternal environment plays a role in determining birth weight and early life growth

of progeny, especially prior to weaning. Milk production by the dam is highly correlated

with lamb growth (Burris and Baugus, 1955). Therefore multiple birth lambs, which

have access to a reduced volume of milk per lamb compared to single born lambs

(Morgan et al., 2006), have reduced weight gain (Burris and Baugus, 1955). The

composition of milk is stable, despite multiple births, so volume is the limiting factor

(Burris and Baugus, 1955). Milk volume is a limiting factor regardless of multiple births

as single born lambs from ewes which produce high milk volumes grow faster than

single born lambs from ewes with low milk production (Burris and Baugus, 1955).

26

Ewe breed impacts milk production and lamb growth. Langlands (1972) showed that

lambs reared by Border Leicester ewes received more milk than lambs reared by Merino

ewes. The reduced milk production of Merino ewes was considered to be of secondary

significance if herbage was available ad libitum as lambs with high voluntary intakes

compensated for the lack of milk by consuming greater amounts of forage. However,

Merino lambs consumed less grass than Border Leicester lambs and grew more slowly

due to their lower overall voluntary food intake.

2.4.5.3 Dam age

As dam age and parity increases so does lamb birth weight (Gbangboche et al., 2008;

London and Weniger, 1995; Yapi-Gnaorè et al., 1997). This relationship is nonlinear

with a decline in lamb birth weight with increasingly older dams (Brown and Reverter,

2002; Huisman et al., 2008). In a study of 1-4 year old dams Afolayan et al. (2007)

demonstrated an average increase of 420 g in birth weight per year of increasing dam

age. The effect was considerably smaller in an earlier study by Ali et al. (2006) which

showed a 25 g increase in birth weight per year of increasing dam age. However, in this

study there was a larger range of dam ages, varying from 1 to 9 years old. In ewes aged

6 years and above, the effect of increasing birth weight was generally smaller and

occasionally negative, which may contribute to the lower average birth weight

compared to that found in the study by Afolayan et al. (2007). Borg et al. (2009) found

4-6 year old dams produced the heaviest lambs, although it was not possible to

determine whether the effect tapered beyond this age dams older than 6 years were not

included in the study. However, when considering these studies there was a trend for

lighter lambs to be born to younger and older dams.

Interestingly, a study on pregnant ewe lambs which were 7 months of age at

insemination showed that ewes fed to achieve moderate growth rates of 75g per day had

27

larger placental cotyledon mass and produced lambs with higher birth weights than

ewes fed to achieve higher growth rates on the same ration. Colostrum yield was also

lower in ewe lambs fed the greater volume of the same ration (Da Silva et al., 2001).

These findings seem counter-intuitive and the authors did not hypothesise why this may

have occurred. Furthermore, no information was provided on the amount (kg) of feed

offered or consumed by the high intake ewe group or expected and actual growth rates

in this group. Young, maiden ewes are known to lose more weight under conditions of

nutritional restriction than mature ewes, resulting in slower growth of progeny through

to slaughter (Hegarty et al., 2006b). However, when nutrition intake is high, the

influence on lambs is expected to be reduced. London and Weniger (1995) suggested

that since maiden ewes are generally still growing they are in competition for nutrients

with the foetus and this effect may be exaggerated in adolescent ewes.

2.5 Growth and fibre type

2.5.1 Muscle fibre classification

Skeletal muscle fibres make up 75-90% of skeletal muscle volume. They are classified

on the basis of several physical and metabolic properties, which include differences in

morphological traits, energy metabolism, contractile properties, metabolic properties

and colour (Lee et al., 2010; Lefaucheur, 2010; Picard et al., 2002).

Histochemical staining of skeletal muscle is a common procedure used to determine

fibre types by staining the contractile component of skeletal muscle, myosin. Myosin,

the major protein of skeletal muscle, is responsible for the contractile component of the

muscle fibres. It is composed of four light chains and two heavy chains (MHC), which

can exist in a number of different isoforms (Klont et al., 1998). The histochemical

28

classification is based on the MHC isoform and differentiates fibre types into type I, IIA

and IIB fibre types (Brooke and Kaiser, 1970).

The classification was revised to include a fourth fibre type when type IIX was

identified in several animals including mouse, rat, guinea pig and rabbit skeletal muscle

(Bär and Pette, 1988; Schiaffino et al., 1989). The IIX fibre type was later identified in

humans (Smerdu et al., 1994), pigs (Lefaucheur et al., 1998) and cattle (Tanabe et al.,

1998). The speed of contraction increases in the rank order I < IIA < IIX < IIB with type

I fibres considered slow twitch and type IIB fibres considered fast twitch (Schiaffino

and Reggiani, 1996).

The four muscle fibre types can also be identified according to their relative oxidative or

glycolytic metabolic characteristics. The metabolic types are differentiated by the

pathway used for regeneration of adenosine triphosphate (ATP) in the muscle

(Dubowitz and Pearse, 1960). The oxidative pathway, which oxidises glycogen,

glucose, lipids, ketone bodies and amino acids in the mitochondria, is aerobic with a

high oxygen requirement. The glycolytic pathway, through which glycogen is rapidly

converted to lactate, is anaerobic with no oxygen requirement (Lefaucheur, 2010).

Oxidative fibres are associated with higher levels of oxidative activity and thus

oxidative enzymes. Rate limiting mitochondrial enzymes such as citrate synthase and

isocitrate dehydrogenase are used to characterise the oxidative capacity of muscles

(Hocquette et al., 1998) with the relative expression of oxidative and glycolytic

enzymes used as an indicator of muscle fibre type in sheep (Gardner et al., 2007;

Gardner et al., 2006b; Lefaucheur, 2010).

The rate limiting enzymes of glycolysis, hexokinase, phosphofructokinase and lactate

dehydrogenase plus glycogen synthase and glycogen phosphorylase are commonly used

29

to qualitatively characterise the glycolytic activity of muscles (Hocquette et al., 1998).

Where oxidative and glycolytic fibres exist within the same muscle, the relative

composition of the muscle determines the predominance of the muscles metabolic

properties (Ozawa et al., 2000; Ryu and Kim, 2005).

The relative importance of these two metabolic pathways, as evidenced by ATPase

activity and enzyme expression (Barnard et al., 1971; Dubowitz and Pearse, 1960;

Engel, 1962; Peter et al., 1972), has been used to characterise muscle fibres as:

oxidative (type I, slow), oxido-glycolytic (type IIA, fast) and glycolytic (type IIX and

IIB, fast) (Ashmore and Doerr, 1971; Peter et al., 1972; Schiaffino and Reggiani, 1996).

In this thesis, metabolic characteristics will be used to define fibre type.

A summary of the contractile and metabolic characteristics, and the strength of

association, of each fibre type can be found in Table 2-1.

Fibre type I IIA IIX IIB

Contraction speed + +++ ++++ +++++

ATPase activity + +++ ++++ +++++

Oxidative metabolism +++++ ++++ ++ +

Glycolytic metabolism + ++++ ++++ +++++

+, very low; ++, low; +++, medium; ++++, high; +++++ very high.

Table 2-1 Contractile and metabolic characteristics of individual fibre types. Adapted from Lefaucheur

(2010).

30

2.5.2 Oxidative fibres

Oxidative fibres are red slow twitch type I fibres. These fibres require a high level of

oxygen and express large amounts of myoglobin for oxygen storage (Lefaucheur, 2010).

Due to the high myoglobin content and heme based pigments in mitochondria they

appear redder in colour (Murray et al., 1996). The mitochondria are also larger and

more populous (Klont et al., 1998) in oxidative compared to glycolytic muscle cells

(Hoppeler, 1985). Oxidative fibres have low ATPase activity, allowing for sustained

prolonged low power work (Lefaucheur, 2010). The ability to sustain low intensity

contractions for basic movements is facilitated by efficient vascularisation (Nelson and

Cox, 2005), small fibre diameter and a high Ca2+

sensitivity (Lefaucheur, 2010).

2.5.3 Glycolytic fibres

Glycolytic fibres are white, fast twitch type II fibres. When compared to oxidative fibres

they have a lower vascularisation, larger fibre diameter and fewer mitochondria (Nelson

and Cox, 2005), and are associated with lower levels of myoglobin, thus appearing paler

in colour. Type IIB glycolytic fibres were first demonstrated in the longissimus in cattle

by Picard and Cassar-Malek (2009). Glycolytic fibres have a high ATPase activity

(Ashmore and Doerr, 1971). These fibres use only glucose as an energy source (Murray

et al., 1996) and sustain brief and intense contractions (Lefaucheur, 2010).

2.5.4 Muscle fibre development

Muscle fibre development begins during early embryonic life when the myogenic

precursor cells (myoblasts) proliferate and fuse into myotubes. Myoblasts have been

isolated at different stages of development and are called embryonic, foetal and adult

myoblasts based on their distinct properties (Harris et al., 1989; Hoh et al., 1988;

Stockdale, 1992). Muscle fibres in adult animals are derived from different myoblast

31

populations (Harris et al., 1989; Vivarelli et al., 1988) and the regulation of this process

is complex (Stockdale, 1992).

In lambs there are three waves of myotube formation from myoblasts which

differentiate to form the origin of the different muscle fibre types. These are the

primary generation, which differentiates for form type I fibres, the secondary

generation, which differentiate to form type II fibres and the third generation which

differentiates to form both type I and type II fibres (Picard et al., 2002).

Muscle fibres express developmental MHC during gestation which is progressively

replaced by adult MHC (Picard et al., 1994; Robelin et al., 1993). Primary fibres in

sheep appear at 32 days of foetal life (Wilson et al., 1992). These fibres express the

slow MHC I isoform, characteristic of type I fibres (Cho et al., 1994; Condon et al.,

1990; Lyons et al., 1990; Picard et al., 1994; Robelin et al., 1993). Secondary fibres in

sheep appear at 38 days of foetal life (Wilson et al., 1992). Most of these develop into

type IIA, IIX or IIB fibres (Ashmore et al., 1972; Picard et al., 1994; Robelin et al.,

1990). Tertiary fibres in sheep appear between 62 and 76 days of foetal life (Wilson et

al., 1992). These develop into a mix of type I and type II fibres (Picard et al., 2002;

Picard et al., 1994).

Thus as developmental MHC expression reduces the proportion of adult MHC increases

resulting in the development of type I and type II fibres. In sheep, foetal MHC

disappears between 140 days of foetal life and 28 days after birth (Maier et al., 1992).

Fast adult MHC mostly occurs after birth in sheep with the proportion of type I fibres

increasing during the first 8 weeks after birth (Suzuki and Cassens, 1983).

Following the waves of myogenesis, the total number of fibres is fixed. In cattle this is

complete by the last third of gestation (Gagniere et al., 1999; Picard et al., 1994) and in

32

pigs by 80% gestation (Wigmore and Stickland, 1983). This is consistent with the

finding that muscle fibre numbers remain constant after birth in cattle (Joubert, 1956;

Robelin et al., 1990; Wegner et al., 2000). Therefore, increase in muscle volume is due

to the increase in size of the existing fibres (Hawkins et al., 1985; Wegner et al., 2000).

2.5.5 Production factors impacting fibre type

Lamb growth rate has a large impact on muscle fibre type. In addition to growth, muscle

fibre type is influenced by a variety of environmental, genetic and production factors.

These include sex, nutrition and breed. The response of fibre type to these factors varies

between different skeletal muscles. Unless stated otherwise the changes reported in

Section 2.5.5 occurred in the longissimus muscle, which is the focus of this thesis due to

its commercial value.

2.5.5.1 Age and growth rate

Change in muscle fibres characteristics is related to several factors however age has the

most significant impact on fibre type (Suzuki and Cassens, 1983; White et al., 1978).

Type I oxidative fibres and activity of oxidative enzyme isocitrate dehydrogenase have

been shown to increase with age in cattle (Brandstetter et al., 1998a) and sheep

(Greenwood et al., 2007; Suzuki and Cassens, 1983) while activity of the glycolytic

enzyme lactate dehydrogenase is reduced (Brandstetter et al., 1998a).

Muscle fibre characteristics also vary with growth rate. The comparison of extreme

genetic models of growth, such as wild compared to domesticated animals within

different species, suggests that selection for increase growth rates correlate with

increased myofibre diameter and a less oxidative metabolism (Ashmore, 1974; Rahelic

and Puac, 1981; Ruusunen and Puolanne, 2004; Solomon and West, 1985; Weiler et al.,

1995). These findings align with a shift between type IIA and IIB fibres with an

33

increased hot carcass weight (Hawkins et al., 1985), which is a proxy for increased

whole of life growth rate. However, the relationship varies within conventional

domestic species and increased growth rates in lambs did not impact the proportions of

muscle fibre types in the Arnold and Meyer (1988) study.

2.5.5.2 Sex

Ryu et al. (2008) found no difference in muscle fibre type proportions between the

sexes in pigs and there was no difference in total number of fibres in pigs (Staun, 1963)

or mice (Rowe and Goldspink, 1969). This is despite significantly larger muscle fibres

being found in female pigs (Larzul et al., 1999; Miller et al., 1975; Solomon et al.,

1990).

Several studies have found that castration has little impact on muscle fibre classification

in cattle (LaFlamme et al., 1973; Ockerman et al., 1984; Young and Bass, 1984).

However, previous work on the m. longissimus dorsi by Brandstetter et al. (1998b)

demonstrated an increase in the proportion of oxidative type I fibres in bulls and of

glycolytic type IIB fibres in steers at 12 months of age. Castration was thought to have

delayed the shift toward oxidative type I fibres that occurs with age by decelerating the

re-conversion process of type IIB to IIA to I which occurs after birth in cattle.

2.5.5.3 Nutrition and feed restriction

Under Australian extensive production systems, season fluctuations have an important

role in determining feed availability and feed intake. Variation in nutrition due to

different proportions of pasture or concentrate or varied protein content has been shown

to impact on muscle fibre type and enzymatic expression (Greenwood et al., 2006b;

Moody et al., 1980). However, the associations between nutrition and muscle fibre

types vary. Hocquette et al. (2001) suggested that oxidative muscles are more sensitive

34

to nutritional variation than glycolytic muscles. Gardner et al. (2006b) suggested that

changes in lamb nutrition may uncouple the link between muscle fibre type and

metabolic enzyme expression, although in contrast to the Hocquette et al. (2001)

findings this was less prevalent for oxidative enzyme expression.

Feed restriction has been shown to induce changes in muscle mass and the proportion of

oxidative fibre types in ruminants. In a study of two groups of lambs Solomon and

Lynch (1988) found that restricted lambs, who were fed a diet containing a third less

metabolisable energy, had a lower proportion of type IIB fibres and a higher proportion

of type IIA. The trend toward a more oxidative metabolism in animals experiencing

feed restriction was also demonstrated by Brandstetter et al. (1998b) in growing bulls

that had been restricted for 3 months. This effect reversed with re-feeding with the

activity of the oxidative enzyme isocitrate dehydrogenase lower in re-fed bulls

compared to bull calves fed a growth ration (Brandstetter et al., 1998b).

2.5.5.4 Breed and genetics

Genetic selection for post weaning growth, muscling and leanness has been shown to

impact on muscle fibre type. Previous work by Greenwood et al. (2006c) has

demonstrated that selection for increased muscling and leanness through the use of sires

with high PEMD and low PFAT, increases the percentage of glycolytic type 2B/X

myofibres. In the same study, selection for growth through the use of sires with high

PWWT had a smaller effect than selection for muscle and leanness (Greenwood et al.,

2006c).

There is evidence demonstrating that a similar effect is seen in beef cattle, with an

increase in muscle linked to an increase in glycolytic muscle fibres. Charolais bulls with

a high muscle growth index experienced muscle hypertrophy with growth, due to a

higher total number of fibres (Picard et al., 2006). These hypertrophied muscles have

35

fibres with the contractile and metabolic properties of type IIX fibres, suggesting that

bulls with high growth potential also have greater numbers of glycolytic fibres.

2.5.6 Fibre type and meat quality

Muscle fibre characteristics are an important determinant of meat quality (Joo et al.,

2013) through its associations with meat colour, intramuscular fat, mineral content and

changes to oxidative capacity (Kim et al., 2013; Lee et al., 2010; Maltin et al., 1998;

Valin, 1988; Valin et al., 1982) . These relationships are discussed in Section 2.6.

An example of the impact of fibre type on meat quality is seen in the selection for

increased muscle in cattle which reduces muscle oxidative capacity and increases the

glycolytic potential of the muscle. This is associated with in a reduced myoglobin

concentration, resulting in paler meat (Bailey et al., 1982; Clinquart et al., 1994), which

is less appealing to the consumer. Myoglobin content is also an important determinant

of the nutritional value of meat as reducing myoglobin concentration proportionately

reduces red meat iron concentration (Hazell, 1982) and negatively affects the nutritional

value of the meat product (Pethick et al., 2006a).

Selection for growth can impact muscle fibre type, and muscle fibre type is associated

with meat quality traits. Therefore growth can impact meat quality traits and the

magnitude of these impacts needs to be quantified.

2.6 Growth and meat quality

Genetic and phenotypic selection for growth is common to increase growth rates and

achieve target weights over a shorter time frame. Selection for increased growth rates

affects muscle fibre type, and muscle fibre type impacts meat quality. Furthermore,

36

early and late life growth have been shown to be different traits (Lewis and

Brotherstone, 2002) so the impact of selection for growth on fibre type and

subsequently meat quality may vary with time. Therefore, the associations between

growth, fibre type and meat quality need to be explored.

Meat quality relates to eating quality, nutritional value and meat colour (Warner et al.,

2010). Consumers demand a consistently high quality lamb product (Pethick et al.,

2005a) and are willing to pay a premium for a high quality product (Pethick et al.,

2006a).

While lamb growth influences a variety of traits, this thesis will be restricted to the

influence of lamb growth on muscle fibre type and the meat traits which are directly

linked to muscle fibre type: intramuscular fat, shear force, consumer overall liking,

isocitrate dehydrogenase activity, myoglobin concentration, iron concentration, zinc

concentration and meat browning.

2.6.1 Intramuscular fat

Animals have multiple fat depots with intramuscular fat being the adipose tissue

contained within muscles. It is comprised of adipocytes which are embedded in a

connective tissue matrix with triacylglycerol (TAG) being the major lipid component

(Murray et al., 2004b).

Intramuscular fat is an early maturing depot in relation to total carcass fat, increasing

only marginally as sheep mature (McPhee et al., 2008; Thompson et al., 1987).

Adipogenic and lipogenic genes are expressed during early life growth in cattle

indicating an early molecular changes associated with intramuscular fat content (Wang

et al., 2009). These findings are in line with studies which indicate the potential for

development of intramuscular fat is fixed during early life growth (Beaulieu et al.,

37

2010; Kirkland and Dobson, 1997; Pethick et al., 2004). Factors which influence

intramuscular fat levels in lambs may therefore exert a greater effect in the perinatal

period than when animals are more mature.

Intramuscular fat is measured by chemical determination using a near infrared (NIR)

procedure with readings validated with chemical fat determination using solvent

extraction (Perry et al., 2001). Measurements are commonly made in the m.

longissimus lumborum due to the economic value of the cut and the importance of

intramuscular fat in meat eating quality (Pannier et al., 2014c). Intramuscular fat can

also be estimated by evaluating the amount of fat visible within the muscle tissue of the

m. longissimus lumborum in cattle, although it is less visible in sheep. This portion of

intramuscular fat, visible at normal storage temperatures, is known as marbling.

Marbling is readily identifiable enabling its use by industry, for example for grading

beef carcasses under the Meat Standards Australia (Polkinghorne et al., 1999). Marbling

is also used by consumers as an indicator of high eating quality, but high levels of

intramuscular fat have adverse effects on consumer acceptability due to increased

visibility of fat in meat. Importantly however, a minimum level of intramuscular fat in

meat is essential due to its positive correlation with the subjective measures of eating

quality, tenderness, flavour, juiciness and consumer rated overall liking (Thompson,

2004).

Consumers rate their satisfaction with lamb based on flavour, tenderness, odour,

juiciness and overall liking (Pethick et al., 2006b). Intramuscular fat is positively

associated with flavour, juiciness, tenderness and overall liking in sheep (Hopkins et al.,

2006; Pannier et al., 2014c) and pigs (Murray et al., 2004a). In beef intramuscular fat is

positively associated with flavour and juiciness (Thompson, 2004) and accounts for up

to 15% of the variation in palatability in beef (Dikeman, 1987). A reduction in

38

intramuscular fat is detrimental to these organoleptic meat qualities (Hocquette et al.,

2010; Wood et al., 2008) including juiciness , flavour (Wood et al., 2008) and

tenderness (Dikeman, 1987). Thus intramuscular fat is a key determinant of eating

quality in red meat (Harper and Pethick, 2004) and in sheep explains more of the

variation in tenderness and overall liking than production factors such as sex and sire

type (Pannier et al., 2014a).

Reduced levels of intramuscular fat result in meat which is perceived by consumers as

dry and less tasty (Channon et al., 2001; McPhee et al., 2008). Acceptable levels in

beef, pork and lamb are quoted by Savell and Cross (1988) as 3-4% while Hopkins et

al. (2006) found a minimum level of 5% was required to achieve an average sensory

score for overall liking for sheep meat.

Oxidative muscle fibres are associated with higher levels of intramuscular fat (Picard et

al., 2002) and intramuscular fat has strong associations with eating quality. Therefore

the impact of growth on fibre type and thus intramuscular fat is likely to impact all of

the meat traits examined in this thesis.

2.6.1.1 Intramuscular fat and growth

Evidence exists for growth impacting intramuscular fat with the rate of accretion

slowing down as animals near their mature size (McPhee et al., 2008).

Nutrition during growth is important for intramuscular fat development. Nutritional

restriction in lambs results in slower growth and weight gain causing a reduction in

intramuscular fat (Murphy et al., 1994). Conversely, pigs with increased growth rates

due to higher protein diets had reduced intramuscular fat (Karlsson et al., 1993).

However this is likely due to reduced maturity at the time of slaughter rather than a

reduction in intramuscular fat deposition.

39

Hot carcass weight at slaughter can be used as a proxy for whole of life growth rate to

determine the impact of growth in intramuscular fat. This approach was taken by

Pannier et al. (2014c) and demonstrated an increase in intramuscular fat in lambs with

increased growth.

Breeding values can be used to select for growth and lean meat yield in lambs. Selection

for increased growth using the PWWT ASBV did not impact intramuscular fat (Pannier

et al., 2014c). Meanwhile, selection for leanness and muscle using the PFAT and

PEMD ASBVs reduced intramuscular fat (Pannier et al., 2014c). Genetic potential for

muscle growth in bulls was found to be correlated with reduction in activity of the

oxidative enzyme citrate synthase, and therefore an increased glycolytic capacity of the

muscle fibre types in bulls and as oxidative capacity is positively linked with

intramuscular fat, bulls with higher muscle growth had lower intramuscular fat

(Hocquette et al., 2012).

Existing evidence for the impact of growth on intramuscular fat therefore indicates that

slow growth due to feed restriction leads to reduced intramuscular fat while increased

growth leads to heavier hot carcass weights and increased intramuscular fat. Increased

growth rates may also be associated with reduced intramuscular fat if animals are

slaughtered when they reach target weights at younger ages when they are less

physiologically mature. The influence of prenatal muscle fibre development on

intramuscular fat (Picard et al., 2002) may also lead to animals with an increased

potential for muscle growth, leading to higher proportions of glycolytic muscle fibres,

having reduced levels of intramuscular fat.

2.6.1.2 Production factors which influence intramuscular fat

Intramuscular fat varies with production factors including age, nutrition, sex, and

genotype. The increase in intramuscular fat with age is well documented in sheep

40

(Hopkins et al., 2007; McPhee et al., 2008; Pannier et al., 2014c; Pethick et al., 2005c)

although the increase is quantitatively small and there is a large degree of overlap

between animals of differing ages (Hopkins et al., 2005).

Nutrition impacts intramuscular fat development however the effect in sheep is small

compared to genetic impacts (McPhee et al., 2008; Pannier et al., 2014c). The timing of

the restriction is important. In cattle, in-utero nutritional restriction has not been shown

to impact intramuscular fat development (Greenwood and Cafe, 2007; Greenwood et

al., 2006a) while a 30% feed restriction in pigs during the growing-finishing phase

caused a reduction in intramuscular fat (Čandek-Potokar et al., 1999). Since nutritional

restriction limits growth, the impact of restriction on intramuscular fat may be delivered

through its impact on growth.

Sex has been shown to influence levels of intramuscular fat. Ewe lambs have been

shown to have greater amounts of intramuscular fat then wether lambs (Pannier et al.,

2014c) and ram lambs (Craigie et al., 2012). While Tejeda et al. (2008) showed no

difference in intramuscular fat between ewe lambs and ram lambs this conclusion was

based on a relatively small study in Merino lambs (n = 48), compared to n = 208 Texel

lambs in the Craigie et al. (2012) study and n = 5,867 lambs from Terminal, Maternal

and Merino sires in the Pannier et al. (2014c) study. Thus the small sample size of the

Tejeda et al. (2008) study may not be sensitive enough to detect a significant effect

between sexes.

Age has also been found to reduce the sex effect on intramuscular fat. Muscle samples

from the m. longissimus dorsi analysed for intramuscular fat from mature sheep showed

no differences between entire and vasectomised rams or between wethers and ewes.

However, intramuscular fat of entire and vasectomised rams was significantly lower

than in either ewes or wethers (Okeudo and Moss, 2007).

41

Intramuscular fat varies with lamb genotype (Díaz et al., 2005). As mature size, and

thus growth rate to attain mature size, varies between breeds the variation in

intramuscular fat with genotype may simply reflect the impact of growth rate on

intramuscular fat rather than a specific breed difference.

Díaz et al. (2005) found that the variation in intramuscular fat between sheep breeds

was similar in magnitude to the variation found by McPhee et al. (2008), with Border

Leicester x Merino animals having the highest increase in intramuscular fat in the loin,

compared to other breeds. The impact of the Border Leicester genotype on

intramuscular fat is consistent with the increase in overall carcass fat levels (Hopkins

and Fogarty, 1998; Hopkins et al., 2007; Ponnampalam et al., 2007).

In a study by Wegner et al. (2000), the level of intramuscular fat did not vary between

four breeds of cattle; German Angus, Galloway and Holstein Friesian. In the same

experiment Belgian Blue cattle, which are selected for double muscling with a mutation

of the myostatin gene causing muscle hyperplasia, had comparatively lower levels of

intramuscular fat. Belgian blue cows within this study also had a greater proportion of

glycolytic type IIB fibres, which are associated with lower intramuscular fat levels

(Hwang et al., 2010). The effect of fibre type on intramuscular fat levels will be

discussed in Section 2.6.1.3.

2.6.1.3 Intramuscular fat and fibre type

The relationship between intramuscular fat and fibre type is not always simple or clear

(Hocquette et al., 2010; Lefaucheur, 2010) however intramuscular fat content is linked

to oxidative muscle fibre types and oxidative enzyme activity. As discussed below,

oxidative muscle fibres are generally associated with higher levels of intramuscular fat.

A notable exception is the study in bison with highly oxidative muscles and low levels

of intramuscular fat (Agabriel et al., 1998).

42

Intramuscular fat is positively correlated with the percentage of oxidative muscle fibres

(Essen-Gustavsson et al., 1992; Maltin et al., 1998) and negatively correlated with

glycolytic muscle fibres (Hwang et al., 2010). In support of these findings oxidative

enzyme activity is increased in oxidative fibres in cattle and in breeds of cattle that have

higher levels of intramuscular fat (Hocquette et al., 2003; Jurie et al., 2007), with

oxidative muscle fibres being associated with higher levels of intramuscular fat (Picard

et al., 2002).

Intramuscular fat content and proportions of myosin heavy chains I (slow) and IIA (fast

oxido-glycolytic) and enzymatic indicators of oxidative metabolism; citrate synthase,

isocitrate dehydrogenase (ICDH) and cytochrome-c oxidase and the expression of H-

fatty acid binding protein have been shown to be higher in the slow oxidative rectus

abdominis muscle compared to the fast glycolytic semitendinosus muscle (Hocquette et

al., 2012). Amongst oxidative metabolic markers cytochrome-c oxidase was better

correlated to TAG than enzymatic indicators within the mitochondrial matrix

(hydroxyacyl-CoA dehydrogenase, citrate synthase and ICDH) (Hocquette et al., 2012).

Genotypes characterised by a high oxidative muscle metabolism have increased levels

of intramuscular fat (Jurie et al., 2007). This is interesting considering high fatty acid

catabolism is a characteristic of oxidative muscles, although a high fatty acid turnover

may also favour fat deposition (Gondret et al., 2001). Jurie et al. (2007) found a high

correlation between TAG deposition and A-fatty acid binding protein, which is

expressed exclusively within adipocytes and not within muscle fibres indicating that

intramuscular adipocyte number (rather than intra fibre fat) is the major biological

mechanism that contributes to TAG accumulation.

Therefore, muscles with higher intramuscular fat levels have a greater number of

oxidative muscle fibres and a greater ability to break down intramuscular fat. This may

43

reflect a synergistic biological relationship between these two tissues as the oxidative

muscle uses the intramuscular fat breakdown products as fuel.

2.6.2 Shear force

Warner-Bratzler shear force is an objective measure of tenderness and describes the

force required, in Newtons or kilograms, to shear a cooked meat sample with a Warner-

Bratzler shear blade (Hopkins et al., 2010). Whilst objective measurements of meat

quality such as shear force and compression have the advantage of being cheaper than

sensory panel testing for tenderness, they are rather simplistic one-dimensional

measures of a complex set of interactions which occur when cooked meat is chewed and

masticated in the mouth (Perry et al., 2001; Watson et al., 2008). Furthermore, they are

destructive to the carcass and time consuming (Hwang et al., 2008; Watson et al.,

2008). Despite this, Huffman et al. (1996) reported that consumer ratings were

consistent with Warner–Bratzler shear values in beef and (Safari et al., 2001) advocated

the use of shear force as a criterion for determining consumer acceptability of loin from

lambs.

Sensory scores are rated out of 100 by consumers. They are subjective measurements of

meat eating quality and have a negative correlation with shear force. Increasing shear

force is associated with a reduction in tenderness, flavour, juiciness and overall liking

(Hopkins et al., 2006; Pannier et al., 2014a; Safari et al., 2001). In the Pannier et al.

(2014a) study sensory scores for tenderness, overall liking, juiciness and flavour

reduced by 11.6, 8.5, 7.6 and 7.4 respectively across the 37 N range in shear force.

The phenotypic correlation between intramuscular fat and shear force after 5 days of

ageing in lamb is -0.30 (Mortimer et al., 2014), suggesting selection for one will also

influence the other. There are similar findings in pigs, where there is also a negative

44

correlation between shear force and intramuscular fat (Karlsson et al., 1993). As shear

force is correlated with intramuscular fat both can be used as indirect predictors of

eating quality. Indirect predictors are useful as the cost of measuring sensory parameters

with taste panels is often prohibitive.

2.6.2.1 Shear force and growth

Growth rate can be difficult to directly quantify and carcass weight, average daily gain

and nutritional restriction are commonly used as proxies for growth rate to investigate

the impact of growth on shear force. In general, shear force does not vary with carcass

weight, average daily gain or growth rate, although there are some exceptions.

Lambs slaughtered at either medium or light live weights had no differences in shear

force, despite differences in carcass fatness between the two groups (Vergara et al.,

1999). This aligns with the findings of Kemp et al. (1980) which also demonstrated no

relationship existed between shear force and slaughter weight in lambs. Conversely,

there was a low negative correlation between hot carcass weight and shear force of -

0.15 found by Vergara et al. (1999), however this was not supported by a phenotypic

association in the same study. There are conflicting reports in beef with a reduction in

shear force seen in crossbred steers as carcass weights increased with age (Bouton et al.,

1978) but no relationship between shear force and carcass weight Japanese Black steers

(Ozawa et al., 2000).

Studies of average daily gain and shear force in cattle had varied results. A group of 48

bulls with average daily gains ranging from between -0.55 kg and 1.36 kg showed no

relationship between shear force and growth rate (Calkins et al., 1987). This aligns with

the findings of Perry and Thompson (2005) of no change in shear force with variations

in average daily gain and Greenwood et al. (2006a) of no influence of pre weaning

growth on shear force.

45

Meanwhile, Fishell et al. (1985) assigned 36 steers to one of three feeding regimes for

120 days pre slaughter. The unrestricted group had an average daily gain of 1.42 kg, the

moderately restricted had an average daily gain of 0.77 kg and the highest restricted

group had an average daily gain of 0.34 kg. In both the longissimus and

semimembranosus shear force was significantly lower in unrestricted animals with

faster growth rates compared to animals with restricted nutrition and slower growth

rates. This relationship is further evidenced by the negative phenotypic correlation

found between average daily gain and shear force in cattle (Shackelford et al., 1994).

The difference is in the Calkins et al. (1987) and Fishell et al. (1985) studies is unlikely

to be related to the effects of testosterone between bulls and steers with Moloney et al.

(2000) reporting no difference in shear force with growth rate in steers.

Evidence for the association between shear force and growth is not clear. While there

was a small negative correlation between carcass weight and shear force in lambs there

was no phenotypic association (Vergara et al., 1999). In cattle there was also a negative

correlation, between average daily gain and shear force (Shackelford et al., 1994),

however the phenotypic association was not present in all studies discussed above.

2.6.2.2 Production factors which influence shear force

Shear force varies with production factors including age, sex, and genotype with shear

force being increased in older sheep (Hopkins et al., 2007) and cattle (Shorthose and

Harris, 1990) demonstrating an increase in sheer force with age at slaughter in cattle due

to an increase in collagen crosslinking.

Unlike age, sex and genotype do not appear to have a significant impact on shear force

in lambs with Kemp et al. (1980) demonstrating no difference between the sexes.

46

Shear force in 60 lambs did not vary between common Australian genotypes for lambs

born to Texel, Poll Dorset, Border Leicester and Merino rams (Safari et al., 2001).

Likewise there was no relationship between shear force and genotype in the Hopkins

and Fogarty (1998) study which involved a total of 436 lambs nor the Speck et al.

(1997) study which examined shear force in New Zealand lambs of different genotypes.

In cattle, shear force does not vary with breed with Wegner et al. (2000) showing no

difference between shear force in German Angus, Galloway and Holstein Friesian

cattle.

While shear force varied with age, there was no variation with sex or genotype. As sex

and genotype can represent differences in growth rates these findings align with the lack

of a phenotypic association between shear force and growth in lambs.

2.6.2.3 Shear force and fibre type

Shear force did not correlate with muscle fibre traits (Wegner et al., 2000) of Holstein-

Friesian, Belgian Blue, German Angus and Galloway which have very different growth

potential.

Muscles with a large fibre size, especially glycolytic type IIB fibres, exhibit tougher

meat than muscles with a narrower fibre diameter in cattle (Renand et al., 2001) and in

pigs (Karlsson et al., 1993). Thus shear force is likely to be reduced as the proportion of

smaller oxidative fibres increases.

Considering the relationship between fibre type and intramuscular fat, as discussed in

Section 2.6.1.3., it follows that as intramuscular fat increases and the oxidative capacity

of muscle increases, the muscle fibre diameter will decrease and shear force is likely to

reduce. This aligns with the findings of Calkins et al. (1981) who showed that muscle

fibre type composition has a stronger relationship with marbling than with shear force.

47

2.6.3 Overall liking

Overall liking, also called overall acceptability, is a subjective measure of meat quality

based on consumer sensory testing. Sensory enjoyment strongly influences the demand

for lamb in Australia and the willingness to pay of consumers (Pethick et al., 2006a).

Untrained consumers assess lamb steaks for overall liking out of 100, with 100 being

the most preferred. Untrained consumer taste panels are used as they are an accurate

reflection of the actual consumer base, despite their greater variance (Hwang et al.,

2008).

Overall liking is highly correlated with tenderness, juiciness and flavour (Pannier et al.,

2014a) and is considered a good subjective measurement of meat quality. Therefore,

factors which influence overall liking, such as intramuscular fat (Hopkins et al., 2006;

Pannier et al., 2014a), will influence consumer acceptability of meat products.

2.6.3.1 Overall liking and growth

The impact of growth on overall liking has not specifically been studied however

relationships between overall liking and sensory traits which have an association with

overall liking, such as juiciness and tenderness, can give an indication of the impact of

growth on overall liking. Unfortunately the relationship between other sensory traits and

growth is also unclear.

For example, juiciness, which is highly correlated to overall liking (Pannier et al.,

2014a), was not affected by growth rate in cattle (Fishell et al., 1985) while tenderness,

which is also highly correlated to overall liking in lambs (Pannier et al., 2014a), was

increased in cattle with higher growth rates (Fishell et al., 1985).

48

Furthermore, there is no clear relationship between selection for growth using the

PWWT ASBV and overall liking. Selection for growth reduced overall liking at only

one of the eight research sites in the Pannier et al. (2014a) study.

Selection for increased muscle using the PEMD ASBV has been found to reduce overall

liking. This effect is independent of the phenotypic impact of intramuscular fat on

overall liking (Pannier et al., 2014a). Therefore, selection for increased PEMD may

negatively influence eating quality for consumers.

2.6.3.2 Production factors which influence overall liking

Overall liking varies with production factors including age, nutrition, sex, and genotype

with older sheep having reduced sensory scores including overall liking (Hopkins et al.,

2006; Pannier et al., 2014a; Pethick et al., 2005c).

Nutrition treatments may impact on the sensory traits such as overall liking. Recent

work by Resconi et al. (2009) found that meat from lambs fed a concentrate only diet

had the highest overall liking as rated by a trained consumer taste panel when compared

to pasture fed lambs or lambs fed a combination of concentrate and pasture and alfalfa

hay. In contrast, there was no difference in overall liking in studies where untrained

consumers rated lamb which was fed different rations (Pethick et al., 2005a). Pannier et

al. (2014a) found that overall liking varied between research sites at which lambs were

raised and the year of birth, reflecting the impact that environment and nutrition has on

consumer sensory scores.

Sex has either no influence (Tejeda et al., 2008) or a small, variable, influence on

sensory scores in lambs. In the Teixeira et al. (2005) study males had an overall liking

score less than one unit higher than female lambs while in the (Pannier et al., 2014a)

study overall liking was higher in female lambs than male lambs by 1.5 units. However,

49

this difference can in part be attributed to the increased amount of intramuscular fat,

0.15%, in the loin of female lambs compared to wether lambs (Pannier et al., 2014c).

There is a variable impact of genotype on overall acceptability of lamb meat reported in

the literature. A study based on 60 lambs from Texel, Poll Dorset, Border Leicester and

Merino sires found no effect of genotype (Safari et al., 2001). Conversely, in a much

larger study based on 210 animals, Hopkins et al. (2005) found meat from Merino

lambs generally have lower sensory scores than meat from Terminal first cross lambs.

Dam breed however appears to have the opposite effect to sire breed with overall liking

higher in lambs from Merino dams when compared to meat from Border Leicester x

Merino dams (Pannier et al., 2014a).

While overall liking reduced with age it varied with nutrition, sex and genotype. These

conflicting findings highlight the need for further work to better describe and quantify

the effect of genotype on overall liking and the mechanisms through which this effect is

mediated.

2.6.3.3 Overall liking and fibre type

There are no studies which directly link overall liking and fibre type. Nonetheless, the

positive relationship between intramuscular fat and overall liking (Pannier et al., 2014a)

can be used to infer a positive relationship between overall liking and the proportion of

oxidative fibre types. As intramuscular fat is associated with increased oxidative fibre

types (Hocquette et al., 2003), factors which cause a reduction in oxidative fibre types

are likely to reduce intramuscular fat and thus overall liking.

50

2.6.4 Isocitrate dehydrogenase activity

Isocitrate dehydrogenase (ICDH; EC 1.1.1.42) is a crucial enzyme in the oxygen-

dependant citric acid cycle of mitochondria, which are larger and more abundant in

oxidative muscle fibres (Hoppeler, 1985). Due to its high correlation with oxidative

metabolism, ICDH activity is used as a marker of oxidative capacity and thus fibre type

(Gardner et al., 2006b).

2.6.4.1 Isocitrate dehydrogenase activity and fibre type

Muscle oxidative capacity is intrinsically linked to fibre type (Brandstetter et al., 1998a;

Staron and Johnson, 1993) with type I fibres having a greater expression of oxidative

enzymes such as ICDH (Brandstetter et al., 1998a; Peter et al., 1972). This is in contrast

to type II fibres which express higher levels of glycolytic enzymes (Brandstetter et al.,

1998a) and reduced levels of oxidative enzymes (Gardner et al., 2007; Gardner et al.,

2006b).

In bulls the transition between fibre types is marked by changes in metabolic

metabolism. As the proportion of oxidative fibres reduce and glycolytic fibres increase

before puberty in cattle, ICDH reduces and the activity of glycolytic enzyme lactate

dehydrogenase increases (Schreurs et al., 2008). This reflects the use of glycolytic

metabolism to supply energy during periods of rapid growth before puberty

(Brandstetter et al., 1998a).

The high correlation between ICDH activity and TAG (Jurie et al., 2007) aligns with

the use of fatty acids as an energy source for oxidative metabolism (Ashmore, 1974).

Considering this relationship, it follows that as intramuscular fat increases, ICDH

activity is likely to increase. This aligns with the findings of Jurie et al. (2007) who

51

showed that ICDH activity is increased in genotypes of cattle which deposit greater

amounts of intramuscular fat.

2.6.4.2 Isocitrate dehydrogenase activity and growth

There is a limited amount of information on how growth impacts muscle ICDH activity

in sheep but work in rabbits and rats has demonstrated higher ICDH activity with

increased growth rates (Pascual and Pla, 2007). This aligns with the findings in sheep

where growth is restricted due to nutritional restriction, resulting in reduced ICDH

activity (Gardner et al., 2006b). These associations may be affected by maturity with no

difference in ICDH Activity in rabbits selected for growth when compared at a similar

degree of maturity (Hernández et al., 2004).

In ruminants the existing evidence does not give a clear association between ICDH

activity and growth however it may be inferred from the association between ICDH

activity and muscle fibre type. ICDH expression is higher with type I muscle fibres

(Brandstetter et al., 1998a) and these fibres increase as animals mature (Brandstetter et

al., 1998a; Greenwood and Cafe, 2007). Therefore, a fast growing animal tending

toward a larger mature size is likely to be less mature at a given age when compared to a

slower growing animal tending toward a small mature size, which is likely to result in

reduced ICDH activity.

Selection for increased muscle in sheep increases the expression of type IIX fibres

(Wegner et al., 2000) resulting in a reduced expression of ICDH activity (Gardner et al.,

2006b).

2.6.4.3 Production factors which influence isocitrate dehydrogenase activity

ICDH activity varies with production factors including age, sex, and genotype. ICDH

activity increases with age (Brandstetter et al., 1998a; Gardner et al., 2007; Jurie et al.,

52

2005) and maturity (Schreurs et al., 2008), most likely due to increased proportions of

type I muscle fibres (Brandstetter et al., 1998a).

There is a small variation in ICDH activity between sexes with activity being 2% higher

in wethers than ewes (Gardner et al., 2007). Castration has also been found to impact in

cattle, with bulls having higher ICDH activity than steers, which aligns with the

increase in oxidative type I muscle fibres (Brandstetter et al., 1998b). Bulls also have a

reduced lipid concentration and which is likely to be due to the use of fatty acids as an

energy source for oxidative metabolism (Ashmore, 1974).

Genotype impacted significantly on ICDH activity with Border Leicester x Merino

sheep having up to 30% lower ICDH activity than Poll Dorset x Merino sheep

depending on the age at which the comparison was made from 8 to 22 months of age

(Gardner et al., 2007). Genotypes which express higher levels of intramuscular fat have

increased ICDH activity in sheep (Pethick et al., 2005c), cattle (Hocquette et al., 2003;

Jurie et al., 2007; Pethick et al., 2005c), pigs (Essén-Gustavsson et al., 1994) and

rabbits (Gondret et al., 1998). Selection for leanness in lambs results in a reduction in

intramuscular fat (Pannier et al., 2014c) therefore a concurrent reduction in ICDH

activity (Gardner et al., 2007; Gardner et al., 2006b) was to be expected.

2.6.5 Myoglobin concentration

Myoglobin is the heme protein primarily responsible for the colour of meat (Fox and

Jay, 1966; Kim et al., 2010; Livingston and Brown, 1981). Meat colour is the most

important attribute influencing purchase decisions (Mancini and Hunt, 2005), with

deviations from the bright cherry red colour of meat responsible for product rejection

(Suman and Joseph, 2013). Therefore myoglobin concentration is an important

consumer trait. Myoglobin is the primary oxygen-carrying pigment of muscle tissues

53

supplying oxygen to the mitochondria within muscle fibres (Wittenberg and Wittenberg,

2007). Other heme proteins, including haemoglobin (Rickansrud and Henrickson, 1967)

and cytochrome-c (Girard et al., 1990) are present but unlike myoglobin have little

effect on the colour of meat (Ledward, 1992).

There are four major chemical forms of myoglobin which are responsible for meat

colour, oxymyoglobin, deoxymyoglobin, metmyoglobin and carboxymyoglobin

(Giddings and Solberg, 1977; Mancini and Hunt, 2005) with the relative proportions of

these forms of myoglobin determining the colour of fresh meat.

Oxymyoglobin has the characteristic cherry red colour that is seen when myoglobin is

exposed to oxygen in the atmosphere. Oxygen is required to maintain this state.

Deoxymyoglobin is the deoxygenated state of myoglobin with a purplish red colour

typically associated with muscle immediately after cutting. Low oxygen tension is

required to maintain this state. Metmyoglobin is brown in colour and is formed when

the ferrous iron in either oxymyoglobin or deoxymyoglobin oxidises to ferric iron

(Livingston and Brown, 1981; Wallace et al., 1982). Carboxymyoglobin is formed

when deoxymyoglobin binds with carbon monoxide forming a bright red colour. This

commonly occurs during packaging with low levels of carbon monoxide (Hunt et al.,

2004; Luno et al., 2000; Sørheim et al., 2001).

The chemical reactions of oxygenation, oxidation, reduction and

carboxymyoglobination cause myoglobin to move from one chemical form to another

(Figure 2-7). During oxygenation, myoglobin is exposed to oxygen causing

deoxymyoglobin to convert to oxymyoglobin (Reaction 1).

Over time, as the oxygen within meat is utilised, oxymyoglobin is deoxygenated back

into the less stable deoxymyoglobin, which then becomes oxidised to form

54

metmyoglobin with oxidative triggers such as low oxygen partial pressure, temperature

and pH (Mancini and Hunt, 2005) (Reaction 3). This is a two-step process as

oxymyoglobin does not convert to metmyoglobin directly (Reaction 2). Enzymatic and

non-enzymatic processes can reduce ferric metmyoglobin back into ferrous

deoxymyoglobin (Reaction 4) with oxygen consumption, metmyoglobin reducing

activity and nicotinamide adenine dinucleotide (NADH) availability significantly

influencing the reaction.

Carboxymyoglobination occurs when carbon monoxide binds to deoxymyoglobin

(Reaction 5). The chemistry of this reaction is not completely understood however

deoxymyoglobin is more readily converted to carboxymyoglobin than oxymyoglobin or

metmyoglobin are. When carboxymyoglobin is exposed to atmospheres free of carbon

monoxide the carbon monoxide will slowly dissociate from myoglobin (Mancini and

Hunt, 2005).

Figure 2-7 Interconversions of myoglobin redox forms in fresh meat. Source: AMSA Color Measurement

Guidelines (AMSA, 2012)

55

2.6.5.1 Myoglobin concentration and fibre type

Myoglobin concentration is linked to muscle fibre type with type I muscles having a

greater concentration of myoglobin (Pethick et al., 2005c). Therefore, along with ICDH

activity, myoglobin concentration can be used as an indicator of oxidative metabolism.

Furthermore, factors which influence oxidative capacity are likely to influence meat

traits linked to muscle fibre type.

The increase in myoglobin concentration with age (Gardner et al., 2007; Pethick et al.,

2005c) is in line with a change to an increasing proportion of type I muscle fibres and a

more oxidative muscle metabolism (Brandstetter et al., 1998a; Greenwood et al., 2007;

Suzuki and Cassens, 1983; White et al., 1978). As expected, higher levels of glycolytic

type IIX fibres are associated with a reduction in myoglobin concentration (Gardner et

al., 2007).

2.6.5.2 Myoglobin concentration and growth

Myoglobin concentration is reduced in slower growing breeds of lamb which are less

mature at a given age, such as Merinos (Gardner et al., 2007) and has been shown to be

higher in fast growing lamb which are likely to be closer to their mature size (Gardner et

al., 2007).

Nutrition has been shown to impact muscle fibre type (Greenwood et al., 2006b; Moody

et al., 1980), which may subsequently alter oxidative capacity including myoglobin

concentration. Nutritional restriction in lambs, which in turn will restrict growth rates,

has been shown to reduce myoglobin concentration with a magnitude of effect four

times greater than as genotypic effects (Gardner et al., 2006b).

Selection for increased muscle using ASBVs for increased PEMD has been shown to

increase the proportion of type IIX glycolytic fibres (Greenwood et al., 2006c; Wegner

56

et al., 2000) and reduce myoglobin concentration (Gardner et al., 2006b). However,

while selection for leanness using ASBVs for reduced PFAT also increased the

proportions of type IIX glycolytic fibres, an associated reduction in myoglobin

concentration was expected but not seen (Gardner et al., 2006b).

2.6.5.3 Production factors which influence myoglobin concentration

Myoglobin concentration is influenced by production factors including age, sex and

genotype. Age is a strong driver of myoglobin concentration in sheep with myoglobin

doubling in all muscles between 4 and 22 months of age in the Gardner et al. (2007)

study. In this study the greatest increase in myoglobin occurred between 4 and 8 months

of age. The increase in myoglobin with age is a consistent finding in sheep (Ledward

and Shorthose, 1971; Pethick et al., 2005b) with Jacob et al. (2007) demonstrating an

increase between 5 and 12 months of age. As with age, myoglobin concentration

increases as lambs approach maturity (Gardner et al., 2007). As higher myoglobin

concentrations are associated with increased redness in meat, at maturity animals have

redder meat (Hopkins et al., 2007).

The influence of sex on myoglobin concentration is not straightforward. Although

wether lambs have been shown to have reduced proportions of type IIX fibres than ewe

lambs (Greenwood et al., 2007) and thus greater oxidative metabolism, Ledward and

Shorthose (1971) showed that ewe lambs had 10% more myoglobin than wether lambs.

In contrast to these findings, and to the hypothesis based on fibre type, Gardner et al.

(2007) found no impact of age on myoglobin concentration (P > 0.05).

The impact of genotype on myoglobin concentration varies, with studies showing

reduced myoglobin concentration in slow growing Merinos compared to faster growing

breeds (Gardner et al., 2007; Gardner et al., 1999) while Hopkins et al. (2005) showed

no difference between Merinos and other genotypes.

57

The inconsistent variation between myoglobin concentration with sexes and genotypes

indicates that changes in fibre type with lamb growth do not explain the entire variation

in myoglobin concentration and further studies are needed to explain the associations.

2.6.6 Iron and zinc concentration

Iron and zinc are key nutrients for human health and consumers assess meat quality by

reliance quality traits which include nutrition (Pethick et al., 2006a; Troy and Kerry,

2010). Red meat is recognised as one of the best sources of iron and zinc because of

their high bioavailability in meat compared to plant sources (Linder, 1991). Australian

lamb is currently marketed as a good source of both minerals (Pannier et al., 2014b;

Pannier et al., 2010).

Iron and zinc concentration are higher in muscles with greater proportions of oxidative

fibres (Cassens et al., 1967) and have a positive association with intramuscular fat.

Therefore, factors which impact oxidative capacity may in turn influence iron and zinc

concentration in meat.

2.6.6.1 Iron and zinc concentration and growth

Iron and zinc concentration are influenced by growth with higher levels in lamb

(Pannier et al., 2014b) and cattle (Kirchgessner et al., 1994) with heavier hot carcass

weights. While a similar finding of increased zinc concentration with increased weight

was found in the Bellof et al. (2007) study in lambs, there was no increase in iron

concentration. The Bellof et al. (2007) study was based on the averaged iron and zinc

concentrations from the total muscle tissue however Lin et al. (1988) found that when

examining retail lamb cuts iron concentration was highest in the loin. The use of total

muscle tissue may explain the lack of alignment with mineral changes found in the loin

in the Bellof et al. (2007) study. In direct contrast to the findings above, Lin et al.

58

(1988) found a reduction in iron and zinc concentration with increasing carcass weight.

However, the magnitude of the effect was small with iron and zinc concentrations

influenced more by age than weight.

Selection for increased growth via the PWWT ASBV was expected to reduce iron and

zinc concentration in the (Pannier et al., 2014b) study however there was no

association. The hypothesis was based on progeny of high PWWT sires having an

increased mature size (Huisman and Brown, 2008) and therefore being less mature at

the same slaughter age. While age has a large impact on both iron and zinc

concentration (Pannier et al., 2014b) the variation in maturity at a given age with

differing PWWT may be too small to impact iron and zinc concentration.

2.6.6.2 Production factors which influence iron and zinc concentration

Iron and zinc concentrations vary with production factors including age, sex, and

genotype. Iron and zinc concentrations increase with age in lambs (Lin et al., 1988; Ono

et al., 1984; Pannier et al., 2010) while iron concentration increases with age in cattle

(Doornenbal and Murray, 1982; Kotula and Lusby, 1982) and zinc concentration

increases with age in pigs (Cassens et al., 1967).

Iron concentration was higher in ram lambs compared to ewe lambs in the Bellof et al.

(2007) study of 108 lambs. However, this association was only present in ram and ewe

lambs up to 30 kg body weight, after which there was no difference in iron

concentration between ram and ewe lambs. In the same study, zinc concentration was

also higher in ram lambs than ewe lambs, which aligns with greater proportion of

glycolytic type IIX fibres in ewe lambs (Greenwood et al., 2007). In contrast, iron

concentration was higher in ewe lambs compared to wether lambs in both the (Pannier

et al., 2014b) and (Pearce et al., 2009) studies, although the reason for this is unknown.

59

Animal breed can impact mineral levels (Underwood, 2012) and iron and zinc

concentration have been found to vary between some sire types. Pannier et al. (2014b)

found that Terminal sired lambs had the lowest iron concentration while it was

intermediate in Maternal sired lambs and highest in Merino sired lambs. Terminal sired

lambs have more glycolytic type IIX fibres (Greenwood et al., 2007) which may reflect

their lower iron levels. There was no difference in zinc concentration between lambs

from either Terminal or Merino sire types however lambs from Maternal sires had the

highest zinc concentrations (Pannier et al., 2014b). As there is no general alignment in

iron and zinc concentration of progeny of different sire types and between the sexes it is

unlikely that these effects are mediated by growth alone.

2.6.6.3 Iron and zinc concentration and fibre type

Iron and zinc concentrations are higher in muscles with greater oxidative, type I fibres

(Beecher et al., 1968; Cassens and Cooper, 1971; Cassens et al., 1967). Thus muscles

with reduced oxidative capacity have been shown to contain reduced iron and zinc

concentrations (Kondo et al., 1991; Pearce et al., 2009). Factors contributing to the

reduction of these minerals are likely to include lower concentrations of myoglobin, less

vascularisation (Choi and Kim, 2009; Lefaucheur, 2010) and fewer mitochondria

(Hoppeler, 1985).

Iron and zinc concentration increases with age (Pannier et al., 2014b) as does oxidative

fibre type (Brandstetter et al., 1998a). The strong association between iron and zinc

concentration and oxidative fibre type is due to the function of these minerals in

oxidative fibres. Iron is a component of myoglobin (Hazell, 1982) and zinc is associated

with antioxidant cascades and the electron transport chain (Powell, 2000) both of which

are also associated with more oxidative fibre types.

60

Given that ICDH activity and myoglobin concentration are indicators of oxidative

capacity, they have a positive association with iron and zinc concentration (Pannier et

al., 2014b). The magnitude of the association was greater for iron, which increased two-

fold more than zinc across the same range of ICDH activity and four-fold more across

the same range of myoglobin concentration, indicating a stronger alignment of iron

concentration with oxidative capacity. This was supported by stronger phenotypic and

genotypic correlations between ICDH activity and myoglobin concentration with iron

than with zinc (Mortimer et al., 2014).

Selection for leanness via a reduced PFAT ASBV has been shown to increase the

proportion of type IIX muscle fibres (Greenwood et al., 2006c), which results in

reduced oxidative capacity indicated by myoglobin levels (Gardner et al., 2006b) as

well as a reduction in iron levels (Pannier et al., 2014b). No association has been found

between selection for a reduced PFAT and zinc concentration (Pannier et al., 2014b).

Selection for an increase in muscle using the PEMD ASBV reduces the proportion of

oxidative muscle fibres (Greenwood et al., 2006c), which aligns with a reduced

myoglobin concentration in the Gardner et al. (2006b) study. Despite the strong

association between myoglobin concentration and iron concentration, Pannier et al.

(2014b) found no subsequent reduction in iron or zinc concentration.

As iron and zinc are highly linked to muscle fibre type, factors which influence muscle

oxidative capacity and fibre type are likely to impact iron and zinc concentrations. The

association between iron concentration and fibre type is stronger, likely to be due to the

direct involvement of iron as a key component of myoglobin, so impacts to fibre type

may be more strongly reflected in iron concentration than zinc concentration.

61

2.6.7 Meat browning

Lamb meat develops a desirable bright red colour after exposure to oxygen allows

myoglobin pigments to transform from the purple pigment deoxymyoglobin into the red

pigment oxymyoglobin. Meat browning then occurs over time on retail display as

oxymyoglobin auto-oxidises to form the brown pigment metmyoglobin

Consumers are able to detect this change in meat colour (Khliji et al., 2010) and

associate browning with a lack of freshness and quality (Faustman and Cassens, 1990)

leading to product rejection and revenue loss (Mancini and Hunt, 2005; Suman and

Joseph, 2013). Retail shelf life for lamb meat in Australia is currently 2 to 3 days

(Calnan et al., 2014; Jacob et al., 2007).

Price discounts due to meat browning in beef are estimated to total more than one

billion dollars in the United States each year (Smith et al., 2000) with an average loss of

meat sales due to colour deterioration estimated to be 3.7% (Williams et al., 1992). In

Australia this would equate to a loss of $353 million based on a slaughter value of

$9,541 million for beef and lamb in the 2009-2010 financial year. While the importance

of meat quality traits varies by country (Warner et al., 2010) a consumer survey carried

out in six European countries identified meat colour at retail display as the most

important intrinsic quality when purchasing beef, pork or chicken (Glitsch, 2000) and in

a Japanese survey 58% of participants identified meat colour as the most important

factor when selecting beef products (Sanders et al., 1997).

Meat colour can be measured using light wavelength reflectance with the ratio of

reflectance at 630 mm and 580 mm (R630/R580) representing the redness:brownness of

the meat surface. This measurement is also known as the oxy/met ratio (Hunt et al.,

2001). The ratio approaches a minimum value after 3 days of simulated retail display in

62

lamb meat (Jacob et al., 2014). The ratio needs to exceed 3.3 units for the average

consumer to deem lamb meat an acceptable red colour (Khliji et al., 2010).

Reducing the rate of meat browning relies on an understanding of oxygen diffusion and

consumption and the rate of auto-oxidation of myoglobin pigments into metmyoglobin

(Jose, 2011). The rate of metmyoglobin reduction may also be important (Faustman

and Cassens, 1990) however its importance remains unclear (Bekhit and Faustman,

2005). In cattle Bekhit et al. (2003) found that NADH, which is integral in

metmyoglobin reduction, was more important in governing meat colour than

metmyoglobin reducing activity. In sheep the importance of the reduction process has

not yet been established as studies in lamb loin have found no correlation between

metmyoglobin reducing ability and meat colour (Bekhit et al., 2001).

Strategies to reduce meat browning include packaging modification and the use of

dietary anti-oxidants supplementation (Mancini and Hunt, 2005). The antioxidant

Vitamin E has been used successfully in the finishing diet of animals to reducing the

rate of browning in beef (Faustman et al., 1989) and lamb (Strohecker et al., 1997).

2.6.7.1 Meat browning and fibre type

Meat browning is linked to fibre type as numerous studies have shown that muscle

oxidative capacity impacts the rate of meat browning (Calnan et al., 2014; Jeong et al.,

2009). More oxidative muscles are prone to more rapid meat browning than more

glycolytic muscles (O'keeffe and Hood, 1982; Renerre and Labas, 1987).

Oxidative, type I, muscle fibres are associated with greater concentrations of

mitochondria (Klont et al., 1998). Increased mitochondrial activity leads to increased

production of free radicals that trigger myoglobin oxidation. Thus, muscle fibre types

63

with high numbers of mitochondria, such as type I muscle fibres, can be detrimental and

accelerate meat browning (Sammel et al., 2002). .

Increased meat browning is also associated with increased intramuscular fat. The

reactive products of lipid oxidation accelerate the oxidation of myoglobin (Buckley et

al., 1995; Faustman et al., 1999) and thus browning. Therefore, strategies to inhibit

lipid oxidation also reduce meat browning (Faustman et al., 2010). Furthermore,

intramuscular fat is increased with oxidative fibre types (Hocquette et al., 2003; Jurie et

al., 2007) which also increase meat browning.

2.6.7.2 Meat browning and growth

Meat browning is influenced by weight and stage of maturity but not by genetic

selection for growth. In Australia the maturity of lambs at slaughter has been reduced

(Hall, 2000) by selection for sires with high post weaning weight progeny leading to

increased growth rates (Hall et al., 2002). Reduced maturity is associated with reduced

muscle oxidative capacity (Brandstetter et al., 1998a) thus the rate of meat browning

may be reduced with selection for growth.

Calnan et al. (2014) used hot carcass weight, as a proxy for whole of life growth rate, to

demonstrate an association between increased growth and a reduced rate of meat

browning. This association did not change with correction for the oxidative enzyme

ICDH activity. This indicated that the impact of phenotypic growth rate on meat

browning was not delivered through an altered oxidative capacity.

Although increased growth caused a reduction in meat browning, selection for increased

growth using the PWWT ASBV did not have any further impact on meat browning

(Calnan et al., 2014). Overall, the impact of growth on meat browning appears to be

independent of growth.

64

2.6.7.3 Production factors which influence meat browning

Environmental and production factors such as diet (Faustman and Cassens, 1990) and

genetics (King et al., 2010; Mortimer et al., 2010) can have an impact on meat

browning. Diet and genetics can increase growth and it is unknown based on current

literature whether the impact of these production factors are simply proxies for the

impact of growth on meat browning.

Calnan et al. (2014) reported an reduction in browning in meat from lambs born to

Terminal and Maternal sires compared to meat from progeny of Merino sires. Lambs

born to Terminal and Maternal sires have a greater potential for increased muscling,

which increases the expression of type IIX glycolytic muscle fibres, reducing muscle

oxidative capacity (Greenwood et al., 2006c; Greenwood et al., 2007) and reducing

meat browning.

While the proportion of type IIX fibres increase with selection for muscling (Wegner et

al., 2000), the proportion of type I oxidative fibres increase with age (Warner et al.,

2007) and maturity (Suzuki and Cassens, 1983; White et al., 1978). The increase in

oxidative fibres with age and maturity leads to increased browning (Renerre, 1990).

Consumers relate browning to meat from mature animals (Jeyamkondan et al., 2000).

While browning is generally more rapid in mature animals due to increased proportions

of oxidative fibres, meat browning can occur in meat from animals at any stage of

maturity.

Several authors have investigated the link between myoglobin concentration and meat

browning given that browning is caused by myoglobin oxidation. However, findings

have been mixed. O'keeffe and Hood (1982) speculated that muscles containing lower

myoglobin would brown faster because their myoglobin oxidised at a greater frequency.

65

However, this was not supported by the work of King et al. (2010) who found that

breeds with reduced myoglobin concentration had the reduced browning or the work of

McKenna et al. (2005) who found no relationship between myoglobin content and meat

colour across 19 muscles in beef.

There are a number of additional intrinsic and extrinsic factors beyond the scope of this

thesis that impact on meat browning including pH, rate of pH decline, temperature and

light, presence of antioxidants, lipid oxidation, bacterial contamination and packaging

(Lawrie, 1983; Mancini and Hunt, 2005; Suman et al., 2014; Warner et al., 2007).

Variation in the response of meat browning to changes in growth rate or with muscle

fibre type may be due in part to some of these factors.

2.7 Hypothesis

This thesis explores the effect of selection for increased growth in lambs on muscle

oxidative capacity and meat quality. The impact of weight and growth rate at key times

during production are investigated. These aspects are covered in the following chapters;

the hypotheses of which are outlined below:

Chapter 3: Comparison of linear and nonlinear functions to estimate lamb growth

The aim of this chapter is to determine the most appropriate method of fitting a growth

curve to lambs, based on residual weight values, to determine reliable estimates of

weight and growth rate at key time points during production.

The specific hypotheses tested were:

The mean squared error would be lowest for a random effects model curve fit rather

than a biologically based Brody curve fit or a cubic polynomial curve fit.

66

Chapter 4: Lamb birth and rearing type influence growth and weight potential of

progeny from high growth sires

The aim of this chapter was to determine the impact of nutritional restriction on growth

of lambs from high growth sires. In addition to environmental variation, sources of

nutritional restriction included both dam and lamb factors with reduced milk production

in some dam breeds and sibling competition in-utero and pre weaning. The impact of

these forms of nutritional restriction on progeny of high post weaning weight sires

achieving expected weights was also investigated. The impact of breeding values for

muscling and leanness on lamb weight were also investigated.

The specific hypotheses tested were:

The magnitude of the sire PWWT effect would vary between sites as nutrition

varies between sites

Lambs born and raised as multiples would have a reduced response to increased

sire PWWT

Lambs born to Merino dams would have a reduced response to increased sire

PWWT

Lambs from high PEMD sires and lambs from low PFAT sires would be heavier

Chapter 5: Selection for lean meat yield in lambs reduces indicators of oxidative

metabolism in the longissimus muscle

The aim of this chapter is to evaluate the impact of selection for lean meat yield on

muscle oxidative capacity. The activity of the oxidative enzyme isocitrate

dehydrogenase and the concentration of the red pigment of muscle myoglobin were

used as indicators of oxidative capacity. The impact of production factors on muscle

oxidative capacity was also investigated.

67

The specific hypotheses tested were:

Selection for leanness via increasing sire PEMD ASBVs and reducing sire PFAT

ASBVs would reduce oxidative capacity in the muscle of their progeny

Selection for growth via an increased sire PWWT ASBV would reduce oxidative

capacity in the muscle of their progeny

Muscle oxidative capacity would vary between sites

The magnitude of the effect of production factors on muscle oxidative capacity

would be greater than the magnitude of the genotypic effects

Chapter 6: Lamb growth increases intramuscular fat levels and reduces shear force in

the longissimus muscle

The aim of this chapter was to evaluate the impact of weight and growth rate on meat

quality traits of intramuscular fat, shear force and consumer overall liking. These traits

were selected due to the direct link with palatability and consumer acceptance of lamb

meat. Hot carcass weight is commonly used as a proxy for lamb growth however this

whole of life growth rate has not been verified as a good measure of lamb growth.

Furthermore, it cannot distinguish the impact of growth at important time points during

production such as weaning and post weaning. The impact of weight and growth rate on

these traits, at weaning and post weaning time points, is compared to the impact of hot

carcass weight on the same traits.

The specific hypotheses tested were:

Intramuscular fat and overall liking would increase with lamb growth

Shear force would reduce with lamb growth

Hot carcass weight would be a poor proxy for lamb growth

68

Chapter 7: Lamb growth increases oxidative capacity and mineral content in the

longissimus muscle

The aim of this chapter was to evaluate the impact of weight and growth rate on meat

quality traits related to oxidative capacity, mineral content and meat browning.

Specifically, isocitrate dehydrogenase activity and myoglobin concentration were

investigated as markers of oxidative capacity and iron and zinc concentration were the

minerals investigated. These meat traits were selected due to their strong links with

muscle fibre type and the impact that changes in fibre type can have on the nutritive

value and retail shelf life of lamb meat. For the reasons previously mentioned, these

impact of weight and growth rate on these traits is compared to the impact of hot carcass

weight on these traits, as in Chapter 6.

Oxidative capacity would increase with lamb growth

Iron and zinc concentration would increase with lamb growth

Lamb meat would brown more rapidly with lamb growth

Hot carcass weight would be a poor proxy for lamb growth

69

Chapter 3. Comparison of linear and nonlinear

functions to estimate lamb growth

3.1 Introduction

Lamb growth is modelled as a function of weight change between birth and maturity. It

follows a sigmoidal pattern with the greatest growth rate after birth in the exponential

phase, a declining growth rate as the animal nears its mature weight in the transitional

phase and a stable mature weight in the plateau phase (Hendrick et al., 1989). As

incomplete weight data recordings are common in industry utilisation of the data that is

available is essential. This requires the fitting of a growth curve to available weight data

so that weight and growth rate can be estimated at specific times of interest.

A variety of growth functions exist. They can be broadly grouped into linear and

nonlinear functions which can be fitted to individual animals or populations of animals.

The Brody function (Brody, 1945) is a nonlinear function which is fitted to the live

weight data of individual animals. It is utilized because of the easy biological

interpretation of the parameters with the shape of the curve being linked to the mature

weight and maximum growth rate of the sheep (Topal et al., 2004). However, the Brody

function tends to overestimate weight during early life growth (Beltrán et al., 1992;

Gbangboche et al., 2008) and gives biased estimates close to the edge of the available

data.

A second group of functions which are fitted to the weight data of individual animals

are polynomial functions. Unlike the Brody function, polynomials are linear functions.

They are attractive as the fit of these functions generally increases with the order of the

70

polynomial, with a third degree cubic polynomial often used to model weight data.

However, there is no biological basis to polynomial function and the asymptote can

cause overestimation of weight close to adulthood.

Random effects models are a population based curve fit which uses a fixed regression to

describe the average shape of a growth curve, and a random regression for each animal

to account for deviations from the fixed regression. With the ability to easily

accommodate missing weight data, overcoming the common difficulty of incomplete

weight data recording in an industry setting, it is a commonly used as a growth function

(Lewis and Brotherstone, 2002). As random effects models are based on a polynomial

function they have similar limitations. However, the ability to incorporate different

durations of data collection results in unbiased estimations of weight than from the

individual based curve fits.

Weight data recorded in an industry setting often differs in lengths with some animals

being weighed frequently for variable lengths of time (i.e. months to years), and others

being weighed less frequently and also for variable lengths of time. While random

effects model methodology is not biased by these differences in data availability,

functions such as the Brody’s are heavily influenced. Therefore, when functions are

fitted on an individual animal basis those animals with data collected for shorter periods

are likely to be somewhat biased for weights estimated at the edge of their available

data compared to those with weights collected for periods that extend well beyond this

point.

Assessment of residual weights (the difference between the estimated weight and the

recorded weight) after a growth function has been fitted is a useful way of assessing

curve fit. Specifically, the mean square error, which is the average value of the squared

residual weights, is commonly used (Topal et al., 2004). It is a useful measure of fit as it

71

can be compared across both linear and nonlinear curves and is not biased by the

number of weight recordings each animals has. As the fit of different functions may

vary with time, the residual weights can also be assessed during specific times of

interest for comparison between curves.

As linear and nonlinear functions are known to have limitations when estimating growth

and random effects models can accommodate missing weight data and different

durations of data collection, it is hypothesised that the fit of the random effects model

will have a lower mean square error.

3.2 Materials and Methods

3.2.1 Experimental Design

The Information Nucleus Flock experiment, from which lamb weight data was sourced,

was conducted at eight sites across Australia, representing a broad range of production

systems (Kirby NSW, Trangie NSW, Cowra NSW, Rutherglen VIC, Hamilton VIC,

Struan SA, Turretfield SA, and Katanning WA). These sites were a component of the

Sheep CRC’s Information Nucleus Flock the details of which were presented by

Fogarty et al. (2007) and van der Werf et al. (2010). A total of 18,185 lambs from

Maternal (n=3,633), Merino (n=6,895) and Terminal sires (n=7,657) were produced

between 2007 and 2011. Lambs were weighed at birth, at weaning (90 days of age) and

approximately every two weeks thereafter. A large proportion of lambs were

slaughtered at a target weight of 23 kg, as early as 145 days. However, some lambs

were retained and continued to have weights recorded. As the final weight recorded for

most lambs was within 300 days, a decision was made to restrict the weight recordings

72

used to 300 days. Lambs were reared under extensive pasture grazing conditions and fed

grain, hay or feedlot pellets when feed supply was limited (Ponnampalam et al., 2014b).

3.3 Statistical Analysis

3.3.1 Data available

Each of the 18,185 lambs had between 1 and 16 records, with an average of 8 records

per lamb. This resulted in 142,572 weight recordings. 18,139 lambs had weight

recordings on the day of birth. The distribution of weight recordings available between

day 1 and day 300 is presented in Figure 3-1. The Brody fit required at least three

weight recordings and the cubit fit required at least four weight recordings. For

comparison, all three functions were fitted to weight data for animals with four or more

live weight recordings. Lambs weights were incorrectly measured or recorded at the

Turretfield site during a 14 day period in 2011. Four hundred and thirty three weight

recordings from this period at the Turretfield site were removed before curve fitting.

This resulted in all three functions being fitted to 17,382 animals with 141,050 weight

recordings. As random effects models can accommodate animals with as little as one

weight recording, the model was also fitted to the entire population 18,185 lambs with

142,572 weight recordings.

Brody and Legendre cubic polynomial functions were fitted to the weight data of each

individual animal. A cubic Legendre polynomial was also fitted using a Bayesian

random effects model to the entire population.

73

Figure 3-1 Graph of the number of weight recordings against lamb age (in days) between 1 and 300 days.

3.3.2 Nonlinear curve fit

A Brody growth curve (Brody, 1945) was fitted to each individual animals weight. As

estimated values for initial weight and mature weight in the nonlinear growth formulas

are highly correlated (Cullis and McGilchrist, 1990; Lewis et al., 2002) the modified

formula, as derived by Doren et al. (1989) was used:

Wt[t] = A – (A – Wt0)e-kt

Where Wt[t] is the weight at time t, expressed in days, Wt0 is the observed birth weight,

A is the asymptotic limit of the weight when t approaches infinity, e is the base of

natural logarithm, and k is a curve parameter representing the ratio of maximum growth

rate to mature size, referred to as the maturing rate. Values for A were restricted to less

than 100 kg and values for k were restricted to lie between 0 and 1. The Brody curve

0

200

400

600

800

1000

1200

0 50 100 150 200 250 300

Nu

mb

er o

f w

eig

ht

reco

rdin

gs

Age (days)

74

was fitted using nonlinear regression with the NLIN procedure of SAS (SAS Version

9.1, SAS Institute, Cary, NC, USA) where the curve parameters A and k were

determined iteratively using a Gauss-Newton algorithm (Hartley, 1961). The curves of 5

animals did not converge. Outlier weights were identified by large residual values (> 15

kg), leading to the removal of a single weight recording from 13 different animals.

3.3.3 Linear curve fit

A third order orthogonal Legendre polynomial function was fitted to each individual

animals weight data:

Wt[t] = c0 + c1t + c2t2 + c3t

3

Where Wt[t] is the weight at time t, expressed in days, and c is the coefficient. The

cubic curve was fitted using the GLM procedure of SAS. Outlier weights were

identified by large residual values (> 15 kg) and a single weight recording was removed

from 4 animals.

3.3.4 Random effects model fit using Bayesian analysis

A Bayesian random effects model of third order polynomial was fitted to the weight

data of the population in R, with each individual animals curve deviating from the

population curve. Outlier weights were identified by large residual values (> 15 kg) and

a single weight recording was removed from 4 animals.

3.3.5 Curve fitting comparisons

Goodness of fit for each curve was assessed using the mean squared error (MSE), as

previously utilised for nonlinear growth curves in both sheep (Gbangboche et al., 2008;

Topal et al., 2004) and cattle (Brown et al., 1976; Goonewardene et al., 1981). A MSE

75

was calculated for the curve fit between 0 and 300 days. An MSE was also calculated

between 90 and 160 days as this captured the period where weights would be estimated,

at 100 and 150 days, from the most appropriately fitting curve. The coefficient of

determination R2 was not reported as it is an inadequate measure of the goodness of fit

in nonlinear models (Spiess and Neumeyer, 2010) therefore this could not be compared

across all three models.

The residual values from each curve fit (the difference between the estimated weight

and the live weight recording) were assessed for normality, skewedness and kurtosis

using the UNIVARIATE procedure of SAS. As there were greater than 2,000 weight

observations, the Kolmogorov-Smirnov rather than the Shapiro-Wilk test for normality

was used. The Kolmogorov-Smirnov test for normality (D) is based on the empirical

distribution function (Stephens, 1974). Small values of D (<0.05) lead to the rejection

of the null hypothesis that that the residuals are normally distributed (Ayyangar, 2007).

Skewedness was used to characterize whether the distribution of residuals is symmetric

(if skewedness = 0 the distribution is symmetrical). Kurtosis was used to characterize

the ‘peakedness’ of the distribution, where a normal distribution has a value of 0 (in

SAS) and smaller values correspond to thinner tails (less peakedness).

76

3.4 Results

3.4.1 Nonlinear curve fit

3.4.1.1 Mean square error

The correlation between the observed weight and the estimated weight was 0.99 and the

least squares estimates parameters for A was 63.92 ± 0.19 kg and for k was 0.008 ±

0.001. The MSE for the Brody curve fit is presented in Table 3-1 The MSE over the

period of interest (90-160 days) was 16% greater than over the whole range of the data

(0-300 days).

Table 3-1 Mean square error (MSE) ± standard error (kg2) for curve fits between 0 and 300 days and 90

and 160 days for Brody, Cubic, Random Effects Model and Random Effects Model with all data

(including animals with less than 4 live weight observations) curve fits.

MSE (0-300) MSE (90-160)

Brody 4.15 ± 0.03 4.82 ± 0.04

Cubic 2.18 ± 0.02 2.77 ± 0.03

Random Effects Model 3.07 ± 0.02 3.71 ± 0.04

Random Effects Model (all data) 3.25 ± 0.03 3.85 ± 0.04

3.4.1.2 Residual graph over time

The average residual value between birth and 300 days was -0.003 kg and residual

values varied between -13.95 and 12.66 kg. The average value of residuals for each 10

day increment was plotted against time (Figure 3-2). Within a 10 day time period, the

average positive residual value was as much as 1.65 kg while the average negative

residual value was as much as -1.35 kg. The average residual value with the Brody

curve fit varied from close to zero at birth to more than 1.5 kg close to the edge of the

77

recorded weights at 300 days. Between birth and 300 days the residual value fluctuated

between positive and negative values at five different ages. During early life growth the

Brody model was a less precise estimator of growth, with a large positive residual value

at around 5 weeks of age followed by a large negative value at around 6 weeks of age.

The fit was also less precise later in life, overestimating weight from 200 days onwards.

At weaning (100 days) the residual value was three quarters of a kilogram while post

weaning the residual value was negative by around 1 kg providing evidence of an

overestimation of weight at weaning and an underestimation of weight post weaning to

the 200 day time point.

Figure 3-2 Graph of average residual weight (kg) per 10 day period with a Brody curve fit (-- -- -- --), a

cubic curve fit (- - - -) and a random effects model fit (····). The solid line at residual weight zero

represents the line the residuals would fall on if the random effects model were a perfect fit to the weight

data.

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

0 50 100 150 200 250 300

Res

idu

al

va

lue

(kg

)

Age (days)

Brody CubicRandom Effects Model Zero Residual

78

3.4.1.3 Residual statistics – Normality, Skewedness and Kurtosis

A test of normality for the residual values and values for skewedness and kurtosis are

presented in Table 3-2 and a histogram for the residuals is presented in Figure 3-3. The

test for the location of mu (the mean of the residuals) indicated that the mean was not

significantly different than zero. The test of normality indicated that the distribution of

the residuals was not normal. Although the Kolmogorov-Smirnov test was used over the

Shapiro-Wilk Statistic due to the large sample size, increasingly small departures from

normality are detected as the sample size increases, so the test for normality was

interpreted in conjunction with the results for skewedness and kurtosis. The small

negative skewedness score indicated that the distribution of the residuals was skewed to

the right, indicating that the Brody function overestimated more weights than it

underestimated. The positive kurtosis score indicated that the tails of the distribution

were heavier than for a normal distribution, suggesting that there were more animals

with extreme residual values than expected.

Table 3-2 Results of tests for the location of mu, normality, skewedness and kurtosis of the average

residual values from a Brody function fit to 17,377 animals.

Test Statistic Result P value

Location of mu Students t-test t -0.55 0.58

Normality Kolmogorov-Smirnov D 0.07 <0.01

Skewedness Mardia - -0.18 -

Kurtosis Mardia - 1.75 -

79

Figure 3-3 Histogram of the distribution of residual weight values (kg) against the percent frequency for

the Brody function fit.

3.4.2 Linear curve fit

3.4.2.1 Mean square error

The correlation between the observed weight and the estimated weight was high at 0.99.

The MSE is reported in Table 3-1. The MSE over the period of interest (90-160 days)

was 27% greater than over the whole range of the data (0-300 days).

0

2

4

6

8

10

12

14

16

18

-14-13-12-11-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Fre

qu

ency

(%

)

Residual value (kg)

80

3.4.2.2 Residual graph over time

The average residual value between birth and 300 days was 5.67x10-16

kg and residual

values varied between -12.02 and 13.01 kg. The average value of residuals for each 10

day increment was plotted against time (Figure 3-2). Within a 10 day time period, the

average positive residual value was as much as 0.57 kg while the average negative

residual value was as much as -0.91 kg. Between birth and 300 days the curve fit

oscillated approximately every two months between underestimating and

overestimating weights. Early life weight was underestimated while at around 100

days, when lambs were weaned, weight was overestimated. This is followed by a

similar pattern during the post weaning and maturation phases. Overall, the

underestimation of weights during early and post weaning growth is larger in magnitude

than the overestimation during weaning and maturation.

3.4.2.3 Residual statistics – Normality, Skewedness and Kurtosis

A test of normality for the residual values and the values for skewedness and kurtosis

are presented in Table 3-3 and a histogram of the residuals is presented in Figure 3-4.

The mean of the residuals was not significantly different than zero, and the test of

normality indicated that the distribution of the residuals was not normal. The

distribution of residuals had a small negative skewedness, indicating the distribution

was skewed to the right. Therefore, the cubic function overestimated more weights than

it underestimated. A positive kurtosis indicated that the tails of the distribution are

heavier than for a normal distribution, so there were more animals with extreme residual

values than expected.

81

Table 3-3 Results of test for the location of mu, skewedness, kurtosis and normality of the average

residual values from a cubic function fit to 17,382 animals.

Test Statistic Result P value

Location of mu Students t-test t 0.00 0.99

Normality Kolmogorov-Smirnov D 0.06 <0.01

Skewedness Mardia - -0.15 -

Kurtosis Mardia - 2.66 -

Figure 3-4 Histogram of the distribution of residual weight values (kg) against the percent frequency for

the cubic function fit.

3.4.3 Bayesian random effects model fit

3.4.3.1 Mean square error

The mean square error for the random effects model fit between birth and 300 days, as

well as between 90 and 150 days is presented in Table 3-1. The MSE over the period of

0

2

4

6

8

10

12

14

16

18

20

-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Fre

qu

ency

(%

)

Residual value (kg)

82

interest (90-160 days) was 21% greater than over the whole range of the data (0-300

days). When all live weight data, including animals with less than 4 weight recording,

was used the MSE increased by 5%.

3.4.3.2 Residual graph over time

The average residual value between birth and 300 days was -0.001 kg and residual

values varied between -13.84 and 13.95 kg. The average value of residuals for each 10

day increment was plotted against time (Figure 3-2). Within a 10 day time period, the

average positive residual value was as much as 0.67 kg while the average negative

residual value was as much as -1.63 kg. During early life growth the average residual

value with the random effects model fit was close to zero. At 50 days weight was vastly

underestimated. At weaning (100 days) the residual value was about half a kilogram

while post weaning the residual value was negative by close to 1 kg. This provides

evidence of an overestimation of weight at weaning and an underestimation of weight

post weaning to the 200 day time point. In general, each function overestimated weight

at weaning, and underestimated weight at the post weaning time point of 150 days, the

magnitude of the difference being largest with the Brody fit and similar with the cubic

and random effects model fits.

3.4.3.3 Residual statistics – Normality, Skewedness and Kurtosis

A test of normality for the residual values and values for skewedness and kurtosis are

presented in Table 3-4 and a histogram of the residuals is presented in Figure 3-5. The

mean of the residuals was not significantly different than zero. The test of normality

indicated that the distribution of the residuals was not normal. The distribution of

residuals had a negative skewedness, indicating the distribution was skewed to the right.

Therefore, the random effects model overestimated more weights than it

underestimated. A positive kurtosis indicates that the tails of the distribution were

83

heavier than for a normal distribution, so there were more animals with extreme residual

values than expected.

Table 3-4 Results of tests for the location of mu, skewedness, kurtosis and normality of the average

residual values from a random effects model fit to 17,382 animals.

Test Statistic Result P value

Location of mu Students t-test t -0.13 0.89

Normality Kolmogorov-Smirnov D 0.04 <0.01

Skewedness Mardia - -0.18 -

Kurtosis Mardia - 1.87 -

Figure 3-5 Histogram of the residual weight values (kg) against the percent frequency for the random

effects model fit.

3.4.4 Comparison of residual graphs over time

A graph of the residual values over time for all three curve fits is presented in Figure

3-2. From this graph it can be seen that during early life growth the random effects

0

2

4

6

8

10

12

14

-15

-14

-13

-12

-11

-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

10

11

12

13

14

15

Fre

qu

ency

(%

)

Residual value (kg)

84

model fit was the best estimator of weight while the Brody curve fit was the poorest

with the largest residuals. Between 90 and 160 days all three curve fits were similar,

although the cubic and random effects model fits had slightly lower residual values.

During late life growth the Brody curve fit residual weights were as much as triple the

values for a cubic or random effects model fit.

3.5 Discussion

3.5.1 Mean square error

Contrary to our hypothesis, the MSE of the random effects model was not the lowest of

the three models tested. Instead it was between the MSE of the Brody model, which was

the highest, and the MSE of the cubic polynomial model which was the lowest. The

hypothesis was based on the ability of random effects model to accommodate different

durations of data collection. While the lower MSE for the cubic fit was unexpected it

was likely to be due to the relatively narrow spectrum of live weights over which the

function was fitted. When weights were not restricted to less than 300 days the MSE of

the cubic function increased significantly (data not shown) (MSE = 2.87).

3.5.2 Curve attributes

Linear models such as the cubic polynomial are uncommonly used as growth curves due

to the lack of biological meaning of the parameters and the inherent sigmoidal nature of

lamb growth. Furthermore, the fit of a polynomial will improve with the degree of the

polynomial however the number of live weight recordings per animal to fit the function

also increases. As industry data often varies in length and may not include a large

number of recordings per animal, linear functions such as polynomials are not

85

frequently used. Where a large number of live weight recordings exist, the increased fit

with the higher degree of polynomial will model factors which do not reflect the true

weight of the lamb, such as gut fill.

Meanwhile, nonlinear models are particularly appealing due to the ease of estimation

and simplicity of interpretation (Bathaei and Leroy, 1996; Bathaei and Leroy, 1998).

They can condense information contained in a data series into a few parameters with

biological meaning (Fitzhugh, 1976) and have been used to model growth in both sheep

(Gbangboche et al., 2008; Malhado et al., 2009) and cattle (Kaps et al., 2000).

Various nonlinear equations are available to estimate animal growth including the

biologically based Brody (1945), von Bertalanffy (1957), Richards (1959), Logistic

(Nelder, 1961) and Gompertz (Laird, 1965) models. Moore (1985) described an ideal

equation as one which adequately estimated the overall shape of a growth curve and can

be extended or modified to give greater flexibility and precision. Based on the

sigmoidal nature of growth it could be expected that the Gompertz curve would give a

better fit than an asymptotic curve such as Brody’s however the best fitting curve

actually varies between species (Brown, 1970; Brown et al., 1976; Malhado et al.,

2009) and even between breeds (Topal et al., 2004).

The lack of a single nonlinear curve to describe the growth of all animals is likely to be

due to the failure of animals to achieve their potential to grow due to inadequate

feeding, disease or adverse environments resulting in real growth data being

inadequately described by a smooth function. Therefore selection of the best nonlinear

function to describe growth in sheep is debatable and often made empirically (Fitzhugh,

1976). In this experiment the Brody function was chosen as the nonlinear model due to

its ability to converge easily and accommodate missing data points (Kaps et al., 1999,

2000), with the number of individual convergences in relation to the total sample size

86

representing the best fit when comparing nonlinear models (Souza and Bianchini

Sobrinho, 1994).

3.5.3 Residual weights

Residual weights, the difference between the estimated and the recorded weights, have

been used as a measure of growth curve fit in both sheep (Lambe et al., 2006; Malhado

et al., 2009; Topal et al., 2004) and cattle (Brown, 1970; Brown et al., 1976). The

ability of each function to estimate lamb weight varied with time, as shown by the

residual weight graph (Figure 3-2). These variations in the residual weights from the

Brody fit mirrored the results in other studies of an underestimation of birth weight

(Malhado et al., 2009), an overestimation of weight during early life growth, including

the weaning period at around 100 days (Gbangboche et al., 2008; Malhado et al., 2009),

and an underestimation of weight during the post weaning period at around 150 days

(Gbangboche et al., 2008). In cattle Beltrán et al. (1992) found that the Brody model

overestimated weights before 18 months, which included overestimation at weaning, as

found in this study with lambs.

Residual weights were squared, summed and averaged in this study to determine the

mean squared error for the fit of each three curves. The MSE for the Brody function was

the highest, indicating the poorest fit to the data and was similar to the MSE reported by

Topal et al. (2004) in Morkaraman sheep. The cubic polynomial had the lowest MSE

although it can be seen from the residual weight graph (Figure 3-2) that the fit of the

cubic polynomial and random effects model is similar. Furthermore, during the period

of interest from 90 to 160 days, the MSE for the random effects model varied less from

the MSE for the whole curve fit than with the cubic fit, indicating the stability of the fit

varied little with time. The MSE for the cubic polynomial fit was lowest, however it had

the highest frequency of high residual weights of all three curve fits (Figure 3-4) and the

87

random effects model fit had the lowest (Figure 3-5). The residual weights for the

random effects model fit also displayed a similar skewedness but, as expected, lower

kurtosis than the cubic polynomial fit.

3.5.4 Random effects model

The random effects model methodology offers a more powerful and flexible means of

evaluating repeated live weight measures than a linear or nonlinear function including

the ability to accommodate missing data points and estimate a growth curve for animals

with as little as one weight recording. Environmental effects specific to the time of

recording can be accounted for and models can accommodates genetic differences in the

shape of each animal’s growth curve (Lewis and Brotherstone, 2002). This reduces

potential bias when weights are estimated near the edge of the available data. As

orthogonal forms of linear models are typically fitted with random effects models,

correlations between coefficients are generally lower than with ordinary polynomials.

While this is a desirable feature, there is no simple biological interpretation of the

growth function parameters.

At present only linear models can pragmatically be fitted with random effects models

although several authors have expressed the opinion that the use of nonlinear model

may have advantages (Lewis and Brotherstone, 2002).

88

3.6 Conclusion

This study demonstrated that the individually fitted cubic polynomial function and the

population based random effects model, which is based on polynomial functions,

provide a better fit to lamb live weight data than the individually fitted nonlinear Brody

function. The cubic polynomial had the lowest mean squared error, indicating the best

overall curve fit. However, the cubic polynomial also had the highest frequency of large

residual weights, while the frequency was lowest for the random effects model. The

ability of each function to estimate lamb weight varied with time although in general,

each function overestimated weight at weaning and underestimated weight at the post

weaning time point of 150 days, the magnitude of the difference being largest with the

Brody fit and similar with the cubic and random effects model fit. The random effects

model was subject to less bias and able to utilise data for a greater number of animals

due to its ability to accommodate data sets of differing sizes, which are common in an

industry setting, and infer information from animals with similar genetic and

environmental information. Therefore, the random effects model methodology offers a

more powerful and flexible means of evaluating repeated live weight measures.

89

Chapter 4. Lamb birth and rearing type influence

growth and weight potential of progeny from high

growth sires

4.1 Introduction

The ability to estimate lamb weights and growth rates at key time points is essential to

implementing appropriate management and feeding strategies. Therefore, factors which

influence lamb weights and growth rates have the potential to impact the economic

viability of a production system. Live weight has been shown previously to be

influenced by multiple births, by dam age and by the sex of the lamb. Thus ewe lambs,

lambs from 1 year old dams and multiple birth lambs tend to be lighter at birth

(Afolayan et al., 2007). Ewe nutrition is also important determinant of lamb weight with

lambs born to ewes with restricted nutrition being 34% lighter at birth (Schinckel and

Short, 1961). As birth weight is phenotypically correlated to adult weight (Huisman and

Brown, 2008) these lambs are likely to be lighter and grow slower throughout the early

life growth period.

Weights and growth rates in lambs can also be influenced by genetics. The Australian

sheep industry uses sire Australian Sheep Breeding Values (ASBVs) to select for

increased growth via an increased sire post weaning weight (PWWT) which has been

shown to correlate with an increased mature size (Huisman and Brown, 2008). The

magnitude of this response can be depressed by poor ewe nutrition (Hegarty et al.,

2006c). Therefore production factors which restrict lamb or ewe nutrition such as birth

90

type, rear type or dam breed may also depress the progeny response to increased sire

PWWT.

Lamb nutrition varies in-utero with the number of placentomes per foetus and the

relative placental weight per foetus declining as litter size increases (Greenwood et al.,

2000b). As placental weight is the largest determinant of foetal weight close to birth

(Greenwood et al., 2000b) placentally mediated foetal growth restriction due to multiple

births is likely to result in lighter lamb birth weights and a reduced capacity to express

the full PWWT potential. Furthermore, milk production has been shown to increase by

just 27% at peak yield in ewes with twin lambs (Morgan et al., 2006) representing a

nutritional restriction between birth and weaning for lambs raised as twins. Therefore

both birth and rearing type are likely to influence the expression of sire PWWT.

Milk production by the dam is highly correlated with lamb growth (Burris and Baugus,

1955; Snowder and Glimp, 1991) and milk yield is reduced in Merino dams compared

to Border Leicester-Merino cross dams (Kleemann et al., 1984) therefore we also expect

increased responses per kg of sire PWWT in lambs from Border Leicester-Merino cross

dams.

In addition to increased growth, ASBVs are used to select for increased muscling via an

increased post weaning eye muscle depth (PEMD) and leanness via a reduced post

weaning fat depth (PFAT) at the c-site (5cm from the midline over the 12th

rib). Gardner

et al. (2015) demonstrated an increase in pre slaughter live weight with a reduction in

PFAT and no association with PEMD. Therefore selection for leanness via a reduced

sire PFAT may result in heavier progeny while selection for increased muscling via an

increased sire PEMD may have no impact on live weight.

91

As a component of the Information Nucleus Flock (van der Werf et al., 2010) run by the

Australian Cooperative Research Centre for Sheep Industry Innovation numerous

measurements, including live weight, were taken on the progeny of sires with divergent

ASBV values enabling the effects of genotype and production factors to be determined.

As the effect of sire PWWT varies with nutrition we hypothesised that the magnitude of

the PWWT effect would vary between sites. We also hypothesised that lambs born and

raised as multiples or born to Merino dams would have a reduced response to increased

sire PWWT. Furthermore, lambs from low PFAT sires will be heavier while selection

for increased muscling via an increased sire PEMD will not impact lamb weight.

4.2 Materials and Methods

4.2.1 Experimental Design

The design of the Sheep CRC’s Information Nucleus Flock was presented in detail by

Fogarty et al. (2007) and van der Werf et al. (2010). Each year for 5 years a total of

approximately 3,500 lambs were produced at eight research sites across Australia,

which represent a broad range of production systems (Kirby NSW, Trangie NSW,

Cowra NSW, Rutherglen VIC, Hamilton VIC, Struan SA, Turretfield SA, and

Katanning WA). The lambs were progeny of 435 sires selected to represent the key

breeds used in Australia, and of different production traits. The sire types included

Merino sires (Merino, Poll Merino), Maternal sires (Border Leicester, Booroola,

Coopworth, Corriedale, Dohne Merino, East Friesian, Prime SAMM), and Terminal

sires (Hampshire Down, Ile De France, Poll Dorset, Southdown, Suffolk, Texel, White

Suffolk). Lambs were reared under extensive pasture grazing conditions and fed grain,

hay or feedlot pellets when feed supply was limited (Ponnampalam et al., 2014a).

92

4.2.2 Lamb weights

Lambs were weighed at birth, at weaning (90 days of age) and approximately every two

weeks thereafter.

4.2.3 Data available

A total of 18,185 lambs with 142,572 weight recordings were available. Of these, a total

of 17,525 lambs with 141,206 weight recordings had data available for the fixed

production effects in the model, which included site, year of birth and sire type. Actual

birth weight was available for 17,480 animals. Actual birth weight and estimated weight

and estimated growth rate ranges are presented in Table 4-1.

Table 4-1 Actual birth weight (kg) ranges and estimated weight (kg) and growth rate (g/day) ranges at

100, 150 and 240 days.

Mean Minimum Maximum Range SD

Birth Weight (kg) 4.91 1.40 10.10 8.70 1.06

Weight 100 days (kg) 28.70 3.38 60.76 57.38 8.17

Weight 150 days (kg) 34.38 10.81 68.78 57.97 7.84

Weight 240 days (kg) 43.42 21.24 74.94 53.70 8.90

Growth Rate 100 days (g/day) 136.96 -184.43 547.70 732.13 80.52

Growth Rate 150 days (g/day) 97.17 -170.23 393.49 563.72 67.23

Growth Rate 240 days (g/day) 125.11 -559.42 897.38 1456.80 167.58

SD: Standard deviation

Of the 435 sires used, 84 Maternal, 167 Merino and 184 Terminal had ASBVs for birth

weight (BWT), PWWT, PFAT and PEMD. The PFAT and PEMD breeding values were

based on live ultrasound measurements while the BWT and PWWT breeding values

were based on live weight. All ASBVs were sourced from Sheep Genetics, Australia’s

93

national genetic evaluation database for sheep (Brown et al., 2007). A correlation

matrix is presented in

Table 4-2. The ASBVs were estimated in April 2012 and were generated within

individual databases for Terminal, Maternal and Merino sires. Progeny from the

Information Nucleus Flock were excluded from the calculation of sire ASBVs.

Table 4-2 Phenotypic correlations (standard errors) of post weaning sire breeding values for weight, fat

depth and eye muscle depth among Maternal, Merino and Terminal sires.

PWWT PFAT PEMD

Maternal

PWWT 1.00 -0.18 (0.09) -0.06 (0.62)

PFAT -0.18 (0.09) 1.00 0.47 (0.00)

PEMD -0.06 (0.62) 0.47 (0.00) 1.00

Merino

PWWT 1.00 0.23 (0.00) 0.29 (0.00)

PFAT 0.23 (0.00) 1.00 0.76 (0.00)

PEMD 0.29 (0.00) 0.76 (0.00) 1.00

Terminal

PWWT 1.00 -0.03 (0.69) -0.26 (0.00)

PFAT -0.03 (0.69) 1.00 0.39 (0.00)

PEMD -0.26 (0.00) 0.39 (0.00) 1.00

4.2.4 Statistical analysis

A mixed effects regression was used in the study due to its ability to accommodate the

correlation of data due to repeated measures for the growth curve, and its ability to

handle unbalanced data. This allowed the estimation of live weight gain, allowing

weights and growth rates to be determined at the key ages in the growth cycle. Markov

chain Monte Carlo (MCMC) stochastic simulation, developed for estimation and

inference, was used within a generalised linear mixed model in the statistical package R

94

to fit a growth curve to the entire population and the growth curve of each individual

lamb deviated from this population curve. The number of iterations was 100,000 with a

burn in period of 5,000 and thinning was set at 100. The priors used were g priors.

Initially base models were developed to describe live weight, which included fixed

production effects for site (representing research station site; Kirby, Trangie, Cowra,

Rutherglen, Hamilton, Struan, Turretfield, Katanning), year of birth (2007, 2008, 2009,

2010, 2011), sex (male, female), birth type – rear type (11, 21, 22, 31, 32, 33), age of

dam (2, 3, 4, 5, 6, 7 years), sire type (Merino, Maternal, Terminal) and dam breed

within sire type (Merino x Merino, Merino x Maternal, Merino x Terminal, Border

Leicester-Merino x Terminal). The production effects and their interactions with animal

age was modelled using linear, quadratic and cubic terms for time, in days, since birth.

Individual identification, sire identification and dam identification by year were

included as random terms and were tested for linear, quadratic and cubic interactions

with time. All relevant first order interactions between fixed effects were tested and

removed if non-significant (P > 0.05).

From this model weights and growth rates, the first derivative from the random effects

model, were estimated for each animal at 100, 150 and 240 days with day 100

representing weaning age, day 150 representing the post weaning time point and day

240 representing near the average age at which the post weaning weight ASBV is

calculated in the Sheep Genetics data base (Brown et al., 2007). All appropriate results

for the individual, including curves for sire, dam, dam by year and animal, were

combined to calculate the estimated individual growth curve and derivative.

The estimated weights and growth rates were then analysed in SAS using a linear mixed

effects model (SAS Version 9.1, SAS Institute, Cary, NC, USA) where fixed and

random effects were tested using likelihood ratios. The model contained the same

95

production effects as described in the base random effects model above however the

effect of the production factors did not vary with time, nor was time included as a

random variable. The models contained relevant first order interactions between the

production effects and these were removed in a stepwise manner, by a process of

backward elimination, if non-significant (P > 0.05). The impact of production effects on

lamb weight at 100, 150 and 240 days could be estimated from R however it was not

possible to determine the impact on growth rate directly from R. As the impact of

production effects on weight as estimated from R and SAS were similar a decision was

made to determine the impact of production effects on both weight and growth rate in

SAS.

To assess the impact of sire birth weight (BWT) ASBV on weight at birth BWT was

included as a covariate in the base model. The model also contained relevant first order

interactions between the fixed effects and covariates, including a quadratic effect of the

covariate, and were removed if non-significant (P > 0.05).

To assess the impact of sire weaning weight (WWT) ASBV on weight and growth rate

at 100 days WWT was included as a covariate in the base model. The model also

contained relevant first order interactions between the fixed effects and covariates,

including a quadratic effect of the covariate, and were removed if non-significant (P >

0.05).

To assess the impact of sire PWWT, PEMD and PFAT on weight and growth rate at

150 and 240 days these were included in the base model. The model also contained

relevant first order interactions between the fixed effects and covariates, including a

quadratic effect of the covariate, and were removed if non-significant (P > 0.05).

96

4.3 Results

4.3.1 Actual birth weights

The phenotypic linear mixed effects model is presented in Table 4-3. The model did not

change when animals with birth weights greater than 3 standard deviations from the

mean were excluded so these animals were retained.

97

97

Table 4-3 Numerator and denominator degrees of freedom and F-values for the base models for birth weight and estimated weights and growth rates at 100, 150 and 240 days.

Birth Weight Day 100 Day 150 Day 240

Wt Gr Wt Gr Wt Gr

NDF;DDF F-value NDF;DDF F-value F-value NDF;DDF F-value F-value NDF;DDF F-value F-value

Site 7;5085 473.89** 7;5099 778.80** 306.43** 7;5099 1836.24** 69.08** 7;5099 19950.40** 42.38**

Year 4;12,000 4.64** 4;12,000 22.19** 5.56** 4;12,000 46.66** 6.49** 4;12,000 275.18** 20.92**

Sex 1;5085 853.25** 1;5099 480.52** 32.55** 1;5099 954.89** 5.06* 1;5099 33221.30** 409.77**

Birth type-rear type 2;5085 3326.84** 5;5099 724.59** 6.36** 5;5099 1009.60** 10.87** 5;5099 4108.29** -

Dam age 6;5085 29.83** 6;5099 27.64** 48.07** 6;5099 25.46** 32.02** 6;5099 300.10** 17.62**

Sire type 2;5085 268.01** 2;5099 1140.22** 32.16** 2;5099 2409.76** 56.59** 2;5099 4686.92** 49.80**

Dam breed (sire type) 1;5085 344.23** 1;5099 719.81** 67.02** 1;5099 1363.61** - 1;5099 8350.83** -

Site by Year 27;5085 42.53** 27;5099 167.95** 193.97** 27;5099 128.92** 53.17** 27;5099 80.71** 100.77**

Site by Sire type - - 14;5099 27.16** 39.86** 14;5099 15.67** - 14;5099 11.70** 13.56**

Year by Sire type - - 8;5099 - 51.75** - - - - - 36.57**

NDF, DDF; Numerator and denominator degrees of freedom; Wt; Weight; Gr; Growth rate; *P < 0.05, **P < 0.01

98

4.3.1.1 Production effects on actual birth weight

The average actual birth weight was 4.91 kg and ranged from 1.40 to 10.10 kg. Birth

weight was greatest at the Turretfield site and least at the Kirby site with average birth

weights of 5.09 kg and 3.90 kg respectively (P < 0.01). Lambs born in 2011 were

heaviest on average weighing 4.77 kg and lambs born in 2010 were the lightest

weighing 4.65 kg (P < 0.01). Within a site birth weight varied between years by as

much as 0.86 kg at Katanning and as little as 0.37 kg at Cowra (P < 0.01).

Predicted means for the impact of the production effects on birth weight are presented in

Table 4-4. Ram lambs were on average 0.32 kg heavier than ewe lambs (P < 0.01)

(Table 4-5). Single born lambs were on average 1.02 kg heavier than twin lambs and

1.74 kg heavier than triplets (P < 0.01). Lambs born to 2 year old dams were the lightest

and were on average 0.50 kg lighter than lambs born to 8 year old dams which were the

heaviest (P < 0.01). Lambs were 0.22 kg heavier on average for each year that dam age

increased up until 4 years of age (P < 0.01) then there was no significant difference in

lamb weight with each year that dam age increased. Terminal sired lambs were on

average 0.43 kg heavier than Maternal sired lambs and 0.53 kg heavier than Merino

sired lambs (P < 0.01). Within the Terminal sired lambs, those from Border Leicester-

Merino dams were on average 0.46 kg heavier than lambs from Merino dams (P <

0.01).

99

97

Table 4-4 Predicted means for birth weight (kg) and estimated weight at 100, 150 and 240 days for different sexes, birth type-rear types, dam ages, sire types and dam breed within

sire type combinations.

Birth Weight Weight Day 100 Weight Day 150 Weight Day 240

(kg) (kg) (kg) (kg)

Level Mean s.e. Coefficient s.e. Coefficient s.e. Coefficient s.e.

Sex F 4.54 0.02 26.54 0.13 32.38 0.11 40.58 0.07

M 4.86 0.02 28.10 0.13 34.14 0.11 43.76 0.07

Birth type-rear type 11 5.62 0.02 31.87 0.12 37.47 0.10 45.47 0.07

21 4.60 0.02 29.41 0.16 35.27 0.13 43.68 0.07

22 - - 26.75 0.12 32.61 0.10 41.76 0.07

31 3.88 0.03 28.33 0.37 34.33 0.30 42.83 0.12

32 - - 25.12 0.23 31.21 0.19 40.61 0.09

33 - - 22.45 0.26 28.66 0.22 38.70 0.10

Dam age 2 4.34 0.05 27.59 0.32 31.78 0.26 38.90 0.11

3 4.57 0.02 27.40 0.16 33.33 0.13 42.52 0.07

4 4.77 0.02 27.68 0.14 33.98 0.12 42.97 0.07

5 4.78 0.02 27.98 0.14 33.76 0.12 42.63 0.07

6 4.80 0.02 28.64 0.16 33.69 0.13 42.43 0.07

7 4.79 0.03 26.81 0.22 33.39 0.18 42.75 0.09

8 4.84 0.06 25.14 0.36 32.89 0.29 43.01 0.12

Dam breed-Sire type Maternal-Merino 4.59 0.03 27.01 0.22 33.43 0.18 42.52 0.14

Merino-Merino 4.49 0.02 22.87 0.16 27.95 0.13 34.57 0.10

Terminal-Merino 4.79 0.02 29.82 0.18 36.02 0.15 47.07 0.10

Terminal-BLM 5.25 0.02 34.34 0.17 40.76 0.14 51.79 0.09

s.e.: Standard Error; BLM: Border Leicester-Merino

100

97

Table 4-5 Magnitude of effect (difference between the highest and lowest values) for production or for genetic effects on weight and growth rate.

Birth Weight Weight Weight Weight Growth Rate Growth Rate Growth Rate

Day 100 Day 150 Day 240 Day 100 Day 150 Day 240

Site 1.19 10.26 13.20 17.90 79.85 88.63 74.52

Year 0.12 2.72 2.79 1.95 12.16 14.28 88.72

Sex 0.32 1.56 1.76 3.18 4.10 3.90 30.56

Birth type-rear type 1.74 9.42 8.81 6.77 10.30 23.77 NS

Dam age 0.50 3.50 2.20 4.11 44.30 98.12 66.47

Sire type 0.53 9.21 10.44 14.86 25.10 30.81 82.41

Dam breed-Sire type 0.46 4.52 4.74 4.72 11.00 NS NS

BWT 0.95 - - - - - -

WWT - 5.98 - - NS - -

PWWT - - 6.62 8.24 - NS NS

PFAT - - 1.41 NS - NS NS

PEMD - - NS NS - NS NS

NS: Not significant; -: Not modelled; Australian Sheep Breeding Values; BWT: Birth weight; WWT: Weaning weight; PWWT: Post weaning weight; PFAT: Post weaning fat depth;

PEMD: Post weaning eye muscle depth

101

4.3.1.2 ASBV effects

In the base model sire as a random term was significant (P < 0.01) with 95% confidence

intervals of sire estimates presented in Table 4-6.

Table 4-6 95% Confidence intervals of sire estimates for birth weight (kg) and estimated weights (kg) and

growth rates (g/day) at 100, 150 and 240 days.

Maternal Merino Terminal

Birth Weight (kg) 4.55 - 4.61 4.46 - 4.51 4.99 - 5.02

Weight 100 days (kg) 26.69 - 27.31 22.72 - 23.10 31.89 - 32.18

Weight 150 days (kg) 33.18 - 33.64 27.83 - 28.15 38.20 - 38.46

Weight 240 days (kg) 42.34 - 42.74 34.40 - 34.74 49.28 - 49.60

Growth Rate 100 days (g/day) 142.16 - 150.03 119.10 - 125.20 142.71 - 151.54

Growth Rate 150 days (g/day) 111.73 - 118.28 82.00 - 85.99 109.57 - 113.56

Growth Rate 240 days (g/day) 92.37 - 109.50 70.11 - 80.94 154.36 - 170.21

BWT had an association with actual weight at birth increasing it by 0.53 kg for each kg

of BWT (P < 0.05) (Table 4-7). Across the 1.8 kg range, increasing BWT was

associated with an increase of 0.95 kg in weight at birth (Table 4-5). The association

varied with year of birth and birth type. The variation between years was as much as

0.75 kg per kg of BWT in 2011 and as little as 0.43 kg in 2010 while the variation

between birth types was as much as 0.70 kg per kg of BWT in single born lambs and as

little as 0.37 kg in triplet born lambs.

102

97

Table 4-7 Regression coefficients for the relationship between sire ASBVs for growth and actual birth weight (kg) or estimated weight and growth rate at 100, 150 and 240 days.

Birth Weight Weight Day 100 Growth Rate Day 100 Weight Day 150 Growth Rate Day 150 Weight Day 240 Growth Rate Day 240 (kg) (kg) (g/d) (kg) (g/d) (kg) (g/d)

Level Coeff. s.e. Coeff. s.e. Coeff. s.e. Coeff. s.e. Coeff. s.e. Coeff. s.e. Coeff. s.e.

BWT ASBV WWT ASBV PWWT ASBV PWWT ASBV

ASBV 0.59 0.09 0.19 0.12 NS - 0.17 0.05 NS - 0.45 0.03 3.99 1.83

ASBV*Site Kirby - - -0.35 0.08 - - -0.13 0.04 - - - - - -

Trangie - - 0.07 0.12 - - 0.10 0.06 - - - - - -

Cowra - - 0.05 0.10 - - 0.09 0.05 - - - - - -

Rutherglen - - -0.16 0.09 - - -0.03 0.04 - - - - - -

Hamilton - - -0.26 0.09 - - -0.13 0.05 - - - - - -

Struan - - -0.22 0.10 - - -0.09 0.05 - - - - - - Turretfield - - -0.09 0.10 - - -0.03 0.04 - - - - - -

Katanning - - 0.00 . 0.00 . - - - -

ASBV*Year 2007 -0.27 0.07 - - - - - - - - -0.21 0.02 - -

2008 -0.22 0.07 - - - - - - - - -0.05 0.02 - -

2009 -0.30 0.08 - - - - - - - - -0.05 0.02 - - 2010 -0.33 0.07 - - - - - - - - -0.03 0.02 - -

2011 0.00 . - - - - - - - - 0.00 . - -

ASBV*Birth type-rear type 11 0.33 0.09 0.46 0.08 - - 0.28 0.04 - - - - -6.13 1.46

21 0.15 0.09 0.39 0.09 - - 0.21 0.04 - - - - -4.35 1.55

22 0.23 0.08 - - 0.15 0.04 - - - - -3.15 1.46 31 0.00 . 0.16 0.14 - - 0.11 0.07 - - - - -4.67 2.19

32 0.12 0.11 - - 0.10 0.05 - - - - -3.52 1.79

33 0.00 . - - 0.00 . - - - - 0.00 .

ASBV*Dam age 2 - - - - - - - - - - -0.07 0.02 - -

3 - - - - - - - - - - -0.02 0.02 - - 4 - - - - - - - - - - -0.01 0.02 - -

5 - - - - - - - - - - -0.01 0.02 - -

6 - - - - - - - - - - -0.05 0.02 - - 7 - - - - - - - - - - -0.03 0.02 - -

8 - - - - - - - - - - 0.00 . - -

ASBV*Dam breed-Sire type Maternal-Merino - - -0.01 0.13 - - -0.06 0.05 - - - - - -

Merino-Merino - - 0.26 0.09 - - 0.06 0.04 - - - - - -

Terminal-Merino - - 0.00 . - - 0.00 . - - - - - - Terminal-BLM - - 0.00 . - - 0.00 . - - - - - -

s.e.: Standard Error; Coeff.: Coefficient; BLM: Border Leicester-Merino; NS: Non significant

103

4.3.2 Estimated live weights and growth rates at 100, 150 and 240 days

The phenotypic linear mixed effects models are presented in Table 4-3

4.3.2.1 Production effects

The average weights at 100, 150 and 240 days were 28.70 ± 8.17, 34.38 ± 7.84 and

43.42 kg ± 8.90 kg. Lambs at Trangie were the heaviest on average at 100, 150 and 240

days weighing 31.87, 40.38 and 53.61 kg, while lambs at Kirby were lightest at 100 and

150 days weighing 21.61 and 27.18 kg (P < 0.01) (Table 4-5). At 240 days the lambs

from Hamilton were lightest on average at 35.71 kg while lambs at Kirby were second

lightest at 35.87 kg. Lambs born in 2011 were heaviest on average at 100 and 150 days

weighing 29.16 and 35.18 kg and at these ages lambs born in 2009 were the lightest

weighing 26.44 and 32.39 kg (P < 0.01). At 240 days, lambs born in 2007 were heaviest

on average weighing 43.08 kg and lambs born in 2008 were lightest weighing 41.13 kg.

At 100 days the difference between years varied across sites by as much as 11.17 kg at

Katanning and as little as 3.89 kg at Cowra. At 150 days the difference between years

varied across sites by as much as 10.72 kg at Trangie and as little as 2.71 kg at Cowra.

At 240 days the difference between years varied across sites by as much as 4.44 kg at

Turretfield and as little as 3.33 kg at Rutherglen (P < 0.01).

Predicted means for the impact of the production effects on weight are presented in

Table 4-4 and predicted means for the impact of the production effects on growth rate

are presented in Table 4-8. Wethers were 1.56 kg heavier than ewe lambs at 100 days,

1.76 kg heaver at 150 days and 3.18 kg heaver at 240 days (P < 0.01). Wethers also

grew faster by 4.10 g/day at 100 days, 3.90 g/day at 150 days and by 30.56 g/day at 240

days (P < 0.01).

104

97

Table 4-8 Predicted means for estimated growth rate at 100, 150 and 240 days for different sexes, birth type-rear types, dam ages, sire types and dam breed within sire type

combinations.

Growth Rate Day 100

Growth Rate Day 150

Growth Rate Day 240

(g/day)

(g/day)

(g/day)

Level

Coefficient s.e.

Coefficient s.e.

Coefficient s.e.

Sex F

136.90 1.84

101.50 2.67

96.44 3.34

M

141.00 1.84

105.40 2.67

127.00 3.34

Birth type-rear type 11

140.20 1.69

90.53 2.26

- -

21

143.70 2.06

97.83 3.48

- -

22

135.70 1.72

104.00 2.17

- -

31

143.10 4.00

104.20 8.49

- -

32

137.60 2.86

110.01 4.84

- -

33

133.40 3.29

114.30 5.22

- -

Dam age 2

118.20 3.71

58.48 6.39

137.50 8.07

3

136.70 2.05

100.90 3.12

116.50 3.93

4

148.10 1.92

109.50 2.84

107.90 3.64

5

138.10 1.93

99.68 2.83

118.40 3.65

6

124.50 2.12

78.51 3.35

137.00 4.24

7

144.60 2.63

120.80 4.59

93.84 5.65

8

162.50 4.09

156.60 8.04

71.03 9.35

Dam breed-Sire type Maternal-Merino

146.50 3.45

114.90 3.68

99.01 6.60

Merino-Merino

122.60 2.42

84.09 3.01

76.89 4.51

Terminal-Merino

142.20 2.50

111.50 2.85

159.30 4.28

Terminal-BLM

153.20 2.47

111.50 2.85

159.30 4.28

s.e.: Standard Error; BLM: Border Leicester-Merino

105

Lambs born and raised as singles were the heaviest, being 2.46 kg heavier at 100 days,

2.20 kg heavier at 150 days and 1.79 kg heavier at 240 days than the next heaviest

lambs, twins raised as singles (P < 0.01) (Table 4-4). This was generally reflected in the

growth rates at 100 days with lambs raised as singles growing faster than lambs raised

as multiples (Table 4-8). At 150 days, lambs born and raised as triplets were growing

between 10 and 24 g/day faster than lambs born as twin or singles. At 240 days growth

rate did not vary between birth and rear types.

At 100 and 150 days the progeny of 4-6 year old dams were heavier (P < 0.01) however

by 240 days this pattern was no longer evident (Table 4-4). At 100 and 150 days the

lambs of 8 year old dams were growing the fastest and by 240 days were also the

heaviest (P < 0.01) (Table 4-8).

The progeny of Terminal sires were 5.07, 4.96 and 6.91 kg heavier than the progeny of

Maternal sires and 9.21, 10.44 and 14.86 kg heavier than the progeny of Merino sires at

100, 150 and 240 days respectively (P < 0.01) (Table 4-4). The progeny of Terminal

sires also grew 25.10, 27.41 and 82.41 g/day faster than the progeny of Merino sires at

100, 150 and 240 days (P < 0.01) (Table 4-8).

Within the Terminal sired lambs, those from Border Leicester-Merino dams were 4.52,

4.74 and 4.72 kg heavier than those from Merino dams at 100, 150 and 240 days (P <

0.01) (Table 4-4) and grew faster by 11.00 g/day at 100 days (Table 4-8).

4.3.2.2 ASBV effects

In the base models for weight and growth rate at 100, 150 and 240 days sire as a random

term was significant (P < 0.01) with 95% confidence intervals of sire estimates

presented in Table 4-6.

9

97

106

4.3.2.2.1 WWT and PWWT on weight at 100, 150 and 240 days

WWT demonstrated an association with weight at 100 days (P < 0.01). Across the 16 kg

range, increasing WWT was associated with an increase in weight of 5.98 kg (Table

4-5). This effect varied between sites, birth type-rear type combinations and sire types

(P < 0.01) (Table 4-7). There was no effect at the Kirby, Hamilton or Struan sites

however within the Trangie, Cowra, Katanning, Turretfield and Rutherglen sites weight

increased by 9.00, 8.75, 7.90, 6.47 and 5.34 kg across the 16 kg range in WWT. The

was no effect in triplet born lambs however the weight of single born lambs increased

by 9.63 kg across the 16kg range in WWT while the weight of 21 and 22 lambs

increased by 8.53 and 6.09 kg respectively. Progeny of Merino sires had the largest

weight response of 0.55 kg per kg of WWT increase while progeny of Terminal sires

increased by 0.29 and progeny of Maternal sires increased by 0.28. This equated to

increases of 6.66, 2.91 and 2.31 kg across the 12.11, 10.04 and 8.26 kg WWT ranges for

Merino, Terminal and Maternal sire types.

PWWT demonstrated an association with weight at 150 days (P < 0.01) (Figure 4-1).

Across the 23 kg range, increasing PWWT was associated with an increase in weight of

6.62 kg. This effect varied between sites, birth type-rear type combinations and sire

types (P < 0.05) (Table 4-7). There was no effect at the Kirby or Hamilton sites

however within Trangie, Cowra, Katanning, Rutherglen, Turretfield, Rutherglen and

Struan sites weight increased by 9.60, 9.28, 7.29, 6.54, 6.49 and 5.09 kg across the 23

kg range in PWWT. There was no effect in triplet born lambs however the weight of

single born lambs increased by 9.83 kg across the 23 kg range in PWWT while the

weight of 21 and 22 lambs increased by 8.29 and 6.81 kg respectively. Progeny of

Merino sires had the largest weight response of 0.35 kg per kg of PWWT while progeny

of Terminal sires increased by 0.29 and progeny of Maternal sires increased by 0.22.

107

This equated to increases of 5.54, 4.98 and 3.11 kg across the 15.83, 17.16 and 14.14 kg

PWWT ranges for Merino, Terminal and Maternal sire types (Figure 4-1).

Figure 4-1 Comparison of sire estimates for post weaning weight (PWWT) Australian Sheep Breeding

Values for Maternal, Merino and Terminal sires with predicted means for weight at 150 days. Lines

represent predicted means for weight (± s.e.) for each sire type. Symbols (∆,+,o) represent individual sire

estimates.

20

25

30

35

40

45

-5 0 5 10 15 20

Lam

b W

eigh

t 150 d

ays

(kg)

Sire PWWT Breeding Value

Maternal Merino Terminal

9

97

108

PWWT demonstrated an association with weight at 240 days (P < 0.01) (Table 4-7).

Across the 23 kg range, increasing PWWT was associated with an increase in weight of

8.24 kg. This effect varied between years and dam ages (P < 0.01) (Table 4-7). In 2011,

2010, 2008, 2009 and 2010 weight increased by 9.81, 9.11, 8.66, 8.63 and 4.97 kg

across the 23 kg range in PWWT. Lambs from dams aged 8, 5, 4, 3, 7, 6 and 2 increased

in weight by 8.85, 8.64, 8.60, 8.46, 8.20, 7.74 and 7.17 kg across the 23kg range in

PWWT.

4.3.2.2.2 WWT and PWWT on growth rate at 100, 150 and 240 days

There was no impact of WWT on growth rate at 100 days and no impact of PWWT on

growth rate at 150 days (Table 4-7). PWWT demonstrated an association with growth

rate at 240 days however this effect was only present in some birth type-rear type

combinations (Table 4-7). The variation was as much as 3.99 g/day in triplet born and

raised lambs and as little as -2.14 in single born lambs. This equated to an increased

growth rate in triplet born and raised lambs of 92.89 g/day and a reduced growth rate in

single born lambs of 49.82 g/day across the 23 kg range in PWWT.

4.3.2.2.3 PEMD on weight and growth rate at 150 and 240 days

PEMD had no effect on weight or growth rate at 150 days (Table 4-9). There was an

effect of PEMD on weight at 240 days however this was only present at the Katanning

site where weight reduced by 0.12 kg per kg of PEMD equating to a reduction of 0.96

kg across the 8 kg PEMD range. There was no effect of PEMD on growth rate at 240

days.

109

Table 4-9 Regression coefficients for the relationship between sire ASBVs for muscling and leanness and

estimated weight and growth rate at 100, 150 and 240 days.

Weight Day 240 Weight Day 150

(kg) (kg)

Coefficient s.e. Coefficient s.e.

PEMD ASBV PFAT ASBV

ASBV -0.12 0.05 -0.28 0.06

ASBV*Site Kirby -0.01 0.04 - -

Trangie 0.01 0.05 - -

Cowra 0.13 0.05 - -

Rutherglen 0.10 0.05 - -

Hamilton 0.06 0.05 - -

Struan 0.15 0.05 - -

Turretfield 0.05 0.05 - -

Katanning 0.00 . - -

ASBV: Australian Sheep Breeding Value; PEMD: Post weaning eye muscle depth; PFAT: Post weaning

fat depth; s.e.: Standard Error

9

97

110

4.3.2.2.4 PFAT on weight and growth rate at 150 and 240 days

PFAT demonstrated an association with weight at 150 days, increasing weight by 1.41

kg across the decreasing 5 kg PFAT range (P < 0.01) (Figure 4-2). PFAT had no effect

on weight at 240 days or growth rate at 150 or 240 days.

Figure 4-2 Comparison of sire estimates for post weaning fat depth (PFAT) Australian Sheep Breeding

Values for Maternal, Merino and Terminal sires with predicted means for weight at 150 days. Lines

represent predicted means for weight (± s.e.) for each sire type. Symbols (∆,+,o) represent individual sire

estimates.

20

25

30

35

40

45

-3 -2 -1 0 1 2 3

Lam

b W

eigh

t 150 d

ays

(kg)

Sire PFAT Breeding Value

Maternal Merino Terminal

111

4.4 Discussion

4.4.1 Association between carcass breeding values and lamb weight

and growth rate

As expected, the progeny of sires with increased BWT were heavier at birth, WWT

were heavier at 100 days and PWWT were heavier at 150 and 240 days. As lamb

survival increases with increased birth weight in Merinos (Holst et al., 2002) increased

sire BWT may increase lamb survival rates. However, the relationship is non-linear

with increased lamb mortality beyond birth weights of 6.5kg, primarily due to dystocia

(Brown et al., 2014; Fogarty et al., 1992; Fogarty et al., 2000; Oldham et al., 2011;

Smith, 1977). Therefore at the highest extreme of BWT lamb mortality will be

increased.

As hypothesised, the magnitude of the responses to WWT and PWWT varied between

sites. This is likely to be due, in part, to differences in nutrition at different sites which

is known to impact on the magnitude of the PWWT effect (Hegarty et al., 2006c). The

largest response to WWT and PWWT were seen at Trangie and Cowra, the two sites at

which lambs were also the heaviest, consistent with improved nutrition. Alternatively,

the smallest response to WWT and PWWT was seen at Kirby where lambs were also

the lightest, consistent with inferior nutrition. These site differences were present

despite the implementation of management strategies such as supplementary feeding

when feed supply was limited or of poor quality (Ponnampalam et al., 2014a).

Nonetheless, as the dams used at each site were of different genetic lines, we cannot

discount the possibility that the varying WWT and PWWT effects have been impacted

by dam genetics. Other factors that vary between sites, and may contribute to the

variation in effect between sites, include weather and worm or other disease burden. At

9

97

112

240 days the response to increased sire PWWT did not differ between sites so it is

possible that at this time point the extremes for the sites with the poorest nutrition were

not enough to depress to response to the PWWT ASBV. There was also no variation

between sites in response to the BWT ASBV. This may indicate a buffering of the

lambs by maternal nutrition with dams mobilising more of their own energy reserves.

Alternatively, the variation may have been present but obscured by the fact that the

error represents a larger proportion of the difference between weights at birth than later

in life. This is more likely to be the case as the difference in magnitude of response

between sites in response the BWT ASBV were similar to that for WWT and PWWT

being high at Trangie and Cowra and lowest at Kirby when the site by BWT interaction

was retained despite being insignificant.

Lamb birth type and rear type also moderated the response to increased BWT, WWT

and PWWT with the magnitude of the weight response at birth, 100 and 150 days

depressed with multiple births. This depressive effect due to in-utero restriction was

partially alleviated when the nutritional restriction between birth and weaning was lifted

through loss of a sibling for multiple birth lambs. The effect on growth rate was also

moderated at 240 days where the response to increased sire PWWT was positive in

lambs born and raised as triplets and by contrast was negative in lambs born as singles.

This would imply that increased PWWT accelerated early life growth such that mature

size is reached sooner in single born lambs thus growth rate slows down by 240 days.

As triplet born and raised lambs are proportionately less of their mature size by 240

days the full PWWT effect is still present.

Sire type moderated the response to increased WWT and PWWT with progeny of

Merino sires having the greatest response at 100 and 150 days. As Merino sired progeny

weighed less and had lower growth rates in general they are likely to be proportionately

113

less of their mature size at weaning and post weaning and therefore have an increased

response to sire growth ASBVs. As sire breeding values represent twice the expected

progeny difference, lamb weight is expected to increase by 0.5kg for each 1kg increase

in sire growth ASBV. Increases in lamb weight with sire growth ASBVs were as

expected for the progeny of Merino sires and reduced to 60% of the expected increase

in progeny of Maternal and Terminal sires. Maternal sires had a reduced accuracy of

sire ASBVs compared to Merino sires which may account for some of the variation in

sire type performances.

Although lambs from Border Leicester-Merino dams were larger and grew faster than

lambs from Merino dams, there was no difference in weight response to WWT or

PWWT in these lambs. This is despite evidence suggesting restricted nutrition to

progeny of Merino dams due to their restricted milk output (Kleemann et al., 1984) and

supporting evidence that pre weaning growth rates are reduced in progeny of Merino

dams compared to progeny of Border Leicester-Merino dams (Atkins and Thompson,

1979). As expected, a weight and growth rate differential was seen in response to dam

breed as a production effect. It is not clear why there was no breeding value differential

however the nutritional restriction due to dam breed may not be enough to reach a

threshold beyond which there is an impact on the expression of breeding values. The

magnitude of the production effects of site, birth type-rear type combination and sire

type was between 1.5 and 3.5 times the magnitude of the effect of dam breed on weight

at 100, 150 and 240 days (Table 4-5). The production effects which vary with WWT

and PWWT follow a similar pattern. As dam breed only had a small impact on lamb

weight the variation in the expression of the breeding value was too small to detect.

9

97

114

As hypothesised, selection for increased muscle via an increased sire PEMD did not

impact on progeny weight. As the breeding values are calculated corrected to a constant

live weight (Brown et al., 2007) their impact on the phenotypic expression of growth

may be limited.

As hypothesised, selection for leanness via a reduced PFAT increased lamb weight

however this effect was only present at 150 days. The effect across the PFAT range was

small, representing an increase of just 4% above the average lamb weight at 150 days.

The phenotypic correlation between c-site ultrasound fat depth and weight was

relatively small in the Huisman and Brown (2008) study which may reflect the lack of a

consistent effect on weight in this study.

4.4.2 Production effects

The site at which lambs were reared impacted on weight throughout life. The

differences between sites are likely to be partly explained by nutrition and dam genetics.

Trangie had consistently high lamb weights and Kirby consistently low lamb weights,

possibly indicating increased and reduced planes of nutrition respectively. However, the

Information Nucleus Flock was not designed to control nutrition and thus conclusions

regarding the nutritional impact on birth weight are difficult to assess. As dam genetics

vary at each site they may also contribute to the variation in weights between sites. If

dam genetics at each site were entirely responsible for differences in lamb weight there

would be minimal differences in lamb weights at each site across different years

however lamb weight at a single site varied between years by nearly 40% of the average

lamb weight. Thus while dam genetics varies between sites and contributes to lamb

weight differences they are not solely responsible for these differences.

115

The year in which lambs were born had no impact at birth and was relatively consistent

between 100 and 240 days. The effect was small being as little as 10% of the magnitude

of the site of rearing, which had the largest effect. While the effect was small it may be

due to post weaning nutritional availability within a year which would explain why the

impact is not present at birth. However, the sites at which lambs were raised were

spread across different regions of Australia where pasture availability varied

dramatically. Therefore it is likely that there are also non nutritional factors which

contributed to the impact of lamb weight within a year.

Male lambs were heavier and grew faster at all times measured and this has been

demonstrated in previous studies (Afolayan et al., 2007; Thatcher et al., 1991;

Thompson et al., 1985b). Despite male lambs being castrated in this study this effect is

consistent with the impact of testosterone, a potent muscle growth stimulant (Lawrence

and Fowler, 2002). The study by Thatcher et al. (1991) compared growth rates in rams,

wethers and ewes and found that rams grew significantly faster than wethers which in

turn grew significantly faster than ewes (P < 0.05). Thus while the wethers in this

experiment grew faster than the ewes it can be surmised that growth rate would have

been further increased if the male lambs were left entire.

Single born lambs were 22% heavier than twin born lambs and 45% heavier than triplet

born lambs at birth which is consistent with findings of Afolayan et al. (2007). The

weight differences due to birth type were present throughout life, although they were

proportionately less with age. Within a birth type, weight also varied with rearing type

with twins raised as singles heavier than twins raised as twins. Triplets raised as singles

were also heavier than triplets raised as twins which were heavier than triplets raised as

triplets. These weight differences were also present throughout life and reducing

throughout age although the weight differences were at most 19%. Thus while both

9

97

116

birth type and rearing type impact on weight, birth type has the greatest effect. The

weight differences at birth were likely to be due to the effects of in-utero nutritional

restriction. As birth type represents a prenatal restriction and rear type represents a

postnatal to weaning restriction it can be concluded that prenatal restriction has a greater

impact on lamb weight. At 100 days the triplet born and raised lambs had the slowest

growth rate however by 150 days their growth rate was the fastest suggesting that

multiple born lambs experience a period of compensatory growth once nutritional

restriction is lifted at weaning. Compensatory growth in immature sheep whose

nutrition has been restricted has been previously demonstrated by Thornton et al.

(1979). In this experiment the 7 month old sheep (approximately 210 days old) had their

nutrition restricted and lost an average of 85g/day over a two month period. When feed

was reintroduced ad libitum these sheep then gained 500 g/day in the first 3 days and

300 g/day on average over 33 days to achieve similar live weights to the non-restricted

sheep. These growth rates exceed the 114.30 g/day seen in the 150 day old triplet born

and raised lambs in this experiment, however the lambs in this experiment were not

severely nutritionally restricted post weaning as in the Thornton et al. (1979)

experiment which demonstrates the difference between restriction in-utero versus just

that imposed during the postnatal period. Interestingly the initial live weight of these

Dorset Horn x Merino wethers at 7 months of age was 22.5 kg in the Thornton et al.

(1979) experiment, more than 50% lighter than current weight estimations.

Dam age influenced lamb birth weight with lambs born to 2 year old dams being the

lightest. This may be due to the utilisation of energy by the 2 year old dam for its own

growth and development at the expense of the lamb, causing in-utero nutritional

restriction. Ali et al. (2006) demonstrated an increase in birth weight of 25 grams for

each one year increase in dam age. As in our study this effect was present until dams

reached 4 years of age however the magnitude was smaller than the 230 – 430 g

117

increase between years seen in our study. This may be a reflection of the variation of the

dam age effect between dam breeds as only Rambouilet ewes were used by Ali et al.

(2006) whereas Afolayan et al. (2007) used crossbred ewes and the increase in lamb

weight at birth between dam years was as much as 630g. While there was no dam age

by dam breed effect in this study Rambouilet ewes were not used and the dam breeds

which were used may not have been divergent enough to drive an effect. Dam age also

influenced weaning and post weaning lamb weights with lambs from 4-6 year old dams

tending to be heavier at 100 and 150 days. Increases in lamb weight at birth, weaning

and post weaning (200 days) with increasing dam age were also found in the Afolayan

et al. (2007), study, although only 1-4 year old dams were included so the effect beyond

this age is unknown. At 240 days, when lambs were close to adult weight, this pattern

was no longer evident. It is possible that the early influence on lamb weight is due to

dam factors such as mothering ability, milk production and colostrum composition, the

effects of which diminish post weaning, however these effects were not measured

during this experiment.

The progeny of Terminal sires were heavier between birth and 240 days while the

progeny of Merino sires were lighter at and had the slowest growth rates which is

consistent with the findings of Fogarty et al. (2000). Furthermore, progeny of elite

Terminal sires grew in excess of 250 g /day at 100, 150 and 240 days. As many

nutritional requirement tables used for ration formulation only accommodate lamb

growth rates up to 250 g / day (ARC, 1980) the nutrient requirements in the progeny of

contemporary elite growth sires need to be updated.

9

97

118

Within the progeny of Terminal sires, those from Border Leicester-Merino dams were

heavier and had higher growth rates than those from Merino dams which is a reflection

of the increased mothering ability (Alexander et al., 1983) and milk yield of Border

Leicester-Merino dams (Kleemann et al., 1984).

Lamb weight was influenced by both production and genotypic factors and the

magnitude of the effects varied (Table 4-5). Site, birth type-rear type and sire type had

the greatest impact on lamb weight. Birth type had the largest effect on weight at birth

and this was 1.8 times the magnitude of the effect across 1.8 units of the BWT breeding

value. The difference in weight between sites had the second largest effect and was 1.3

times the magnitude of the BWT effect on birth weight. At 100, 150 and 240 days site

had the largest effect on weight. At 100 days the differences between sites was 1.7 times

the magnitude of the effect across 16 units of the WWT breeding value and between 1.1

and 6.6 times the magnitude of the other production effects. At 150 and 240 days the

differences between sites was twice the magnitude of the effect across 23 units of the

PWWT breeding value and between 1.2 and 9.2 times the magnitude of the other

production effects.

Lamb growth rate was also influenced by both production and genotypic factors (Table

4-5). At 100 and 150 days site had the largest effect on growth rate whereas at 240 days

year of birth had the largest effect. At 100 days there was no effect of WWT on growth

rate however the difference in growth rate between sites was between 1.8 and 19.5 times

the magnitude of the other production effects. At 150 days there was no effect of

PWWT on growth rate however the difference in growth rate between sites and dam

ages was similar and between 2.8 and 25.2 times the magnitude of the other production

factors. At 240 days there was no main effect of PWWT and the difference in growth

rates between years and sire types was similar and between 1.1 and 2.9 times the

119

magnitude of other production effects. Therefore, while sire ASBVs do impact on

weight and growth rate in their progeny the effect is smaller than the impact of the

environment, year of birth or birth type.

9

97

120

4.5 Conclusion

This study demonstrated that both genotypic and production factors had significant

effects on lamb growth. Selection for growth by selection of sires with an increased

BWT, WWT and PWWT resulted in lambs that were heavier at birth, weaning and post

weaning. The magnitude of these responses was moderated by nutrition hence varied

between sites and with birth type-rear type combinations. Lambs born as multiples, with

nutritional restriction from conception to weaning, had a reduced response to sire

PWWT. Selection for leanness and muscling by selection of sires with a reduced PFAT

and increased PEMD resulted had little impact on weight with increasing PEMD having

no effect at all. Reducing PFAT increased lamb weight, however this effect was only

present at the post weaning time point and was relatively small. As expected, single

born lambs, male lambs and lambs born to Terminal sires were heavier and grew faster.

Weight and growth rate were heavily influenced by environmental effects at each site

and the effects were stable with minimal re-ranking of sites based on lamb weight over

time. In conclusion, while selection for growth impacts on lamb growth, these effects

and those of production factors such as sex and sire type are generally small compared

to the impact of the environment.

121

Chapter 5. Selection for lean meat yield in lambs

reduces indicators of oxidative metabolism in the

longissimus muscle

This chapter has been published in Meat Science:

Kelman, K.R., Pannier, L., Pethick, D.W. and Gardner, G.E. (2014) Selection for lean

meat yield in lambs reduces indicators of oxidative metabolism in the longissimus

muscle. Meat Science. 96 (2, Part B): 1058-1067.

5.1 Introduction

Lamb is marketed in Australia as a nutritious red meat and a ‘good source’ of iron and

zinc (Pannier et al., 2014b). The qualities of redness and mineral content of meat are

highly linked to muscle fibre type and thus oxidative capacity of muscles, with reduced

oxidative capacity being associated with reduced levels of iron and zinc (Pannier et al.,

2014b) and a less red appearance. For this reason oxidative capacity is an important

determinant of consumer appeal impacting both the colour and nutritional content of the

product.

Muscle oxidative capacity is intrinsically linked to fibre type (Brandstetter et al., 1998a;

Staron and Johnson, 1993) with type I muscle fibres having greater expression of

oxidative enzymes in cattle, guinea pigs and rabbits (Brandstetter et al., 1998a; Peter et

al., 1972) and higher levels of myoglobin in sheep (Pethick et al., 2005c), in contrast to

type II muscle fibres which express higher levels of glycolytic enzymes (Brandstetter et

al., 1998a). Therefore the activity of oxidative enzymes such as isocitrate

9

97

122

dehydrogenase (ICDH; EC 1.1.1.42) and myoglobin concentration can be used as

indicators of oxidative metabolism, an approach previously taken to study the muscle of

lamb (Gardner et al., 2006b).

Production factors such as lamb nutrition and age have been shown to influence

oxidative metabolism. A study by Gardner et al. (2006b) demonstrated the impact of

plane of nutrition in lambs, with nutritional restriction reducing muscle ICDH activity

and myoglobin concentration. In comparison the genotypic effects on these traits were

smaller, having only half (for ICDH activity) or a quarter (for myoglobin concentration)

the impact of the nutritional effects (Gardner et al., 2006b).

Animal age is also a strong driver of muscle oxidative capacity with older more mature

lambs having decreased proportions of type IIX glycolytic muscle fibres and increased

proportions of type I and type IIA fibres (Greenwood et al., 2007; Suzuki and Cassens,

1983; White et al., 1978) leading to an increased ICDH activity (Brandstetter et al.,

1998a) and myoglobin concentration (Gardner et al., 2007; Pethick et al., 2005c).

Fibre type and thus oxidative capacity can also be influenced by genetics. The

Australian sheep industry uses Australian Sheep Breeding Values (ASBVs) to select for

increased growth, muscularity and leanness, which are all likely to impact on oxidative

capacity. Selection for increased growth via an increased post weaning weight (PWWT)

has been shown to correlate with an increased mature size (Huisman and Brown, 2008).

Thus at the same weight lambs from high PWWT sires will be less mature resulting in

reduced oxidative capacity. Increased muscularity in double muscled cattle and

increased muscling in lamb, achieved by selection for an increased post weaning eye

muscle depth (PEMD), have been shown to result in a greater proportion of type IIX

glycolytic fibres (Greenwood et al., 2006c; Wegner et al., 2000). Likewise, selection for

leanness via a reduced post weaning c-site fat depth (PFAT) has been shown to increase

123

eye muscle depth and increase the proportion of type IIX fibres (Greenwood et al.,

2006c), reducing ICDH activity and myoglobin concentration (Gardner et al., 2006b).

Thus current selection for increased PWWT and PEMD and reduced PFAT are all likely

to reduce the oxidative capacity of muscle.

As a component of the Information Nucleus Flock (van der Werf et al., 2010) run by the

Australian Cooperative Research Centre for Sheep Industry Innovation numerous

measurements, including muscle ICDH activity and myoglobin concentration, were

taken on the progeny of sires with divergent ASBVs enabling the effects of genotype

and production factors to be determined. We hypothesised that increasing sire PWWT

and PEMD ASBVs and reducing sire PFAT ASBVs would all reduce the oxidative

capacity of muscle in their progeny. As production factors varied between sites, we also

hypothesised that this would cause differences in muscle oxidative capacity and that the

magnitude of these effects would be larger than the magnitude of the genotypic effects.

5.2 Materials and Methods

5.2.1 Experimental Design

The design of the Sheep CRC's Information Nucleus Flock was presented in detail by

(Fogarty et al., 2007). Each year a total of approximately 2,000 lambs were produced at

eight research sites across Australia, which represent a broad range of production

systems (Kirby NSW, Trangie NSW, Cowra NSW, Rutherglen VIC, Hamilton VIC,

Struan SA, Turretfield SA, and Katanning WA). The lambs were progeny of 267 key

industry sires, including Merino sires (Merino, Poll Merino), Maternal sires (Border

Leicester, Bond, Booroola, Coopworth, Corriedale, Dohne Merino, East Friesian, Prime

SAMM, White Dorper), and Terminal sires (Hampshire Down, Ile De France, Poll

9

97

124

Dorset, Southdown, Suffolk, Texel, White Suffolk). Lambs were reared under extensive

pasture grazing conditions and fed grain, hay or feedlot pellets when feed supply was

limited (Ponnampalam et al., 2014a) to enable them to reach an average carcass weight

of approximately 21.5 kg. The time taken to reach this weight varied between 134 and

504 days. Lambs were yarded for 6 h on the day before slaughter then transported to

one of five commercial abattoirs where they were held in lairage overnight. After

slaughter, all carcasses were subjected to medium voltage electrical stimulation (Pearce

et al., 2010) and dressed according to AUS-MEAT specifications (Anonymous, 2005).

Carcasses were chilled overnight (3-4 °C) before a wide range of carcass and meat traits

were sampled.

5.2.2 Collection of phenotypic measurements

Hot carcass weight (HCWT) was measured at slaughter, and within 5 h post-mortem a 1

g sample of m. longissimus lumborum (short loin muscle) was excised from the carcass

(over the 12th rib) for every animal. Subcutaneous fat was removed and the samples

were snap-frozen in liquid nitrogen and stored at−80 °C until subsequent ICDH

determination. The ICDH activity was determined by the method of Briand (1981),

which was adapted to run on an Olympus AU400 auto analyser. At 24 h post-mortem

the entire short loin muscle was removed and weighed (short loin muscle weight), and

additional samples for myoglobin, intramuscular fat and protein determination were

collected. Subcutaneous short loin fat was also dissected and weighed (short loin fat).

For myoglobin, a 1 g sample of short loin muscle was excised and stored at −20 °C until

myoglobin concentration determination by the method of Trout (1991). For lambs born

in 2007 only, extractable protein was measured for the ICDH and myoglobin samples

by the method of (Bradford, 1976). For intramuscular fat, a 40 g sample of loin muscle

was excised and stored at −20 °C until subsequent freeze drying using a Cuddon FD

1015 freeze dryer (Cuddon Freeze Dry, NZ). The intramuscular fat content was

125

determined using a near infrared (NIR) procedure in a Technicon Infralyser 450 (19

wavelengths) using the method described by Perry et al. (2001).

5.2.3 Data available

For the ICDH analysis muscle samples for a total of 3,178 lambs, which were the

progeny of 178 sires across 2 years were available, while for the myoglobin analysis

muscle samples for a total of 5,580 lambs, which were the progeny of 267 sires across 3

years were available.

5.2.4 Statistical analysis

ICDH activity and myoglobin concentration were analysed using linear mixed effects

models (SAS Version 9.1, SAS Institute, Cary, NC, USA). Initially, base models for

ICDH activity and myoglobin concentration were developed, which included fixed

production effects for site (representing research station site; Kirby, Trangie, Cowra,

Rutherglen, Hamilton, Struan, Turretfield, Katanning), year of birth (2007 and 2008 for

ICDH and 2007, 2008 and 2009 for myoglobin), sex (male, female), birth type–rear

type (11, 21, 22, 31, 32, 33), age of dam (2, 3, 4, 5, 6, 7 years), sire type (Merino,

Maternal, Terminal), dam breed within sire type (Merino ×Merino, Merino × Maternal,

Merino × Terminal, Border Leicester– Merino × Terminal) and kill group within site by

year. Sire identification and dam identification by year were included as random terms.

All relevant first order interactions between fixed effects were tested and removed in a

stepwise manner, by a process of backward elimination, if non-significant (P > 0.05).

To assess the relationship of ICDH activity and myoglobin concentration with aspects

of carcass phenotype traits, HCWT, short loin fat weight, short loin muscle weight and

intramuscular fat were included as covariates in the above base models. Initially HCWT

was included, and then either short loin muscle or fat weight was also included. All

models contained all relevant first order interactions between the fixed effects and

9

97

126

covariates, including a quadratic effect of the covariate, and were removed in a stepwise

manner if non-significant (P > 0.05).

To assess the impact of sire PWWT, PFAT and PEMD ASBVs on ICDH activity and

myoglobin concentration, the three ASBVs were analysed concurrently as covariates in

the base model. Additionally the carcass traits of HCWT, short loin fat weight, short

loin muscle weight and intramuscular fat were analysed one at a time within this ASBV

model. Each ASBV was also analysed as an individual covariate within the base model.

All models contained relevant first order interactions between the fixed effects and

covariates, including a quadratic effect of the covariate, and were removed in a stepwise

manner if non-significant (P > 0.05).

To assess the impact of age at slaughter on ICDH activity and myoglobin concentration,

age at slaughter was included as a covariate in the base model. The model contained all

relevant first order interactions between the fixed effects and the age at slaughter

covariate, including a quadratic effect of age, and these were removed in a stepwise

manner if non-significant (P > 0.05). The kill group within site by year term was not

included as a fixed effect in this model as within each site by year the age at slaughter

was confounded by kill group. This is because the variation in age within each kill

group was less than 10 days. Removing the kill group term as a fixed effect allowed the

effect of age at slaughter to be estimated. As kill group accounts for factors the age at

slaughter term does not two versions of the model were used to estimate the impact of

age, the first with kill group as a random term and the second without kill group as a

random term, for comparison.

For lambs born in 2007 only, protein was determined and tested as a covariate within

this subset of the data. For both ICDH activity and myoglobin concentration, protein

had a large effect, increasing ICDH activity by 1.92 μmol/min/g tissue across the 35 g

protein range (Figure 5-1) and myoglobin concentration by 3.98 mg/g tissue across the

127

60 g protein range. However, in both cases the magnitude of the other genotype and

environmental effects described within these models did not change whether the models

were corrected for protein or not. As the protein correction did not alter the magnitude

of the effects, a pragmatic decision was made to reduce assay workload by

discontinuing protein determination. Therefore protein was not measured in subsequent

years, and the analyses presented are expressed on a per gramme of tissue basis, with no

protein correction.

Figure 5-1 Linear relationship between muscle sample protein concentration (mg/g tissue) and isocitrate

dehydrogenase activity (± s.e.) in μmol/min/g tissue per gramme of protein. Each icon (o) represents an

individual lamb.

4

4.5

5

5.5

6

6.5

7

7.5

8

8.5

9

20 30 40 50 60 70 80 90

ICD

H A

ctiv

ity

(µ

mo

l/m

in/g

tis

sue)

Protein Concentration (mg/g tissue)

9

97

128

5.3 Results

5.3.1 Isocitrate Dehydrogenase

5.3.1.1 Production effects

The phenotypic linear mixed effects model utilised ICDH activity for 3,178 animals and

described 68% of the total variance. The model is presented in Table 5-1.

Table 5-1 Numerator degrees of freedom, denominator degrees of freedom and F-values for the base

models for isocitrate dehydrogenase activity (μmol/min/g tissue) and myoglobin concentration (mg/g

tissue).

Isocitrate dehydrogenase Myoglobin concentration

(μmol/min/g tissue) (mg/g tissue)

NDF, DDF F-value NDF, DDF F-value

Site 7, 532 43.32** 7, 972 228.59**

Year of birth 1, 2409 1593.09** 2, 4242 16.76**

Sex 1, 532 8.75** 1, 972 34.81**

Sire type 2, 532 3.17* 2, 972 7.10**

Dam breed (sire type) 1, 532 10.60** NA NA

Kill group (site by year) 38, 532 18.57** 64, 972 40.21**

Site by year 6, 532 4.22** 13, 972 58.95**

NDF, DDF: Numerator and denominator degrees of freedom; * P < 0.05; ** P < 0.01; NA: Not

applicable (non-significant term)

129

97

The number of progeny analysed in the base model is presented in Table 5-2.

Table 5-2 Number of progeny at each site for year of birth, sex, sire type and dam breed within sire type from the base ICDH activity model.

Year of birth Sex Sire type Dam breed within sire type

2007 2008 Female Male Maternal Merino Terminal MATM MM TM TBLM

Kirby 168 298 121 345 105 143 212 102 142 124 85

Trangie * 218 68 150 43 43 133 43 43 69 63

Cowra 290 157 143 304 65 65 317 64 64 213 101

Rutherglen 298 203 179 322 54 71 376 54 71 0 376

Hamilton 203 194 143 250 61 63 270 61 63 266 0

Struan 89 131 77 143 42 51 128 41 51 38 88

Turretfield 66 184 47 203 64 84 102 64 84 102 0

Katanning 392 409 239 560 184 161 446 184 161 442 0

*No animals were produced; MATM: Maternal x Merino; MM: Merino x Merino; TM: Terminal x Merino; TBLM: Terminal x Border Leicester-Merino

9

97

130

The raw mean and ranges for ICDH activity and phenotypic covariates are presented in

Table 5-3.

Table 5-3 Raw mean ± standard deviation, and minimum to maximum value, for isocitrate dehydrogenase

activity and the phenotypic covariates tested within the base model.

Isocitrate Hot carcass Intramuscular Short loin Short loin Slaughter

Dehydrogenase weight fat muscle weight fat weight age

(µmol/min/g tissue) (kg) (%) (g) (g) (days)

Site

Kirby 4.75 ± 1.62 26.2 ± 5.3 4.8 ± 1.0 404.8 ± 84.7 289.2 ± 166.3 365.4 ± 49.4

(1.02 - 9.01) (16.0 - 40.0) (2.8 - 9.1) (140 - 670) (30 - 865) (270 - 420)

Trangie 3.78 ± 1.10 24.6 ± 3.0 4.4 ± 0.9 373.3 ± 62.9 147.8 ± 53.0 226.7 ± 53.3

(1.67 - 7.05) (18.5 - 34.7) (2.9 - 7.8) (245 - 583) (40 - 330) (193 - 333)

Cowra 6.07 ± 1.42 23.4 ± 3.2 4.0 ± 0.8 360.1 ± 63.1 159.2 ± 63.1 193.4 ± 57.2

(2.67 - 10.29) (16.3 - 35.3) (2.3 - 7.1) (215 - 565) (28 - 385) (138 - 325)

Rutherglen 4.72 ± 1.47 21.9 ± 2.5 3.9 ± 1.0 326.4 ± 53.5 175.6 ± 56.5 266.6 ± 54.6

(1.43 - 8.71) (14.4 - 31.8) (1.9 - 7.2) (157 - 558) (47 - 429) (215 - 392)

Hamilton 5.48 ± 1.63 21.1 ± 2.6 4.3 ± 1.0 341.3 ± 55.1 167.4 ± 53.0 313.2 ± 62.3

(2.42 - 11.29) (14.2 - 32.2) (2.1 - 7.7) (189 - 525) (11 - 354) (270 - 456)

Struan 4.73 ± 1.58 20.9 ± 3.2 4.3 ± 0.9 317.4 ± 71.7 206.2 ± 84.5 280.6 ± 63.7

(2.16 - 10.46) (12.8 - 29.0) (2.2 - 8.0) (165 - 549) (59 - 531) (219 - 426)

Turretfield 5.03 ± 1.53 20.5 ± 2.8 4.2 ± 1.2 332.6 ± 59.2 252.6 ± 78.2 262.7 ± 73.2

(2.90 - 11.27) (14.6 - 29.8) (2.6 - 10.5) (211 - 496) (0 - 555) (188 - 361)

Katanning 5.58 ± 1.72 21.5 ± 3.3 4.7 ± 1.1 343.8 ± 66.6 194.9 ± 87.3 280.4 ± 96.8

(2.25 -11.40) (13.6 - 31.3) (1.5 - 9.6) (173 - 598) (15 - 560) (178 - 499)

Year of birth

2007 6.44 ± 1.35 21.8 ± 2.8 4.2 ± 1.0 339.6 ± 57.7 182.4 ± 73.2 262.9 ± 71.3

(2.81 - 11.40) (12.8 – 32.2) (1.5 - 9.1) (140 - 580) (0 - 531) (157 - 456)

2008 4.10 ± 1.06 23.2 ± 4.4 4.5 ± 1.0 359.8 ± 77.5 211.2 ± 115.9 290.0 ± 93.5

(1.02 - 9.54) (13.6 – 40.0) (1.8 - 10.5) (173 - 670) (15 - 865) (138 - 499)

Sex

Female 5.00 ± 1.67 22.7 ± 3.7 4.3 ± 0.9 357.2 ± 65.7 213.4 ± 99.6 250.2 ± 61.0

(1.02 - 11.33) (12.8 – 39.6) (1.9 - 8.4) (165 - 625) (53 - 865) (138 - 420)

Male 5.24 ± 1.66 22.5 ± 3.9 4.4 ± 1.1 348.4 ± 72.2 192.3 ± 100.7 289.9 ± 91.4

(1.36 - 11.40) (13.6 – 40.0) (1.5 - 10.5) (140 - 670) (0 - 740) (138 - 499)

Sire type

Maternal 5.15 ± 1.63 21.6 ± 3.5 4.4 ± 1.0 319.8 ± 59.1 206.1 ± 96.1 250.5 ± 64.2

(1.70 - 10.72) (14.3 - 36.6) (2.2 - 9.6) (186 - 565) (63 - 585) (138 - 420)

Merino 5.62 ± 1.64 21.3 ± 3.8 4.6 ± 1.3 330.4 ± 65.8 164.6 ± 110.7 394.4 ± 57.0

(2.42 - 11.29) (12.8 - 33.8) (1.8 - 10.5) (140 - 580) (0 - 610) (178 - 499)

Terminal 5.02 ± 1.67 23.3 ± 3.8 4.2 ± 0.9 367.5 ± 69.8 207.8 ± 94.4 245.7 ± 60.4

(1.02 - 11.40) (14.8 – 40.0) (1.5 - 8.8) (193 - 670) (53 - 865) (138 - 420)

Dam breed within sire type

MATM 5.15 ± 1.63 21.7 ± 3.5 4.4 ± 1.0 319.9 ± 59.0 206.3 ± 96.3 250.2 ± 63.9

(1.70 - 10.72) (14.3 - 36.6) (2.2 - 9.6) (186 - 565) (63 - 585) (138 - 420)

MM 5.61 ± 1.64 21.3 ± 3.8 4.6 ± 1.3 330.6 ± 65.8 164.9 ± 110.6 393.9 ± 58.2

(2.42 - 11.29) (12.8 - 33.8) (1.8 - 10.5) (140 - 580) (0 - 610) (178 - 499)

TM 5.21 ± 1.64 22.6 ± 3.6 4.3 ± 0.9 367.1 ± 66.9 204.9 ± 94.1 244.5 ± 61.5

(1.90 - 11.40) (14.8 - 38.8) (1.5 - 8.8) (195 - 598) (53 - 695) (138 - 420)

TBLM 4.65 ± 1.66 24.4 ± 3.9 4.1 ± 0.9 368.6 ± 74.5 213.7 ± 95.1 247.6 ± 58.4

(1.02 - 9.41) (16.0 – 40.0) (1.9 - 7.0) (193 - 670) (61 - 865) (138 - 420)

MATM: Maternal x Merino; MM: Merino x Merino; TM: Terminal x Merino; TBLM: Terminal x Border

Leicester - Merino

131

The average activity level of ICDH was 5.17 ± 0.06 μmol/min/g tissue and ranged from

1.02 to 11.40 μmol/min/g tissue. Wethers (5.27 ± 0.03) had on average 0.13 μmol/min/g

tissue higher ICDH activity than ewe lambs (5.14 ± 0.05) (P < 0.01). Maternal (5.27 ±

0.08) and Merino (5.27 ± 0.12) sired lambs had 0.19 μmol/min/g tissue higher ICDH

activity than Terminal (5.08 ± 0.06) sired lambs (P < 0.05). Within the Terminal sired

lambs, those from Merino dams (5.19 ± 0.07) had 0.23 μmol/min/g tissue higher ICDH

activity than lambs from Border Leicester–Merino dams (4.96 ± 0.08) (P < 0.01).

ICDH activity was greatest at the Cowra site and least at the Trangie site with activities

of 5.76 and 3.99 μmol/min/g tissue (P < 0.01). It was also greater in lambs born in 2007

compared to lambs born in 2008 with activities of 6.45 and 4.12 μmol/min/g tissue (P <

0.01). However the difference between years varied across sites by as much as 2.69

μmol/min/g tissue at Hamilton to as little as 2.11 μmol/min/g tissue at Kirby (P < 0.01)

(Figure 5-2).

Figure 5-2 The effect of site by year of birth on isocitrate dehydrogenase activity (μmol/min/g tissue) (±

s.e.) from the base model.

0

1

2

3

4

5

6

7

8

ICD

H A

ctiv

ity

(µ

mo

l/m

in/g

tis

sue)

9

97

132

Within each site and year there were marked differences between kill groups which

varied in ICDH activity by as little as 0.29 μmol/min/g tissue in Hamilton in 2008 to as

much a 2.36 μmol/min/g tissue in Katanning in 2007 (P < 0.01). ICDH activity

generally increased with older kill groups although this trend was not always consistent

across all sites. Age at slaughter was associated with an increase in ICDH activity of

0.26 μmol/min/g tissue across the 370 day age range (Figure 5-3). When kill group

within site by year was included as a random term the magnitude of the effect increased

to 1.23 μmol/min/g tissue.

Figure 5-3 Linear relationship between isocitrate dehydrogenase activity (μmol/min/g tissue) (± s.e.) and

slaughter age in days. Symbols represent predicted means for kill groups. Lines represent predicted means

for age at slaughter (± s.e.) from base model with no kill group term.

2

3

4

5

6

7

8

9

120 170 220 270 320 370 420 470

ICD

H A

ctiv

ity

(µ

mo

l/m

in/g

tis

sue)

Age at slaughter (days)

Kirby Trangie Cowra Rutherglen

Hamilton Struan Turretfield Katanning

133

When phenotypic (i.e. short loin muscle weight, short loin fat weight, intramuscular fat,

HCWT) or genotypic (PWWT, PFAT, PEMD) covariates were included in this model,

the production effects remained unchanged. The only exception to this was for sire type

which rarely impacted on ICDH activity in the presence of phenotypic or genotypic

covariates.

5.3.1.2 Association between ICDH activity and carcass phenotypic indicators

HCWT varied between 15 and 35 kg and demonstrated an association with ICDH

activity, but this was only present at the Turretfield and Kirby sites (P < 0.05). At the

Turretfield site ICDH activity increased by 1 μmol/min/g tissue across the 20 kg HCWT

range while at the Kirby site it decreased by 0.7 μmol/min/g tissue. Short loin muscle

weight varied between 200 and 550 g and demonstrated an association with ICDH

activity, but only at some sites. At the Turretfield site, ICDH activity increased by 0.88

μmol/min/g tissue across the 350 g short loin muscle weight range while at the Cowra

and Katanning sites it decreased by 0.52 and 0.59 μmol/min/g tissue (P < 0.05). This

association did not change when HCWT was included in the model. Short loin fat

demonstrated no association with ICDH activity. Intramuscular fat demonstrated a

curvilinear association with ICDH activity (P < 0.01), which increased by 0.72

μmol/min/g tissue between 2 and 6 intramuscular fat percent (Figure 5-4). The

relationship appeared to plateau beyond 6 intramuscular fat percent. The magnitude of

this effect was not changed by the inclusion of the phenotypic covariates short loin fat,

short loin muscle weight or HCWT.

9

97

134

Figure 5-4 Curvilinear relationship between intramuscular fat percentage and isocitrate dehydrogenase

activity (± s.e.) in μmol/min/g tissue. Symbols (o) represent individual lamb residuals from the

intramuscular fat model.

5.3.1.3 Effect of sire and sire breeding values

In the base model sire as a random term was significant with individual sire estimates

for ICDH activity varying between 5.04 and 5.54 μmol/min/g tissue for Merino, 4.89

and 5.60 for Maternal and 4.59 and 5.78 for Terminal sires (P < 0.01) (Figure 5-5).

When PWWT, PFAT and PEMD were included simultaneously as covariates in the

base model each had an association with ICDH activity (P < 0.01). Across the 21 kg

range, increasing PWWT was associated with an increase in ICDH activity of 0.43

μmol/min/g tissue. This effect was no longer significant when the model was corrected

for short loin muscle weight, HCWT or intramuscular fat.

0

1

2

3

4

5

6

7

8

9

10

0 2 4 6 8 10 12

ICD

H A

cti

vit

y (

µm

ol/m

in/g

ti

ssu

e)

Intramuscular Fat (%)

135

Figure 5-5 Relationship between sire estimates for isocitrate dehydrogenase activity (μmol/min/g tissue)

and subcutaneous post weaning c-site fat depth (PFAT) for Maternal, Merino and Terminal sires. Lines

represent predicted means for isocitrate dehydrogenase activity (± s.e.) for all sire types combined.

Symbols represent individual sire estimates.

Across the 5 mm range, reducing PFAT was associated with a reduction in ICDH

activity of 0.46 μmol/min/g tissue (Fig. 5). This effect was no longer significant when

the model was corrected for short loin muscle weight.

Across the 7 mm range, increasing PEMD was associated with a reduction in ICDH

activity of 0.50 μmol/min/g tissue. This effect was no longer significant when the model

was corrected for short loin muscle weight.

The magnitudes of the PWWT and PEMD responses were similar when analysed as

single covariates in the base model, however the PFAT effect was no longer significant.

4.5

4.7

4.9

5.1

5.3

5.5

5.7

5.9

-2.5 -1.5 -0.5 0.5 1.5 2.5

ICD

H A

ctiv

ity

(µ

mo

l/m

in/g

tis

sue)

Post-weaning fat depth (mm)

Maternal Merino Terminal

136

97

The mean and ranges for PWWT, PFAT and PEMD are presented in Table 5-4.

Table 5-4 Number of sires and Australian Sheep Breeding Values mean (min, max) for each sire type from the isocitrate dehydrogenase (ICDH) activity and myoglobin

concentration data.

Sire type No. of sires PFAT (mm) PEMD (mm) PWWT (kg)

ICDH Activity Maternal 37 -0.05 (-1.06, 2.56) 0.05 (-1.28, 1.82) 4.79 (-3.66, 10.49)

Merino 70 -0.25 (-1.89, 1.42) -0.18 (-2.02, 2.27) 1.45 (-5.00, 7.29)

Terminal 71 -0.82 (-2.41, 2.27) 1.12 (-1.33, 4.92) 11.92 (1.13, 16.05)

Myoglobin Concentration Maternal 61 -0.07 (-1.61, 2.56) 0.06 (-1.44, 1.82) 5.19 (-3.66, 10.49)

Merino 101 -0.22 (-1.89, 1.50) -0.10 (-2.02, 2.32) 1.81 (-5.00, 8.39)

Terminal 105 -0.85 (-2.44, 2.27) 1.07 (-1.33, 4.92) 12.06 (1.13, 18.08)

ICDH: Isocitrate dehydrogenase; PFAT: Post weaning c-site fat depth, PEMD: Post weaning eye muscle depth, PWWT: Post weaning weight

137

5.3.2 Myoglobin Concentration

5.3.2.1 Production effects

The phenotypic linear mixed effects model utilised myoglobin concentration for 5,580

animals and described 62% of the variance. The model is presented in Table 5-1. The

number of progeny analysed in the base model is presented in Table 5-5.

.

138

97

Table 5-5 Number of progeny at each site for year of birth, sex and sire type from the base myoglobin concentration model.

Year of Birth

Sex

Sire type

2007 2008 2009 Female Male Maternal Merino Terminal

Kirby

247 399 393

324 708

197 203 628

Trangie

* 216 199

145 270

76 77 262

Cowra

291 154 199

208 436

105 90 449

Rutherglen

298 213 172

325 358

88 103 492

Hamilton

198 194 175

196 368

90 80 395

Struan

296 123 177

207 389

108 67 420

Turretfield

269 185 214

207 461

136 119 411

Katanning 399 409 359 364 801 269 199 684

*No animals were produced

139

The raw mean and ranges for myoglobin concentration and phenotypic covariates are

presented in Table 5-6.

Table 5-6 Raw mean ± standard deviation and minimum to maximum values for myoglobin concentration

and the phenotypic covariates tested within the base model.

Myoglobin Hot carcass Intramuscular Short loin Short loin Slaughter

Concentration weight fat muscle weight fat weight age

(mg/g tissue) (kg) (%) (g) (g) (days)

Site

Kirby 8.18 ± 2.03 25.6 ± 5.1 4.7 ± 1.0 388.9 ± 78.6 270.6 ± 160.4 327.3 ± 56.5

(3.09 - 13.65) (15.4 - 40.0) (2.4 - 9.1) (140 - 670) (30 - 880) (235 - 420)

Trangie 6.18 ± 1.65 23.3 ± 3.3 3.8 ± 1.0 359.4 ± 62.6 122.6 ± 55.3 208.0 ± 65.3

(2.99 - 12.66) (16.2 - 34.7) (1.6 - 7.8) (228 - 583) (30 - 330) (154 - 342)

Cowra 6.13 ± 1.48 23.4 ± 3.4 3.9 ± 0.8 362.3 ± 65.2 147.7 ± 63.8 195.2 ± 54.8

(3.17 - 13.80) (14.0 - 35.3) (1.7 - 7.1) (160 - 585) (20 - 385) (138 - 325)

Rutherglen 6.69 ± 1.43 22.3 ± 2.9 3.8 ± 0.9 335.0 ± 58.6 190.2 ± 70.3 270.0 ± 50.9

(3.12 - 13.09) (14.2 - 31.8) (1.9 - 7.2) (157 - 558) (46 - 590) (215 - 392)

Hamilton 6.25 ± 1.69 21.8 ± 3.1 4.2 ± 0.9 341.5 ± 59.1 176.3 ± 73.0 295.6 ± 57.5

(2.15 - 11.33) (13.4 - 32.2) (2.1 - 7.7) (160 - 525) (10 - 412) (229 - 456)

Struan 5.81 ± 1.50 22.0 ± 3.3 4.1 ± 1.0 336.7 ± 71.8 195.6 ± 112.7 260.4 ± 49.7

(2.48 - 10.75) (12.8 - 31.4) (1.6 - 8.2) (155 - 590) (21 - 543) (210 - 426)

Turretfield 5.83 ± 1.58 21.6 ± 2.7 4.1 ± 1.0 335.6 ± 58.3 193.9 ± 107.4 259.9 ± 51.2

(2.24 - 11.57) (14.6 - 29.8) (1.9 - 10.5) (190 - 558) (0 - 555) (188 - 361)

Katanning 6.81 ± 1.66 22.3 ± 3.6 4.6 ± 1.1 364.3 ± 72.4 212.6 ± 88.1 274.6 ± 84.4

(2.95 -15.62) (13.6 - 35.8) (1.5 - 9.8) (173 - 661) (15 - 560) (178 - 499)

Year of birth

2007 6.50 ± 1.97 22.4 ± 3.1 4.3 ± 1.0 349.0 ± 62.3 203.2 ± 87.3 257.7 ± 63.3

(2.15 - 15.62) (12.8 – 35.8) (1.5 - 9.1) (140 - 661) (0 - 543) (157 - 456)

2008 6.76 ± 1.85 23.0 ± 4.4 4.5 ± 1.0 359.1 ± 76.5 205.7 ± 115.4 288.9 ± 91.0

(2.48 - 13.66) (13.6 – 40.0) (1.8 - 10.5) (173 - 670) (15 - 865) (138 - 499)

2009 6.70 ± 1.70 23.4 ± 4.0 4.0 ± 1.1 360.7 ± 71.3 187.6 ± 126.0 260.3 ± 58.6

(2.99 - 13.80) (13.4 – 40.0) (1.6 - 9.8) (155 - 620) (10 - 880) (154 - 359)

Sex

Female 6.45 ± 1.61 23.1 ± 3.7 4.2 ± 1.0 361.6 ± 64.5 212.5 ± 113.1 250.5 ± 57.2

(2.47 - 12.88) (12.8 – 40.0) (1.9 - 9.8) (165 - 625) (31 - 880) (138 - 420)

Male 6.74 ± 1.95 22.8 ± 3.9 4.3 ± 1.1 353.3 ± 72.8 192.2 ± 109.1 277.6 ± 78.9

(2.15 - 15.62) (13.4 – 40.0) (1.5 - 10.5) (140 - 670) (0 - 740) (138 - 499)

Sire type

Maternal 6.55 ± 1.76 21.9 ± 3.5 4.4 ± 1.1 324.1 ± 60.5 194.9 ± 105.9 252.7 ± 60.4

(2.24 - 14.41) (14.0 - 36.6) (1.6 - 9.7) (155 - 565) (20 - 585) (138 - 420)

Merino 8.22 ± 2.08 21.1 ± 3.6 4.5 ± 1.2 324.6 ± 66.0 157.1 ± 107.2 373.2 ± 59.7

(2.83 - 15.62) (12.8 - 33.8) (1.8 - 10.5) (140 - 580) (0 - 610) (178 - 499)

Terminal 6.27 ± 1.58 23.7 ± 3.8 4.1 ± 1.0 373.0 ± 68.1 210.2 ± 110.0 246.8 ± 55.3

(2.15 - 12.88) (14.8 – 40.0) (1.5 - 9.8) (193 - 670) (28 - 880) (138 - 420)

9

97

140

The average myoglobin concentration was 6.65 ± 0.02 mg/g tissue and ranged from

2.15 to 15.62. Ewe lambs (6.89 ± 0.04) had on average 0.23 mg/g tissue higher

myoglobin concentration than wethers (6.66 ± 0.03) (P < 0.01). Maternal (6.87 ± 0.06)

and Merino (6.84 ± 0.09) sired lambs had higher myoglobin concentrations than

Terminal (6.61 ± 0.04) sired lambs (P < 0.01).

Myoglobin concentration was greatest at the Kirby site and least at the Struan site with

8.19 and 6.07 mg/g tissue (P < 0.01). It was also greater in lambs born in 2008 or 2009

than lambs born in 2007 with values of 6.93, 6.79 and 6.58 mg/g tissue (P < 0.01).

However these year differences varied across sites by as much as 2.09 mg/g tissue at

Hamilton to as little as 0.18 at Trangie (P < 0.01) (Figure 5-6).

Figure 5-6 The effect of site by year of birth on myoglobin concentration (mg/g tissue) (± s.e.) from the

base model.

0

1

2

3

4

5

6

7

8

9

My

og

lob

in C

on

cen

tra

tion

(m

g/g

tis

sue)

141

Within each site and year there were marked differences between kill groups which

varied by as little as 0.40 mg/g tissue in Struan in 2008 to as much as 4.77 in Kirby in

2008 (P < 0.01). Myoglobin concentration generally increased with later kill groups and

this result aligned well with the age at slaughter effect which was associated with an

increase in myoglobin concentration of 5.01 mg/g tissue across the 370 day age range

(Figure 5-7). When kill group within site by year was included as a random term the

magnitude of the effect increased to 5.73 mg/g tissue.

Figure 5-7 Linear relationship between myoglobin concentration (mg/g tissue) and slaughter age in days.

Symbols represent predicted means for kill groups. Lines represent predicted means for age at slaughter

(± s.e.) from base model with no kill group term.

When phenotypic (i.e. short loin muscle weight, short loin fat weight, intramuscular fat,

HCWT) or genotypic (PWWT, PFAT, PEMD) covariates were included in this model,

the production effects remained unchanged.

3

4

5

6

7

8

9

10

11

100 150 200 250 300 350 400 450 500 550

My

og

lob

in C

on

cen

tra

tio

n (

mg

/g t

issu

e)

Age at slaughter (days)

Kirby Trangie Cowra Rutherglen

Hamilton Struan Turretfield Katanning

9

97

142

5.3.2.2 Association between myoglobin concentration and carcass phenotypic

indicators

HCWT demonstrated an association with myoglobin concentration increasing by 1.56

mg/g tissue across the 20 kg HCWT range (P < 0.01). This effect varied between sites

and sire types (P < 0.01). At the Kirby, Cowra, Turretfield, Hamilton, Katanning,

Rutherglen, Struan and Trangie sites myoglobin concentration increased by 0.66, 1.23,

1.28, 1.46, 1.72, 1.87, 2.08 and 2.22 mg/g tissue across the 20 kg range in HCWT.

Progeny with Merino sires had the largest increase in myoglobin concentration across

the HCWT range of 1.90 mg/g tissue, while progeny with Maternal and Terminal sires

increased by 1.82 mg/g tissue and 0.97 mg/g tissue across the HCWT range.

Short loin muscle weight did not demonstrate an association with myoglobin

concentration.

Short loin fat weight varied between 50 and 400 g and demonstrated an association with

myoglobin concentration, increasing by 0.83 mg/g tissue across the 350 g short loin fat

weight range (P < 0.01). This effect varied between sire types with progeny of Merino

sires increasing by 1.29 mg/g, progeny of Terminal sires increasing by 0.64 mg/g tissue

and progeny of Maternal sires increasing by 0.57 mg/g tissue across the short loin fat

weight range.

Intramuscular fat demonstrated a curvilinear association with myoglobin concentration

(P < 0.01). Between 2 and 6 intramuscular fat percent myoglobin concentration

increased by 0.61 mg/g tissue (Figure 5-8). The relationship appeared to plateau beyond

6% intramuscular fat. The association varied between sire types with increases of 0.35,

0.66 and 0.84 mg/g tissue in the progeny of Terminal, Maternal and Merino sires

respectively (P < 0.05).

143

Figure 5-8 Curvilinear relationship between intramuscular fat percentage and myoglobin concentration (±

s.e.) in mg/g tissue. Symbols (o) represent individual lamb residuals from the intramuscular fat model.

When ICDH activity was included as a covariate in the base model it was a significant

predictor of myoglobin concentration, which increased by 0.64 mg/g tissue across the

10.05 μmol/min/g tissue ICDH activity range (P < 0.01).

0

2

4

6

8

10

12

14

0 2 4 6 8 10 12

Myo

glo

bin

Co

ncen

tra

tio

n (

mg

/g t

issu

e)

Intramuscular Fat (%)

9

97

144

5.3.2.3 Effect of sire and sire breeding values

In the base model sire as a random term was significant with individual sire estimates

for myoglobin concentration varying between 6.44 and 7.24 mg/g tissue for Maternal,

6.26 and 7.56 for Merino and 6.03 and 7.49 for Terminal sires (P < 0.01) (Figure 5-9).

When PWWT, PFAT and PEMD were included simultaneously as covariates in the

base model each had an association with myoglobin concentration (P < 0.05). Across

the 23 kg range, increasing PWWT was associated with an increase in myoglobin

concentration of 0.50 mg/g tissue. This effect varied between sites being greatest at the

Hamilton and Struan sites where myoglobin concentration increased by 0.89 and 1.00

mg/g tissue across the PWWT range (P < 0.05). These effects were no longer

significant when the model was corrected for short loin muscle weight, short loin fat

weight or intramuscular fat.

Across the 6 mm range, reducing PFAT was associated with a reduction in myoglobin

concentration of 0.67 mg/g tissue and this effect varied between sexes with a reduction

of 0.26 mg/g tissue in wethers and a reduction of 1.07 mg/g tissue in ewe lambs. There

was no impact to the magnitude of the PFAT effect when short loin muscle weight and

short loin fat weight were included separately as covariates. The magnitude of the

PFAT effect reduced to 0.58 mg/g tissue when intramuscular fat was included as a

covariate.

There was also an association between PEMD and myoglobin concentration however

this was only present within the Merino and Terminal sire types (P < 0.01). Across the

6 mm range, increasing PEMD was associated with a reduction in myoglobin

concentration of 0.49 mg/g tissue in the progeny of Terminal sires and an increase of

1.36 in the progeny of Merino sires (Figure 5-9). When short loin muscle weight, short

145

loin fat weight and intramuscular fat were included separately as covariates in the

ASBV model there was a reduction of the PEMD effect in progeny of Merino sires to

1.22 mg/g tissue and increase to 0.57 mg/g tissue in the progeny of Terminal sires.

When these ASBVs were analysed as single covariates in the base model the magnitude

of the PWWT response increased to 0.74 mg/g tissue and the PFAT response reduced to

0.57 mg/g tissue. The magnitude of the PEMD response remained unchanged.

The mean and ranges for PWWT, PFAT and PEMD are presented in Table 5-4.

146

97

Figure 5-9 Relationship between sire estimates for myoglobin concentration (mg/g tissue) and post weaning eye muscle depth (PEMD) for Maternal, Merino and Terminal sires.

Lines represent predicted means for myoglobin concentration (± s.e.) for each sire type. Symbols (o) represent individual sire estimates.

1

147

5.4 Discussion

5.4.1 Association between carcass breeding values and oxidative

capacity

Aligning with our hypothesis, the progeny of sires with an increased PEMD had a

reduced oxidative capacity, as indicated by reduced ICDH activity across all sire types

and reduced myoglobin concentration within Terminal sired lambs. This aligns well

with previous studies, with Gardner et al. (2006a) demonstrating reduced myoglobin

concentration in response to PEMD with in Poll Dorset x Merino lambs, and Gardner et

al. (2007) demonstrating reduced ICDH activity within Poll Dorset x Merino and

Border Leicester x Merino lambs. A further study by Gardner et al. (2006b) also

showed a reduced myoglobin concentration within the progeny of all sire types with

increased PEMD but failed to show this effect on ICDH activity. This is likely to be due

to the limited number of sires used, 9 compared to 267 in this study, where the PEMD

range of 5 mm compared to 7 mm in this study may not have been divergent enough to

elicit a response.

The association between PEMD and reduced oxidative capacity may be due to a shift in

muscle fibre type with increased PEMD leading to a greater proportion of type IIX

fibres (Greenwood et al., 2006c) which have been associated with increased glycolytic

(Gardner et al., 2006b) and reduced oxidative enzyme activities (Brandstetter et al.,

1998a).

The use of the PEMD breeding value within the Australian sheep industry has been

shown to increase loin muscle weight (Gardner et al., 2010). As there is an association

between increased phenotypic muscle weight and reduced muscle oxidative capacity it

could be concluded that the PEMD effect on muscle oxidative capacity is merely a

148

correlate of its impact on short loin muscle weight. However, in this study there was no

phenotypic association between short loin muscle weight and myoglobin concentration,

and a variable association with ICDH activity.

Furthermore, correcting the models for short loin muscle weight had no impact on the

association between PEMD and myoglobin concentration. This inconsistent overall

phenotypic association between short loin muscle weight and muscle oxidative capacity

is unusual and appears to suggest that PEMD is affecting myoglobin concentration

through a mechanism other than the phenotype that it directly delivers.

As hypothesised, progeny of sires with a reduced PFAT had a reduced oxidative

capacity, as indicated by reduced ICDH activity and myoglobin concentration. Previous

studies have demonstrated similar results with Gardner et al. (2006b) and (Gardner et

al., 2007) both demonstrating reduced ICDH activity in lambs with low PFAT sires, and

Greenwood et al. (2006c) showing higher levels of glycolytic type IIX fibres. Although

the Gardner et al. (2006b) study showed a reduction in ICDH activity there was no

impact on myoglobin concentration. The study was performed on wether lambs which

in our study had a smaller response to a reduction in PFAT than ewe lambs. As the

Gardner et al. (2006b) study included only 56 wethers this may not have been enough to

show an effect.

Intramuscular fat is reduced in the progeny of sires with a reduced PFAT in the animals

in this study (Pannier et al., 2014c) however inclusion of intramuscular fat in the ASBV

model did not diminish the impact of PFAT on ICDH activity or myoglobin

concentration, indicating that the impact of PFAT is not delivered through its correlated

impact on intramuscular fat.

149

Contrary to our initial hypothesis, isocitrate dehydrogenase activity and myoglobin

concentration were both increased in the progeny of sires with an increased PWWT.

Our hypothesis was based on the assumption that the progeny of high PWWT sires

would be less mature and therefore have a lower oxidative capacity at a given slaughter

age due to their larger mature size (Huisman and Brown, 2008). However, this effect

has clearly been over-ridden by other factors. The samples for ICDH activity and

myoglobin concentration were taken from the short loin, and the inclusion of short loin

muscle weight as a phenotypic covariate nullified the impact of the PWWT ASBV on

both ICDH activity and myoglobin concentration. Evidence from computed tomography

scanning of a subset of these same animals suggests an altered tissue distribution within

high PWWT lambs. In particular, the weight of muscle within the saddle region of the

carcass was 7% higher at any given carcass weight (Gardner et al., 2012). As such the

increased growth potential of these tissues would have required greater energy

producing capacity, of which the TCA cycle is a vital component. As such it is possible

that the impact of PWWT on muscle oxidative capacity is limited to the saddle region.

5.4.2 Associations with carcass phenotypic measurements

Heavier carcasses were associated with increased myoglobin concentration, although

there was no effect of HCWT on ICDH activity except at the Turretfield and Kirby

sites. In this study HCWT at a constant slaughter age (i.e. kill group corrected for age as

per Figure 5-3 and Figure 5-7) reflects growth rate, which is most likely delivered

through variation in environment and nutrition. As such the faster growing lambs are

likely to be closer to their mature size, aligning well with the myoglobin results which

have previously been shown to increase as lambs approach maturity (Gardner et al.,

2007). Yet this increase in myoglobin concentration may have occurred without a

150

corresponding increase in muscle oxidative capacity, as there was no increase in ICDH

activity. Alternatively the higher growth rate may have been due to a larger mature size,

in which case faster growing animals slaughtered at the same age would be less mature

(Butterfield, 1988), but this seems unlikely as it would have resulted in proportionately

less oxidative muscle tissue (Brandstetter et al., 1998a), the opposite of what was found

here.

The relationship between muscle oxidative capacity and phenotypic carcass muscularity

(short loin muscle weight) and adiposity (short loin fat weight) was also assessed.

Computed tomography was used to determine whole carcass muscle and fat weights on

a subset of the animals used in this study, and these measurements demonstrated that

short loin muscle weight was strongly correlated with whole carcass muscle weight

(0.84), and short loin fat weight was strongly correlated with whole carcass fat weight

(0.83) (Anderson, unpublished data). As such these measurements are good indicators

of whole carcass muscularity and adiposity. As mentioned, the muscularity indicator,

short loin muscle weight, demonstrated no association with myoglobin and an

inconsistent relationship with ICDH. Alternatively the carcass adiposity indicator, short

loin fat weight, demonstrated a consistently positive association with myoglobin,

although no association with ICDH. Thus phenotypic carcass composition demonstrates

some alignment with the oxidative capacity of muscle, albeit an inconsistent

association. The phenotypic correlation between ICDH activity and myoglobin

concentration in this experiment was 0.14 (Mortimer et al., 2014) which may explain

some of the variation in response of the oxidative indicators.

In contrast, a far more consistent association was evident for the relationship between

muscle oxidative capacity and intramuscular fat. ICDH activity and myoglobin

concentration both increased with higher levels of intramuscular fat. Similar effects

151

have been demonstrated in cattle, where the genotypes which expressed higher levels of

intramuscular fat also had a more oxidative muscle type, characterised by increased

ICDH activity (Hocquette et al., 2003; Jurie et al., 2007; Pethick et al., 2005c).

Likewise, this has also been demonstrated in pigs (Essén-Gustavsson et al., 1994) and

rabbits (Gondret et al., 1998) in which slow twitch oxidative muscles had higher levels

of intramuscular fat. More recently, Pannier et al. (2014b) demonstrated a positive

association between intramuscular fat and iron and zinc levels.

The association between intramuscular fat and oxidative capacity appears to be

independent of maturity, or whole carcass adiposity and muscularity. This was

demonstrated by also correcting the ICDH activity and myoglobin models for HCWT,

short loin muscle weight and short loin fat weight. In all cases the intramuscular fat

association with ICDH activity or myoglobin concentration remained unchanged. As

such the association between intramuscular fat and muscle oxidative capacity appears to

be more than just a correlate of whole carcass fatness or muscularity. It seems unlikely

that intramuscular fat would directly mediate the oxidative capacity of adjacent muscle

cells, however this result may highlight the presence of some localised control that

ensures that more oxidative muscle cells have greater amounts of stored oxidative fuel

(fat) in the near vicinity (i.e. intramuscular adipocytes).

5.4.3 Production effects

As hypothesised, the site at which lambs were reared impacted on oxidative capacity

and the magnitude of the effect of site was larger than most other genotypic and

production effects. Between sites, the average age at slaughter varied which may

explain part of the differences between sites. This was particularly evident in the

myoglobin analysis where the Kirby site had the oldest average age at slaughter as well

as the highest myoglobin concentration, and of the four sites with the oldest average

152

age, three sites also had higher than average myoglobin concentrations. In line with

these findings, a previous study showed a significant increase in Chromameter a* values

indicating an increase in the relative redness of meat as lambs age (Hopkins et al.,

2007). In the ICDH analysis the site differences appeared to be less age dependant,

however this was not entirely unexpected as ICDH activity demonstrated a weaker

association with age than myoglobin concentration (Figure 5-3 and Figure 5-7). In

addition the effect of age at slaughter on ICDH activity was small compared to the

magnitude of the production or breeding value effects while the effect on myoglobin

concentration was large compared to all other effects indicating that age was a critical

factor in determining myoglobin concentration but not ICDH activity. When kill group

was included as a random term the effect of age on ICDH activity was larger. As age

and kill group are strongly confounded within the Information Nucleus Flock the true

magnitude of the impact of age is likely to lie somewhere between the two results.

The differences between sites are also likely to be partly explained by nutrition.

Nutritional variation has been shown to impact on muscle fibre type (Greenwood et al.,

2006b; Moody et al., 1980) and subsequently may alter oxidative capacity. Oxidative

capacity was highest at the Kirby and Cowra sites after correcting for age and weight

(maturity), perhaps indicating an increased plane of nutrition at these sites. However,

the Information Nucleus Flock was not designed to control nutrition and thus

conclusions regarding the nutritional impact on oxidative capacity are difficult to assess.

There was a difference in oxidative capacity between wether and ewe lambs although

ICDH activity was greater in wethers and myoglobin concentration was greater in ewe

lambs. Gardner et al. (2007) demonstrated an increased ICDH activity in wether lambs

when compared to ewe lambs. Consistent with these results, Greenwood et al. (2007)

sampled the same lambs as used in the Gardner et al. (2007) study and demonstrated

153

that wether lambs had reduced proportions of type IIX myofibres. While higher

myoglobin levels in ewe lambs were unexpected, a previous study by Ledward and

Shorthose (1971) found a similar result, with ewe lambs having a 10% higher

myoglobin concentration than wethers. The reason for myoglobin concentration to be

increased in ewe lambs is unclear, thus gender demonstrates some alignment with the

oxidative capacity of muscle however the association is inconsistent.

The progeny of Terminal sires had reduced oxidative capacity compared Maternal or

Merino sired lambs. This is due to the selection pressure on muscling in these sires

(Hall et al., 2002) which has also been shown in this study to reduce oxidative capacity.

Muscle oxidative capacity was influenced by both production and genotypic factors,

with site having the largest effect overall. The difference between sites was about three

and a half times the magnitude of the effect of PWWT, PFAT or PEMD, and about 10

times the magnitude of the other production effects on ICDH activity and myoglobin

concentration. The only exception to this was for year of birth which had a larger effect

than site on ICDH activity. Therefore, while sire ASBVs do impact on oxidative

capacity in their progeny, the effect is small compared to the impact of the environment.

154

5.5 Conclusion

This study demonstrated that both genotypic and production factors had significant

effects on oxidative capacity. Selection for leanness and muscling by selection of sires

with a reduced PFAT and increased PEMD resulted in a reduction in oxidative capacity

in muscle of the progeny. This effect was also seen in the progeny of Terminal sires,

which have received greater selection pressure for lean growth and muscling. Selection

for lambs with a larger mature size using sires with an increased PWWT resulted in

increased oxidative capacity however this may be an effect which is localised to the loin

region. An increase in intramuscular fat is linked to an increase in oxidative capacity

that is independent of its association with PFAT or whole body adiposity or

muscularity. Oxidative capacity also increased with age however the effect was more

evident on myoglobin concentration than ICDH activity. Increased growth rate, which is

heavily influenced by environmental effects at each site, was associated with an

increase in oxidative capacity however the effect of growth rate was smaller than the

effect of site itself. In conclusion, while selection for leanness, muscling or mature size

using ASBVs impacts on oxidative capacity, these effects and those of production

factors such as sex and sire type are small compared to the impact of the environment.

155

Chapter 6. Lamb growth increases intramuscular

fat levels and reduces shear force in the longissimus

muscle

6.1 Introduction

Lamb eating quality is becoming increasingly important to consumers (Pethick et al.,

2006a). It is strongly influenced by intramuscular fat due to its associations with

consumer sensory scores. As intramuscular fat increases the sensory scores for

tenderness in lamb increase (Pannier et al., 2014a), which is supported by a negative

phenotypic correlation between intramuscular fat and shear force (Mortimer et al.,

2014). There is also an increase in consumer overall liking, a score out of 100 based on

the Meat Standards Australia sensory system, with an increase in intramuscular fat

(Pannier et al., 2014a) and a reduction with a reduced intramuscular fat (Hopkins et al.,

2006). Therefore changes in intramuscular fat will impact eating quality through

changes in shear force and consumer sensory scores.

Lamb growth may impact intramuscular fat although this has not been thoroughly

investigated in sheep. In cattle, pre finishing growth restrictions result in reduced levels

of intramuscular fat at the end of finishing, with the development of intramuscular fat

linked to carcass fatness (Pethick et al., 2004). Pannier et al. (2014c) used hot carcass

weight as a proxy for whole of life growth in sheep to suggest that intramuscular fat is

increased with lamb growth. However, the accuracy of the hot carcass weight effect as

an estimate of the impact of growth is untested against actual growth rates. As hot

carcass weight is dependent on age at slaughter it is less representative of growth and

156

more representative of environmental influences as animals get older and is likely to be

a poor proxy for whole of life growth. Furthermore, hot carcass weight is unable to

differentiate the effect during different phases of growth.

As intramuscular fat increases with hot carcass weight (Pannier et al., 2014c) and shear

force and consumer sensory scores are influenced by intramuscular fat (Pannier et al.,

2014a) quantification of the true impact of lamb growth is important to highlight any

potential effects on meat quality. We hypothesise that intramuscular fat and overall

liking will increase with growth and shear force will reduce.

6.2 Materials and Methods

6.2.1 Experimental Design

The design of the Sheep Cooperative Resource Centre (CRC) Information Nucleus

Flock was presented in detail by Fogarty et al. (2007) and van der Werf et al. (2010).

Each year, for 5 years between 2007 and 2011, a total of approximately 3,500 lambs

were produced at eight research sites across Australia, which represent a broad range of

production systems (Kirby NSW, Trangie NSW, Cowra NSW, Rutherglen VIC,

Hamilton VIC, Struan SA, Turretfield SA, and Katanning WA). The lambs were the

progeny of sires selected to represent the key breeds used in Australia. These included

Merino sires (Merino, Poll Merino), Maternal sires (Bond, Border Leicester, Booroola,

Coopworth, Corriedale, Dohne Merino, East Friesian, Prime SAMM, White Dorper),

and Terminal sires (Hampshire Down, Ile De France, Poll Dorset, Southdown, Suffolk,

Texel, White Suffolk). Lambs were reared under extensive pasture grazing conditions

and fed grain, hay or feedlot pellets when feed supply was limited (Ponnampalam et al.,

2014a).

157

6.2.2 Collection of phenotypic measurements

Lambs were weighed at birth, at weaning (90 days of age) and approximately every two

weeks thereafter. Hot carcass weight was measured at slaughter. After slaughter all

carcasses were subjected to medium voltage electrical stimulation (Pearce et al., 2010)

and dressed according to AUS-MEAT specifications (Anonymous, 2005). Carcasses

were chilled overnight (3-4°C). Twenty four hours post slaughter the entire short loin

muscle was removed from the carcass saddle region (cut between the 12th and 13th

ribs) and the m. longissimus lumborum muscle dissected. The subcutaneous short loin

fat was weighed (short loin fat weight). The topside muscle (semimembranosus) was

also excised from the carcass. The two muscles were vacuum packed and stored at 2°C

to age for 5 days. Five steaks from each muscle of 15 mm thick were cut, vacuum

packed and frozen at −20°C for subsequent sensory testing. The contralateral loin and

topside of the carcass were also excised and used for subsequent objective meat quality

measurements of intramuscular fat and shear force.

For intramuscular fat, a 40 g sample of loin muscle was excised and stored at -20ºC

until subsequent freeze drying using a Cuddon FD 1015 freeze dryer (Cuddon Freeze

Dry, NZ). Intramuscular fat content, expressed as a percentage, was determined using a

near infrared procedure (NIR) in a Technicon Infralyser 450 (19 wavelengths) using the

method described by Perry et al. (2001). NIR readings were validated with chemical fat

determinations using solvent extraction.

For shear force, a 65 g sample of loin muscle was excised and vacuum packed. After 5

days of aging the samples were stored at -20ºC until subsequent shear force testing.

Samples were cooked at 71ºC in plastic bags in water baths for 35 minutes before being

cooled in running water for 30 minutes. Shear force was measured on six cores from

each loin sample using a Lloyd texture analyser (Model LRX, Lloyd Instruments,

158

Hampshire UK) with a Warner-Bratzler shear blade fitted as described by Hopkins et al.

(2010).

Consumer overall likability was determined using a score scale out of 100 was based on

the Meat Standards Australia sensory system using untrained consumers and has been

explained in detail by Pannier et al. (2014a). Loin and topside steaks were also assessed

for tenderness, juiciness, liking of flavour and liking of odour. Each steak (5 steaks per

muscle) was halved to form ten test samples which were uniformly grilled using a Silex

griller to a medium degree of doneness (internal temperature of 65°C). Samples were

rested for 2 min before being served. Each consumer tested an initial sample of average

quality with an overall liking score of 72 (Pethick et al., 2006b) followed by 6 test

samples (3 loin and 3 topside). Samples were allocated using a Latin square design to

balance position and carry over effects (Thompson et al., 2005). In total 60 consumers

were recruited in each consumer group to taste 36 samples to obtain 10 responses per

sample. The sensory testing protocol is described in detail by Thompson et al. (2005)

and Watson et al. (2008).

6.2.3 Data available

The range for weights and growth rates, derived in Chapter 4, are presented in Table 6-1

and the number of samples analysed, years of data and number of sires with post

weaning weight (PWWT), post weaning c-site fat depth (PFAT) and post weaning eye

muscle depth (PEMD) Australian sheep breeding values (ASBVs) for each trait are

presented in Table 6-2. The PFAT and PEMD breeding values were based on live

ultrasound measurements while the PWWT breeding values were based on live weight.

All ASBVs were sourced from Sheep Genetics, Australia’s national genetic evaluation

database for sheep (Brown et al., 2007) and a correlation matrix is presented in Table

6-3. Sire ASBVs ranged by as much as 23 kg, 5 mm and 8 mm for PWWT, PFAT and

159

PEMD respectively. The number of kill groups within a site and year varied between 2

and 5, and ages varied from 5 weeks to 40 weeks apart.

160

97

Table 6-1 Weight and growth rate ranges for intramuscular fat, shear force and overall liking.

Variable Minimum Maximum Range Mean Standard Deviation

Intramuscular Fat (n = 8,254) Birth Weight (kg) 1.70 9.80 8.10 5.05 1.05

Weight 100 days (kg) 3.38 60.76 57.38 31.58 8.01

Weight 150 days (kg) 11.33 68.78 57.45 37.26 7.52

Growth Rate 100 days (g/day) -170.09 547.70 717.79 134.61 80.92

Growth Rate 150 days (g/day) -170.23 393.49 563.72 99.77 67.12

Weight Change 100 to 150 days (kg) -7.25 20.01 27.26 5.68 3.26

Shear Force (n = 8,433) Hot Carcass Weight (kg) 13.20 40.00 26.80 23.13 3.75

Birth Weight (kg) 1.70 10.10 8.40 5.08 1.06

Weight 100 days (kg) 3.38 60.76 57.38 31.67 7.93

Weight 150 days (kg) 11.33 68.78 57.45 37.39 7.45

Growth Rate 100 days (g/day) -170.09 496.22 666.31 135.20 81.21

Growth Rate 150 days (g/day) -107.37 393.49 500.86 100.53 65.86

Weight Change 100 to 150 days (kg) -5.01 20.01 25.02 5.72 3.22

Overall Liking (n = 1,411) Hot Carcass Weight (kg) 15.00 40.00 25.00 23.70 3.86

Birth Weight (kg) 1.98 9.50 7.52 4.83 1.15

Weight 100 days (kg) 3.71 57.97 54.26 29.14 7.60

Weight 150 days (kg) 11.33 59.36 48.03 34.76 6.90

Growth Rate 100 days (g/day) -89.03 416.19 505.22 127.34 89.04

Growth Rate 150 days (g/day) -107.37 366.30 473.67 103.46 77.56

Weight Change 100 to 150 days (kg) -5.01 18.79 23.80 5.63 4.04

161

Table 6-2 Number of animals, years of data and number of sires available for intramuscular fat, shear

force and overall liking.

Number

of animals

Years

of data

Number

of sires

Maternal

sires

Merino

sires

Terminal

sires

Intramuscular Fat 8,254 5 428 83 163 182

Shear Force 8,433 5 425 83 159 183

Overall Liking 1,411 2 156 39 52 65

Table 6-3 Phenotypic correlations (standard errors) of post weaning sire breeding values for weight, fat

depth and eye muscle depth among Maternal, Merino and Terminal sires.

PWWT PFAT PEMD

Maternal

PWWT 1.00 -0.18 (0.09) -0.06 (0.62)

PFAT -0.18 (0.09) 1.00 0.47 (0.00)

PEMD -0.06 (0.62) 0.47 (0.00) 1.00

Merino

PWWT 1.00 0.23 (0.00) 0.29 (0.00)

PFAT 0.23 (0.00) 1.00 0.76 (0.00)

PEMD 0.29 (0.00) 0.76 (0.00) 1.00

Terminal

PWWT 1.00 -0.03 (0.69) -0.26 (0.00)

PFAT -0.03 (0.69) 1.00 0.39 (0.00)

PEMD -0.26 (0.00) 0.39 (0.00) 1.00

6.2.4 Statistical analysis

Intramuscular fat, shear force and overall liking were analysed using linear mixed

effects models (SAS Version 9.1, SAS Institute, Cary, NC, USA). Overall liking in the

loin and topside were analysed separately. For intramuscular fat and overall liking base

statistical models were developed and then compared to the base models previously

published, based on less data (Pannier et al., 2014a; Pannier et al., 2014c). In the case of

162

shear force a base statistical model was developed for this paper (data not shown). The

base models included fixed production effects for site (representing research station site;

Kirby, Trangie, Cowra, Rutherglen, Hamilton, Struan, Turretfield, Katanning), year of

birth (2007, 2008, 2009, 2010, 2011), birth type – rear type (11, 21, 22, 31, 32, 33), age

of dam (2, 3, 4, 5, 6, 7, 8 years), sire type (Merino, Maternal, Terminal), sex-dam breed

within sire type (Female x Merino x Terminal, Male x Merino x Terminal, Female x

Border Leicester-Merino x Terminal, Male x Border Leicester-Merino x Terminal) and

kill group within site by year. Two base models for overall liking were developed using

the average value of the 10 consumer responses, one for loin and one for the topside cut.

Sire identification and dam identification by year were included as random terms.

Consumer group was included as a random term in the overall liking analyses. All

relevant first order interactions between fixed effects were tested and removed in a

stepwise manner, by a process of backward elimination, if non-significant (P > 0.05).

To assess the impact of growth on intramuscular fat, shear force and consumer overall

liking measures of weight and growth rate were included separately as covariates in the

above base models. For weight these included weight at birth and estimates of weight at

100 and 150 days. The effect of weight on the trait being analysed reflected the

cumulative impact of weight from conception to its point of measurement. For growth

rate, we examined the impact of weight change between 100 and 150 days (weaning and

post weaning) as well as instantaneous growth rate, estimated as the first derivative of

the growth curve, at 100 or 150 days. The advantage of also including the instantaneous

growth rate measures was that the simple weight change between 100 and 150 days

provides no ability to determine where growth rate was having the greater effect, closer

to weaning or post weaning time points. Furthermore, given the smoothed nature of a

growth curve the tangent, or growth rate, will be similar on days that are close to the

day being examined. Therefore, the impact of instantaneous growth rate on the trait in

163

question can be extrapolated to a limited degree either side of its point of estimation,

providing information on the likely change in association between live weight and the

trait in question closer to these time points. Lamb weights and growth rates were

obtained in Chapter 4, on the same animals which had intramuscular fat, shear force and

overall liking measured.

To determine the accuracy of hot carcass weight as a proxy for growth the impact of hot

carcass weight on intramuscular fat, as previously assessed by (Pannier et al., 2014c),

was compared to the impact weight at 100 and 150 days. For shear force and overall

liking, hot carcass weight was included as a covariate in both the shear force base model

and the overall liking base model to determine its impact. To determine whether the

impact of growth on the meat traits was due to adiposity at the site of sampling or whole

carcass adiposity rather than growth, intramuscular fat and short loin fat weight were

included separately as covariates in the base models for each trait. All models contained

all relevant first order interactions between the fixed effects and covariates, including a

quadratic effect of the covariate, and were removed in a stepwise manner if non-

significant (P > 0.05).

The impact of PWWT, PFAT and PEMD ASBVs on intramuscular fat and overall

liking has previously been assessed (Pannier et al., 2014a; Pannier et al., 2014c). To

determine the impact of PWWT, PFAT and PEMD ASBVs on shear force all three

ASBVs were analysed concurrently as covariates in the base model. Subsequently,

weight at 150 days or growth rate at 150 days were included separately in the base

models for each trait, along with the three ASBVs. The aim of this analysis was to

determine whether the impact of these ASBVs on intramuscular fat, shear force and

overall liking were delivered via their impact on growth. The impact of the ASBV for

birth weight (BWT) and weaning weight (WWT) on intramuscular fat were determined

164

by including the ASBVs separately in the intramuscular fat base model. Subsequently

weight at birth was included in the base model, along with BWT and weight at 100 days

was included in the base model, along with WWT. The aim of this analysis was to

determine whether the impact of BWT or WWT on intramuscular fat, were delivered

via their impact on growth. All ASBV models contained relevant first order interactions

between the fixed effects and covariates, including a quadratic effect of the covariate,

and were removed in a stepwise manner if non-significant (P > 0.05).

Partial correlations of weight and growth rate with intramuscular fat, shear force and

overall liking were determined using a multivariate analysis (Version 9.1, SAS Institute,

Cary, NC, USA). The analysis included the fixed effects and their interactions described

for the base models.

165

6.3 Results

6.3.1 Intramuscular fat

6.3.1.1 Base model

The base model for this study (not shown) utilised intramuscular fat for 8,254 animals,

and has previously been published by Pannier et al. (2014c) who analysed a subset of

this data (n = 5,642). There were minor variations to the Pannier et al. (2014c) model

with the inclusion of small effects for dam age, a site by sire type interaction, and the

absence of a site by sex effect.

6.3.1.2 Birth weight

Birth weight demonstrated an association with intramuscular fat reducing by 0.85%

across the 8.1 kg weight range (P < 0.05) (Table 6-4). This association varied in twin

born lambs and between sire types. Intramuscular fat reduced by 0.14% units for each

kg of birth weight in twin born lambs. The variation between sire types is detailed in

Table 6-5.

.

Table 6-4 The magnitude of effect (difference between the highest and lowest values) of lamb weight and

growth rate on intramuscular fat or on shear force across the entire weight or growth rate range.

Trait Intramuscular Fat (%) Shear Force (N)

Birth Weight (kg) -0.85 -

Weight 100 days (kg) 1.09 -10.43

Weight 150 days (kg) 0.49 -6.32

Growth Rate 100 days (g/day) -0.95 7.01

Growth Rate 150 days (g/day) -1.24 12.89

Weight Change 100 to 150 days (kg) -1.17 12.90

166

Table 6-5 Variation in slopes between sire types for the impact of birth weight, weight at 100 days and

weight at 150 days on intramuscular fat.

Intramuscular Fat

Trait Sire type Slope Trait -Min. Trait - Max. Magnitude

Birth weight Maternal -0.13 1.90 9.50 -0.97

Merino -0.12 2.20 8.60 -0.80

Terminal -0.06 1.70 9.80 -0.51

Weight 150 Maternal 0.001 15.46 55.81 0.04

Merino 0.008 13.93 51.20 0.31

Terminal 0.016 11.33 68.78 0.94

Min: Minimum value for the variable; Max: Maximum value for the variable; Magnitude: The total

magnitude of the effect of the trait on intramuscular fat.

6.3.1.3 Effect of weight at 100 and 150 days

Weight at 100 days was associated with an increase in intramuscular fat of 0.02% units

for each kg of weight, which was equivalent to 1.09% across the 57 kg weight range (P

< 0.01). For weight at 150 days, intramuscular fat increased by 0.01% units for each kg

of weight, which was equivalent to 0.49% across the 57 kg weight range. This

association varied between sire types and is detailed in Table 6-5.

6.3.1.4 Effect of change in weight between 100 and 150 days and growth rate at

100 and 150 days

Growth rate measured at 100 and 150 days or the change in weight measured between

these time points were all associated with reducing intramuscular fat (P < 0.01). Change

in weight between 100 and 150 days varied between -7.25 and 20.01 kg, growth rate at

100 days varied between -170.09 and 547.70 g/day, and growth rate at 150 days varied

between -170.23 and 393.49 g/day and across these ranges intramuscular fat diminished

by 1.17%, 0.95% and 1.24%. The association between change in weight between 100

167

and 150 days and intramuscular fat varied between sites and years. The variation

between sites reduced intramuscular fat by as much as 0.09% units and as little as

0.04% units per kg of weight change. The variation between years reduced

intramuscular fat by as much as 0.07% units and as little as 0.05% units per kg of

weight change.

6.3.1.5 Effect of adiposity

When correcting the growth models for short loin fat weight (P < 0.01) the magnitude

of the growth covariates diminished by a third to a half.

6.3.1.6 Effect of ASBVs

When BWT, or WWT were included separately as covariates in the base model, neither

had an association with intramuscular fat. When the BWT and WWT models were

corrected for weight at birth and weight at 100 days respectively, there was no change to

the lack of association of the breeding values with intramuscular fat.

When the PWWT, PFAT and PEMD ASBV model was corrected for weight at 150

days or growth rate at 150 days the magnitude of the effect of the breeding values

reported by (Pannier et al., 2014c) did not change.

6.3.2 Shear force

6.3.2.1 Base model

The base model utilised shear force for 8,433 animals. When phenotypic (i.e. weight,

growth rate, hot carcass weight) or genotypic (PWWT, PFAT, PEMD) covariates were

included in this model, the production effects generally remained unchanged. The

168

exceptions were sire type which was not significant when the base model was corrected

for hot carcass weight, and dam age which became significant when the base model was

corrected for birth weight, weight at 100 days and growth rate at 150 days.

6.3.2.2 Hot carcass weight

When hot carcass weight was included as a covariate in the base model it demonstrated

a curvilinear association with shear force reducing it by 0.34 N for each kg of hot

carcass weight, which was equivalent to a 9.06 N across the 27 kg hot carcass weight

range (P < 0.01). This association varied between sex-dam breed within sire type

combinations. Within the Terminal sired lambs, shear force varied between males and

female lambs with Border Leicester-Merino compared to Merino dams. The reduction

in shear force was 0.05 N less in male compared to female lambs from Merino dams

while it was 0.25 N greater in male compared to female lambs from Border Leicester-

Merino dams (P < 0.05).

6.3.2.3 Birth weight

Birth weight demonstrated an association with shear force however this was only

present in progeny of 4 year old dams where there was a small increase of 0.15 N for

each kg of birth weight (P < 0.05).

6.3.2.4 Effect of weight at 100 and 150 days

Weight at 100 days was associated with a reduction in shear force of 0.18 N for each kg

of weight, which was equivalent to 10.43 N across the 57 kg weight range. The

association was reduced by about 35% with a reduction in shear force of 0.11 N for

each kg of weight at 150 days, which was equivalent to 6.32 N across the 57 kg weight

range.

169

6.3.2.5 Effect of change in weight between 100 and 150 days and growth rate at 100

and 150 days

Growth rate measured at 100 and 150 days or the change in weight measured between

these time points were all associated with increasing shear force (P < 0.01). Change in

weight between 100 and 150 days varied between -5.01 and 20.01 kg, growth rate at

100 days varied between -170.09 and 496.22 g/day, and growth rate at 150 days varied

between -107.37and 393.49 g/day and across these ranges shear force increased by

12.90 N, 7.01 N and 12.89 N. The association between growth rate at 100 days and

shear force varied between years increasing shear force by as much as 0.02 N and as

little as 0.01 N per g of increase growth rate.

6.3.2.6 Effect of adiposity

When correcting the growth models for short loin fat weight (P < 0.01) or intramuscular

fat (P < 0.01) the magnitude of the growth covariates diminished by a third to a half,

except for weight at 150 days which was no longer had a significant association with

shear force when correcting for short loin fat weight.

6.3.2.7 Effect of ASBVs

When PWWT, PFAT and PEMD were included simultaneously as covariates in the

base model, only PFAT had an association with shear force (P < 0.01). Shear force was

increased by 0.60 N for each mm reduction in PFAT, which equated to an increase in

shear force of 3.02 N across the 5 mm PFAT range.

When the PWWT, PFAT and PEMD ASBV model was corrected for weight at 150

days the magnitude of the breeding value effects did not change. When the ASBV

model was corrected for growth rate at 150 days PFAT was no longer significant.

170

6.3.3 Overall liking

6.3.3.1 Base model

The base model for this study utilised overall liking for 1,411 animals, and has

previously been published by Pannier et al. (2014a).

6.3.3.2 Effect of weight

When hot carcass weight was included as a covariate in the base model it demonstrated

no association with overall liking in the loin or topside. Birth weight, weight at 100 days

and weight at 150 days also demonstrated no association with overall liking.

6.3.3.3 Effect of growth rate

Weight change between 100 and 150 days and growth rate at 100 and 150 days

demonstrated no association with overall liking in the loin or topside.

6.3.3.4 Effect of ASBVs

When the PWWT, PFAT and PEMD ASBV model reported by Pannier et al. (2014a)

was corrected for weight at 150 days or growth rate at 150 days there was no variation

in the magnitude of the breeding values on overall liking in the loin or topside.

6.3.4 Phenotypic correlations

The phenotypic correlations of the meat traits with weight and growth rate are presented

in Table 6-6. Intramuscular fat had a negative correlation with all measures of growth

except weigh at 150 days. Shear force was negatively correlated with weight from

weaning onwards and positively correlated with growth rate. There were no significant

correlations between overall liking and growth.

171

97

Table 6-6 Phenotypic correlations of intramuscular fat, shear force and overall liking with lamb weight and growth rate. Standard error values in brackets. Significant correlations are

in bold.

Birth weight Weight 100 days Weight 150 days Growth rate 100 days Growth rate 150 days

Intramuscular Fat -0.06 (0.00) 0.10 (0.00) -0.06 (0.00) -0.07 (0.00) -0.11 (0.00)

Shear Force 0.01 (0.47) -0.09 (0.00) -0.04 (0.00) 0.06 (0.00) 0.12 (0.00)

Overall liking (Loin) 0.01 (0.64) 0.01 (0.81) -0.02 (0.44) 0.01 (0.68) -0.06 (0.04)

Overall liking (Topside) 0.02 (0.50) -0.01 (0.77) -0.03 (0.22) 0.01 (0.69) -0.06 (0.03)

172

6.4 Discussion

6.4.1 Intramuscular fat

Intramuscular fat had a negative association with birth weight after which the

association inverted and intramuscular fat had a strong positive association with weight

at 100 days (Table 6-4). The association between intramuscular fat and growth rate was

similar in magnitude to the association between intramuscular fat and weight at 100

days however it was negative and strengthened from 100 to 150 days. This resulted in a

diminished association between weight and intramuscular fat at 150 days, which was

less than half that observed at 100 days. In general, these phenotypic associations

aligned well with the phenotypic correlations between intramuscular fat and weight and

growth rate (Table 6-6).

To determine whether the change in intramuscular fat with growth was simply a

reflection of its association with whole carcass adiposity the growth models were

corrected for short loin fat weight. In general, this reduced the association between

intramuscular fat and growth by half and in the case of weight at 100 days completely

accounted for this association. This suggests that the impact of growth on intramuscular

fat is largely delivered through the associated increase in whole body fatness as animals

get older. Alternatively, there was still significant variation that was not associated with

whole body fatness.

The positive association between weight at 100 and 150 days and intramuscular fat

aligned well with our hypothesis, and provides more conclusive evidence to confirm the

earlier suggestions of Pannier et al. (2014c) that intramuscular fat is linked to growth.

This suggestion was based on the observed positive association between intramuscular

fat and carcass weight at slaughter in lambs of a similar age, although was proposed

173

with some uncertainty given that carcass weight at slaughter is potentially influenced by

factors other than just whole of life growth. What is less clear is why the phenotypic

association between intramuscular fat and hot carcass with weight in the Pannier et al.

(2014c) study was greater than the magnitude seen for weight at 100 or 150 days in this

study, particularly when the association with weight was diminishing over time. In part

this may be because weight early in life also reflects changes in non-carcass tissues,

which are not represented by hot carcass weight. There is no significant phenotypic

correlation between intramuscular fat and hot carcass weight (Mortimer et al., 2014)

which makes the increased phenotypic association with hot carcass weight is difficult to

explain.

Although contrasting with the positive association between intramuscular fat and post

weaning weight, the negative association with birth weight was not entirely unexpected.

A similar phenomenon has been described in pigs (Beaulieu et al., 2010; Gondret et al.,

2006; Rehfeldt et al., 2008), and is hypothesised to be due to the slow and inefficient

growth of the pigs resulting in reduced lean percentages at slaughter. However, the

apparent inversion of this association between birth and weaning highlights this phase

of growth as a key period in which growth rate may modulate intramuscular fat

development, with the magnitude of the association between intramuscular fat and

growth increasing during this period by as much as 1.98 intramuscular fat percentage

units. This aligns with research indicating that adipocyte development is fixed relatively

early in life (Kirkland and Dobson, 1997), as has been demonstrated in sheep (McPhee

et al., 2008), pigs (Beaulieu et al., 2010) and cattle (Pethick et al., 2004), with greater

adipogenic and lipogenic gene expression in young cattle with increased intramuscular

fat (Wang et al., 2009). Alternatively the inversion of the association between birth and

weaning may reflect a poor association between birth weight and intramuscular fat due

174

to weight close to birth being controlled by a number of factors that are not linked to

maturity, such as dam milk production.

Increased growth rate in lambs at 100 and 150 days was associated with reduced levels

of intramuscular fat. Post weaning weight and adult weight are positively correlated

(Huisman and Brown, 2008) so lambs growing faster at 100 and 150 days are likely to

be growing to an increased mature size. Therefore, they are proportionately less of their

mature size at 150 days resulting in reduced levels of intramuscular fat.

The impact of weight at birth and 150 days on intramuscular fat varied between sire

types (Table 6-5). At birth, the effect of weight on intramuscular fat was strongest in

progeny of Maternal sires. When the model was corrected for short loin fat weight the

variation between sire types was no longer present. This suggests that the variation

between sire types was simply a reflection of whole body fatness. At 150 days, the

effect of weight on intramuscular fat was strongest in the progeny of Terminal sires.

When the model was corrected for short loin fat weigh the variation between sire types

was still present, and the magnitude of effect was again largest in Maternal sires. This

may indicate these animals have higher levels of intramuscular fat at maturity,

independent of whole body adiposity.

6.4.2 Shear force

Shear force was reduced in lambs that were heavier at 100 and 150 days, yet at birth

there was no association. The lack of relationship between lamb birth weight and shear

force is consistent with findings in pigs (Bérard et al., 2010) and cattle (Greenwood et

al., 2006a). This transition from no main effect of weight at birth to a large negative

effect of weight at 100 days highlights early life pre weaning growth as an important

developmental period impacting on shear force. Similarly, early life pre weaning growth

175

is also a key developmental period for intramuscular fat deposition. Given the strong

association between shear force and intramuscular fat (Mortimer et al., 2014; Pannier et

al., 2014a) it is not surprising that these traits are impacted during the same period of

growth.

In contrast to the association with weight, the impact of growth rate at 100 and 150 days

was opposite, with lambs of higher growth rates having increased shear force. This also

explains why the magnitude of the effect of weight on shear force diminished by 35%

between 100 and 150 days (Table 6-4). Although the mechanism for this growth rate

association is unclear, the changing associations across the life of the animal align well

with the phenotypic correlations (Table 6-6) evident in these same animals.

Shear force is reduced with increased intramuscular fat in lamb (Pannier et al., 2014a)

and pigs (van Laack et al., 2001). As intramuscular fat increases with weight, a

concurrent reduction in shear force with weight at 100 and 150 days was expected.

When the growth models were corrected separately for intramuscular fat and short loin

fat weight both measures of fatness generally reduced the association between shear

force and growth by a third to a half. This suggests that the impact of growth on shear

force is largely delivered through the associated increase in local or whole body fatness,

although there was still significant variation that was not associated with whole body

fatness.

Published associations between animal growth and shear force vary. In sheep, several

studies have demonstrated an increased shear force with weight, the opposite to what

was found in this study, however none of these studies were controlled for age (Beriain

et al., 2000; Sañudo et al., 1996; Teixeira et al., 2005). In cattle, neither weight at

slaughter (Lucero-Borja et al., 2014) or average daily gain (Calkins et al., 1987;

Moloney et al., 2000), both representing growth rate of a given period, have an

176

association with shear force. Perry and Thompson (2005) showed no overall impact of

growth path on shear force in beef cattle which aligns with the findings of Greenwood

et al. (2006a).

Although not a live weight measure itself hot carcass weight, and its association with

shear force, was taken into account due to its strong phenotypic correlation with live

weight at slaughter. Therefore, what is less clear is why the phenotypic association

between shear force and slaughter weight, as indicated by the association between shear

force and hot carcass weight in this study, was greater than the magnitude seen for

weight at 100 or 150 days, particularly when the association with weight was

diminishing over time. Furthermore, there was no significant phenotypic correlation

between shear force and hot carcass weight (Mortimer et al., 2014) so the increased

phenotypic association with hot carcass weight is difficult to explain. Shear force is

influenced by intramuscular fat so, as expected, these findings parallel the findings

between intramuscular fat and growth.

6.4.3 Overall liking

Contrary to our hypothesis, overall liking showed no association with lamb growth,

contrasting with the findings of Fisher et al. (2000) who demonstrated that lambs from

heavier breeds had improved overall liking scores. This lack of effect is all the more

surprising given that growth impacted on intramuscular fat and shear force, both of

which have been shown to influence overall liking scores (Hopkins and Mortimer,

2014; Pannier et al., 2014a). In the Pannier et al. (2014a) study, which used the same

lambs as used in the present study, the magnitude of the impact of shear force and

intramuscular fat on overall liking represented around 12 - 14% of the average overall

liking score. Thus other factors must play a role in determining overall liking, beyond

just intramuscular fat and shear force.

177

While the associations of intramuscular fat and shear force with overall liking did not

amount to an effect on overall liking it is possible, based on the phenotypic associations,

that this will change over time. Furthermore, changes in overall liking with lamb growth

may not have been captured due to the highly variable nature of consumer data.

There was no phenotypic association between overall liking and slaughter weight, as

indicated by the lack of association between overall liking and hot carcass weight, in

this study. This is in contrast to the positive correlation between hot carcass weight and

overall liking in the Lambe et al. (2009) study. In the Lambe et al. (2009) study on two

lamb breeds, the correlation was only significant in the loin, and not in the leg. Hence

there was inconsistency between cuts and the lack of correlation in the leg aligns with

the lack of effect in in the topside in the present study. The lack of a relationship

between overall liking and lamb growth was reflected in the phenotypic correlations

between overall liking and growth, which were generally not significant (Table 6-6).

The published impact of PWWT, PFAT and PEMD on intramuscular fat (Pannier et al.,

2014c), shear force and overall liking (Pannier et al., 2014a) did not change when

corrected for weight at 150 days or growth rate at 150 days. The effects of PFAT and

PEMD did not change with correction for weight or growth rate at 150 days indicating

that they are not mechanistically driven through the impact of growth, as evidenced by

the lack of a PWWT effect. The impact of birth weight and weight at 100 days on

intramuscular fat did not vary with the inclusion of ASBVs for birth and weaning

weights indicating that the impact of weight at birth and 100 days on intramuscular fat

is not due to genotype.

178

6.5 Conclusion

This study demonstrated that lamb growth impacts on both intramuscular fat and shear

force, and had no impact on consumer overall liking. The key period for modulation of

intramuscular fat and shear force by growth was in the pre weaning period with the

magnitude of the responses of both traits to weight reducing after weaning.

Intramuscular fat may also be influenced by in-utero growth as the association with

weight inverts between birth and weaning. Lambs heavier at birth had the lowest

intramuscular fat and lambs heavier at weaning had the highest intramuscular fat. A

third to a half of the association between lamb growth and either intramuscular fat or

shear force was delivered through an associated increase in whole carcass adiposity.

This indicated that a large portion of the impact of growth on these two traits is

delivered through the associated increase in whole body fatness as animals get older.

However, there was still significant variation that was not associated with whole body

fatness, indicating that growth was still having an impact on intramuscular fat and shear

force which was independent of whole carcass adiposity. Currently, lamb growth is not

impacting overall liking. However the associations between overall liking and both

intramuscular fat and shear force are such that growth can be expected to impact overall

liking as a result of continual selection for growth. Hot carcass weight, as a proxy for

whole of life growth, overestimated the impact of growth on intramuscular fat and shear

force and was unable to describe the impact of growth during different phases of

growth. Hot carcass weight varies with age at slaughter, while growth at weaning and

post weaning do not, contributing to the poor alignment of hot carcass weight and

growth effects.

179

Chapter 7. Lamb growth increases oxidative

capacity and mineral content in the longissimus muscle

7.1 Introduction

The nutritional value of meat is a key driver of the consumer decision making process

(Pethick et al., 2006a). Currently, the level of minerals such as iron and zinc are

sufficient to claim lamb meat as a good source for the majority of the population

(Pannier et al., 2014b; Pannier et al., 2010). These mineral levels are influenced by

muscle fibre type with reduced iron and zinc in muscles exhibiting low oxidative fibre

types and therefore having a low oxidative capacity (Kondo et al., 1991; Pearce et al.,

2009). Oxidative capacity is indicated by levels of the oxidative enzyme isocitrate

dehydrogenase (ICDH; EC 1.1.1.42) and by myoglobin concentration which are both

elevated in oxidative type I and type IIA muscle fibres (Brandstetter et al., 1998a;

Pethick et al., 2005c). Therefore, changes in fibre type with lamb growth may impact

muscle oxidative capacity as well as iron and zinc concentrations.

The impact of lamb growth on muscle fibre type with subsequent changes in oxidative

capacity and mineral content has not been thoroughly investigated. Kelman et al. (2014)

used hot carcass weight at a given age as a proxy for whole of life growth to suggest

that oxidative capacity is increased with lamb growth rate. Given the association

between oxidative capacity and iron and zinc concentrations (Pannier et al., 2014b), it

seems reasonable that iron and zinc will also increase with increased growth. Aligning

with this suggestion, Pannier et al. (2014b) demonstrated in these same lambs an

180

association between increasing iron and zinc concentration and heavier hot carcass

weight.

The use of hot carcass weight as a proxy for whole of life growth is based on the

assumption that lamb growth is linear. However, lamb growth is sigmoidal with rapid

exponential growth during early life followed by a deceleration and then a late tapering

of growth close to adult weight (Hendrick et al., 1989). Thus hot carcass weight

measured at a fixed time point only is a crude representation of growth. Therefore a

more detailed analysis of growth path is required to understand the impact of this on

oxidative capacity and mineral content and thus consumer acceptability.

The surface browning of meat at retail display slowly increases as red myoglobin

oxidises forming “brown” metmyoglobin. Browning is accelerated when type I muscle

fibres (O'keeffe and Hood, 1982; Renerre and Labas, 1987) and thus oxidative capacity

(Calnan et al., 2014) increase. Therefore, any changes to muscle fibre type with growth

may also impact meat browning. Furthermore meat browning is increased with

increasing intramuscular fat due to the auto-oxidation of the lipid accelerating the

oxidisation of myoglobin and the formation of brown pigment (Faustman et al., 2010).

As oxidative capacity and intramuscular fat increase with hot carcass weight (Kelman et

al., 2014; Pannier et al., 2014c) meat browning is likely to increase with increased

growth.

Lambs from the Information Nucleus Flock (van der Werf et al., 2010) (van der Werf et

al., 2010) run by the Australian Cooperative Research Centre (CRC) for Sheep Industry

Innovation were measured for live weight, markers of oxidative capacity, mineral

content and retail meat browning. Live weights of these lambs were used to model

growth (Chapter 4) allowing growth at different phases of lamb development and the

association with the muscle measurements to be assessed. Given the associations with

181

hot carcass weight we hypothesised that increased growth would be associated with

higher myoglobin concentration, iron concentration and zinc concentration, increased

activity of ICDH and accelerated meat browning.

7.2 Materials and Methods

7.2.1 Experimental Design

The design of the Sheep CRC’s Information Nucleus Flock was presented in detail by

Fogarty et al. (2007) and van der Werf et al. (2010). Each year, for 5 years between

2007 and 2011, a total of approximately 3,500 lambs were produced at eight research

sites across Australia, which represent a broad range of production systems (Kirby

NSW, Trangie NSW, Cowra NSW, Rutherglen VIC, Hamilton VIC, Struan SA,

Turretfield SA, and Katanning WA). The lambs were the progeny of sires selected to

represent the key breeds used in Australia. These included Merino sires (Merino, Poll

Merino), Maternal sires (Bond, Border Leicester, Booroola, Coopworth, Corriedale,

Dohne Merino, East Friesian, Prime SAMM, White Dorper), and Terminal sires

(Hampshire Down, Ile De France, Poll Dorset, Southdown, Suffolk, Texel, White

Suffolk). Lambs were reared under extensive pasture grazing conditions and fed grain,

hay or feedlot pellets when feed supply was limited (Ponnampalam et al., 2014a).

7.2.2 Collection of phenotypic measurements

Lambs were weighed at birth, at weaning (90 days of age) and approximately every two

weeks thereafter. Hot carcass weight was measured at slaughter. After slaughter all

carcasses were subjected to medium voltage electrical stimulation(Pearce et al., 2010)

and dressed according to AUS-MEAT specifications (Anonymous, 2005). Carcasses

182

were chilled overnight (3-4°C). After slaughter and within 5 h post–mortem, a 1 g

sample of m. longissimus lumborum (short loin muscle) was excised from the carcass

(over the 12th

rib) of every animal. Subcutaneous fat was removed and the samples were

snap-frozen in liquid nitrogen and stored at -80C until subsequent ICDH

determination. The ICDH activity was determined by the method of (Briand, 1981),

which was adapted to run on an Olympus AU400 auto analyser. At 24 h post-mortem

the entire short loin muscle was removed from the carcass and the subcutaneous short

loin fat was dissected and weighed (short loin fat weight). A 40 g sample of short loin

muscle was excised and stored at -20ºC until subsequent freeze drying using a Cuddon

FD 1015 freeze dryer (Cuddon Freeze Dry, NZ). Approximately 0.2 g of dry matter was

collected for iron and zinc concentration analysis.

These were prepared according to the USEPA method 200.3 (USEPA, 1991) and iron

and zinc concentrations were determined using a Vista AX CCD simultaneous ICP-AES

(Varian Australia Pty Ltd). Intramuscular fat content, expressed as a percentage, was

measured on the 40 g sample of loin muscle. Intramuscular fat was determined using a

near infrared procedure (NIR) in a Technicon Infralyser 450 (19 wavelengths) using the

method described by Perry et al. (2001). NIR readings were validated with chemical fat

determinations using solvent extraction.

For myoglobin, a 1 g sample of short loin muscle was excised and stored at -20C until

myoglobin concentration determination by the method of Trout (1991). To determine

meat browning a 50 mm long by 30 mm deep section was excised from the loin,

vacuum packed in clear gas impermeable plastic and aged in a chiller at 3-4°C for 5

days. The loin samples were then incised perpendicular to the long axis to expose a

fresh surface and wrapped with oxygen-permeable polyvinyl chloride film. The samples

were then placed under simulated retail display conditions of temperature and light as

183

described by Calnan et al. (2014). After 3 days of display light reflectance colour

measures were taken of each loin sample through the film within the chiller using a

Hunterlab spectrophotometer XE Plus (Cat. No. 6352, model No. 45/0-L, aperture of

3.18 cm). A ratio of red oxymyoglobin pigment to brown metmyoglobin pigment was

calculated as (% reflectance at 630 nm) / (% reflectance at 580 nm) or R630/R580.

Higher ratios indicate increased redness with lower values representing increased

surface browning. Two spectrophotometric readings of each sample were taken at 90

degrees on the horizontal plane and the ratios were then averaged.

7.2.3 Data available

The ranges for weights and growth rates, derived in Chapter 4, are presented in Table

7-1 and the number of samples analysed, years of data and number of sires with post

weaning weight (PWWT), post weaning c-site fat depth (PFAT) and post weaning eye

muscle depth (PEMD) Australian sheep breeding values (ASBVs) for each trait are

presented in Table 7-2. The PFAT and PEMD breeding values were based on live

ultrasound measurements while the PWWT breeding values were based on live weight.

All ASBVs were sourced from Sheep Genetics, Australia’s national genetic evaluation

database for sheep (Brown et al., 2007) and a correlation matrix is presented in Table

7-3. Sire ASBVs ranged by as much as 23 kg, 5 mm and 8 mm for PWWT, PFAT and

PEMD respectively. The number of kill groups within a site and year varied between 2

and 5, and ages varied from 5 weeks to 40 weeks apart.

.

184

97

Table 7-1 Weight and growth rate ranges for isocitrate dehydrogenase (ICDH) activity, myoglobin concentration, iron concentration, zinc concentration and meat browning.

Variable Minimum Maximum Range Mean Standard Deviation

ICDH Activity (n = 4,571) Birth Weight (kg) 1.70 9.80 8.10 5.04 1.01

Weight 100 days (kg) 3.38 57.97 54.59 31.37 8.19

Weight 150 days (kg) 13.93 60.20 46.27 36.94 7.46

Growth Rate 100 days (g/day) -170.09 547.70 717.79 132.16 74.06

Growth Rate 150 days (g/day) -170.23 393.49 563.72 97.93 66.29

Weight Change 100 to 150 days (kg) -7.25 16.95 24.20 5.57 3.05

Myoglobin Concentration (n = 8,987) Birth Weight (kg) 1.90 10.10 8.20 5.08 1.08

Weight 100 days (kg) 3.38 60.76 57.38 31.67 7.93

Weight 150 days (kg) 11.33 68.78 57.45 37.39 7.48

Growth Rate 100 days (g/day) -170.09 547.70 717.79 137.33 81.17

Growth Rate 150 days (g/day) -170.23 393.49 563.72 99.39 65.80

Weight Change 100 to 150 days (kg) -7.25 20.01 27.26 5.73 3.20

Iron Concentration (n = 8,434) Birth Weight (kg) 1.90 9.80 7.90 5.05 1.04

Weight 100 days (kg) 3.38 60.76 57.38 31.54 7.97

Weight 150 days (kg) 11.33 68.78 57.45 37.25 7.48

Growth Rate 100 days (g/day) -170.09 547.70 717.79 135.60 81.23

Growth Rate 150 days (g/day) -143.78 393.49 537.27 99.82 66.49

Weight Change 100 to 150 days (kg) -5.01 20.01 25.02 5.70 3.24

Zinc Concentration (n = 7,236) Birth Weight (kg) 1.90 9.80 7.90 5.03 1.05

Weight 100 days (kg) 3.38 57.97 54.59 31.32 8.00

Weight 150 days (kg) 11.33 60.79 49.46 36.94 7.36

Growth Rate 100 days (g/day) -170.09 547.70 717.79 131.27 78.05

Growth Rate 150 days (g/day) -170.23 393.49 563.72 100.41 67.92

Weight Change 100 to 150 days (kg) -7.25 20.01 27.26 5.62 3.28

Meat Browning (n = 5,033) Birth Weight (kg) 1.70 10.10 8.40 5.18 1.07

Weight 100 days (kg) 7.33 60.76 53.43 33.12 7.81

Weight 150 days (kg) 13.93 68.78 54.85 39.35 7.33

Growth Rate 100 days (g/day) -170.09 496.22 666.31 149.71 80.23

Growth Rate 150 days (g/day) -170.23 393.49 563.72 106.66 66.78

Weight Change 100 to 150 days (kg) -7.25 19.31 26.56 6.23 3.25

185

97

Table 7-2 Number of animals, years of data and number of sires available for isocitrate dehydrogenase (ICDH) activity, myoglobin concentration, iron concentration, zinc

concentration and meat browning.

Number of animals Years of data Number of sires Maternal sires Merino sires Terminal sires

ICDH Activity 4, 571 3 258 60 93 105

Myoglobin Concentration 8, 987 5 429 83 163 183

Iron Concentration 8, 434 5 428 83 163 182

Zinc Concentration 7, 236 4 336 75 126 135

Meat Browning 5, 933 5 416 82 151 183

186

Table 7-3 Phenotypic correlations (standard errors) of post weaning sire breeding values for weight, fat

depth and eye muscle depth among Maternal, Merino and Terminal sires.

PWWT PFAT PEMD

Maternal

PWWT 1.00 -0.18 (0.09) -0.06 (0.62)

PFAT -0.18 (0.09) 1.00 0.47 (0.00)

PEMD -0.06 (0.62) 0.47 (0.00) 1.00

Merino

PWWT 1.00 0.23 (0.00) 0.29 (0.00)

PFAT 0.23 (0.00) 1.00 0.76 (0.00)

PEMD 0.29 (0.00) 0.76 (0.00) 1.00

Terminal

PWWT 1.00 -0.03 (0.69) -0.26 (0.00)

PFAT -0.03 (0.69) 1.00 0.39 (0.00)

PEMD -0.26 (0.00) 0.39 (0.00) 1.00

7.2.4 Statistical analysis

ICDH activity and myoglobin, iron and zinc concentrations were analysed using linear

mixed effects models (SAS Version 9.1, SAS Institute, Cary, NC, USA). Initially, base

statistical models were developed and then compared to the base models previously

published, based on less data, for ICDH activity, myoglobin concentration (Kelman et

al., 2014), iron and zinc concentrations (Pannier et al., 2014b) and meat browning

(Calnan et al., 2014). The base models included fixed production effects for site

(representing research station site; Kirby, Trangie, Cowra, Rutherglen, Hamilton,

Struan, Turretfield, Katanning), year of birth (2007, 2008, 2009, 2010, 2011), birth type

– rear type (11, 21, 22, 31, 32, 33), age of dam (2, 3, 4, 5, 6, 7, 8 years), sire type

(Merino, Maternal, Terminal), sex-dam breed within sire type (Female x Merino x

Terminal, Male x Merino x Terminal, Female x Border Leicester-Merino x Terminal,

Male x Border Leicester-Merino x Terminal) and kill group within site by year. Sire

identification and dam identification by year were included as random terms. All

187

relevant first order interactions between fixed effects were tested and removed in a

stepwise manner, by a process of backward elimination, if non-significant (P > 0.05).

To assess the impact of growth on intramuscular fat, shear force and consumer overall

liking measures of weight and growth rate were included separately as covariates in the

above base models. For weight these included weight at birth and estimates of weight at

100 and 150 days. The effect of weight on the trait being analysed reflected the

cumulative impact of weight from conception to its point of measurement. For growth

rate, we examined the impact of weight change between 100 and 150 days (weaning and

post weaning) as well as instantaneous growth rate, estimated as the first derivative of

the growth curve, at 100 or 150 days. The advantage of also including the instantaneous

growth rate measures was that the simple weight change between 100 and 150 days

provides no ability to determine where growth rate was having the greater effect, closer

to weaning or post weaning time points. Furthermore, given the smoothed nature of a

growth curve the tangent, or growth rate, will be similar on days that are close to the

day being examined. Therefore, the impact of instantaneous growth rate on the trait in

question can be extrapolated to a limited degree either side of its point of estimation,

providing information on the likely change in association between live weight and the

trait in question closer to these time points. Lamb weights and growth rates were

obtained in Chapter 4, on the same animals which had these traits measured.

To determine the accuracy of hot carcass weight as a proxy for growth the impact of hot

carcass weight on isocitrate dehydrogenase activity and myoglobin concentration, as

previously assessed by Kelman et al. (2014), iron concentration and zinc concentration,

as previously assessed by Pannier et al. (2014b) and meat browning, as previously

assessed by Calnan et al. (2014) was compared to the impact of weight at 100 and 150

days. To determine whether the impact of growth on the meat traits was due to adiposity

188

at the site of sampling or whole carcass adiposity rather than growth, intramuscular fat

and short loin fat weight were included separately as covariates in the base models for

each trait. All models contained all relevant first order interactions between the fixed

effects and covariates, including a quadratic effect of the covariate, and were removed

in a stepwise manner if non-significant (P > 0.05).

The impact of PWWT, PFAT and PEMD ASBVs on ICDH activity, myoglobin, iron

and zinc concentration and meat browning has previously been assessed (Calnan et al.,

2014; Kelman et al., 2014; Pannier et al., 2014b). This analysis aimed to determine

whether the impact of these ASBVs was delivered via their impact on growth. Thus

weight at 150 days or growth rate at 150 days were included separately in the base

model for each trait, along with the three ASBVs. All models contained relevant first

order interactions between the fixed effects and covariates, including a quadratic effect

of the covariate, and were removed in a stepwise manner if non-significant (P > 0.05).

Partial correlations of weight and growth rate with ICDH activity, myoglobin, iron and

zinc concentration and meat browning were determined using a multivariate analysis

(Version 9.1, SAS Institute, Cary, NC, USA). The analysis included the fixed effects

and their interactions described for the base models.

7.3 Results

7.3.1 Isocitrate Dehydrogenase

7.3.1.1 Base model

The base model for this study (not shown) utilised ICDH activity for 4,571 animals, and

has previously been published by Kelman et al. (2014) who analysed a subset of this

189

data (n = 3, 178). There were no significant variations to the Kelman et al. (2014) model

with the additional data, and the magnitude of the effects were similar.

7.3.1.2 Birth weight

Birth weight varied between 1.70 and 9.80 kg and across this range was associated with

a reduction in ICDH activity of 0.57 µmol/min/g tissue (P < 0.01) (Table 7-4).

190

97

Table 7-4 The magnitude of effect (difference between the highest and lowest values) of lamb weight and growth rate on isocitrate dehydrogenase activity, myoglobin concentration,

iron concentration, zinc concentration and meat browning.

ICDH Activity Myoglobin Concentration Iron Concentration Zinc Concentration Meat Browning

Birth Weight (kg) -0.57 -1.22 -0.25 - -0.14

Weight 100 days (kg) -0.38 2.22 0.26 0.13 -0.86

Weight 150 days (kg) -0.78 1.81 0.23 - -0.88

Growth Rate 100 days (g/day) -0.78 -0.91 -0.17 -0.35 -

Growth Rate 150 days (g/day) -0.60 -0.72 -0.22 -0.32 0.67

Weight Change 100 to 150 days (kg) -0.83 -1.47 -0.26 -0.35 0.60

191

7.3.1.3 Effect of weight at 100 and 150 days

Weight at 100 days was associated with a reduction in ICDH activity of 0.01

µmol/min/g tissue for each kg of weight, which was equivalent to 0.38 µmol/min/g

tissue across the 55 kg weight range. For weight at 150 days the association was twice

as strong with a reduction in ICDH activity of 0.78 µmol/min/g tissue across the 46 kg

weight range.

7.3.1.4 Effect of change in weight between 100 and 150 days and growth rate at 100

and 150 days

Growth rate measured at 100 and 150 days or the change in weight measured between

these time points were all associated with reducing ICDH activity (P < 0.01). Change in

weight between 100 and 150 days varied between -7.25 and 16.95 kg, growth rate at

100 days varied between -170.09 and 547.70 g/day, and growth rate at 150 days varied

between -170.23 and 393.49 g/day and across these ranges ICDH activity reduced by

0.83, 0.78 and 0.60 µmol/min/g tissue.

7.3.1.5 Effect of adiposity

When correcting the growth models for short loin fat weight, short loin fat weight did

not impact on ICDH activity, nor did it diminish the association between growth and

ICDH activity. When correcting the growth models for intramuscular fat (P < 0.01) the

magnitude of the growth covariates was not diminished for weight, and diminished by

less than a third for growth rate.

192

7.3.1.6 Effect of ASBVs

When the ASBV model was corrected for weight at 150 days or for growth rate at 150

days the magnitude of the breeding value effects reported by Kelman et al. (2014) did

not change.

7.3.2 Myoglobin Concentration

7.3.2.1 Base model

The base model for this study (not shown) utilised myoglobin concentration for 8,987

animals, and has previously been published by Kelman et al. (2014) who analysed a

subset of this data (n = 5,580). There was a minor variation to the Kelman et al. (2014)

model with the inclusion of a small effect for birth type-rear type combination.

7.3.2.2 Birth weight

Birth weight demonstrated an association with myoglobin concentration reducing it by

0.15 mg/g tissue for each kg of weight, which was equivalent to 1.22 mg/g tissue across

the 8.2 kg weight range (P < 0.01) (Table 7-4). This association varied between sites

and sire types. Between sites, myoglobin concentration reduced by as much as 0.28

mg/g tissue and as little as 0.11 mg/g tissue for each kg of weight. The variation with

sire type is detailed in Table 7-5.

.

193

97

Table 7-5 Variation in slopes between sire types for the impact of birth weight, weight at 100 and 150 days, weight change between 100 and 150 days and growth rate at 100 and 150

days on myoglobin concentration, iron concentration and zinc concentration.

Myoglobin Concentration Iron Concentration Zinc Concentration

Trait Sire type Slope Trait - Min. Trait - Max. Magnitude Slope Trait - Min. Trait - Max. Magnitude Slope Trait - Min. Trait - Max. Magnitude

Birth weight Maternal -0.12 1.90 9.70 -0.91 -0.03 1.90 9.50 -0.24 - - - -

Merino -0.26 2.20 9.10 -1.77 -0.05 2.28 8.60 -0.33 - - - -

Terminal -0.08 1.98 10.10 -0.61 -0.02 1.98 9.80 -0.18 - - - -

Weight 100 Maternal ND 0.006 6.87 51.06 0.28 ND

Merino ND 0.001 8.65 51.60 0.06 ND

Terminal ND 0.006 3.38 60.76 0.36 ND

Weight 150 Maternal 0.04 15.46 53.06 1.55 0.006 15.46 55.81 0.25 - - - -

Merino 0.01 13.93 51.20 0.35 -0.002 13.93 51.20 -0.07 - - - -

Terminal 0.04 11.33 68.78 2.54 0.008 11.33 68.78 0.44 - - - -

Weight Change Maternal -0.02 -4.18 15.27 -0.46 -0.007 -4.19 15.27 -0.13 ND

100 to 150 Merino -0.12 -7.25 15.88 -2.78 -0.022 -4.45 15.88 -0.45 ND

Terminal -0.02 -5.01 20.01 -0.45 -0.003 -5.01 20.12 -0.07 ND

Growth Rate 100 Maternal -0.15* -89.23 348.68 -0.07 -0.09* -89.23 348.68 -0.04 -0.47* -89.23 348.68 -0.21

Merino -3.79* -168.78 359.54 -2.00 -0.75* -168.78 359.54 -0.40 -0.81* -89.03 359.54 -0.37

Terminal 0.17* -170.09 547.70 0.12 0.14* -170.09 547.70 0.10 -0.19* -170.09 547.70 -0.13

Growth Rate 150 Maternal -2.27* -85.69 297.86 -0.87 ND ND

Merino -5.35* -170.23 277.49 -2.39 ND ND

Terminal -2.22* -107.37 393.49 -1.11 ND ND

Min: Minimum value for the variable; Max: Maximum value for the variable; Magnitude: The total magnitude of the effect the variable has on myoglobin concentration, iron

concentration and zinc concentration; *(slope) x10-3

; ND: No difference in slope between sire types; -: No significant association

194

7.3.2.3 Effect of weight at 100 and 150 days

Weight at 100 days demonstrated a curvilinear association with myoglobin

concentration increasing it by 0.04 mg/g tissue for each kg of weight, which was

equivalent to 2.22 mg/g tissue across the 57 kg weight range (P < 0.01). For weight at

150 days, the association increased myoglobin concentration by 1.81 mg/g tissue across

the 57.45 kg range (P < 0.01). This association varied between sites and sire types.

Between sites, myoglobin concentration increased by as much as 0.05 mg/g tissue and

as little as 0.05x10-2

mg/g tissue for each kg of weight. The variation with sire type is

detailed in Table 7-5.

7.3.2.4 Effect of change in weight between 100 and 150 days and growth rate at 100

and 150 days

Growth rate measured at 100 and 150 days or the change in weight measured between

these time points all demonstrated a relationship with myoglobin concentration (P <

0.01). Change in weight between 100 and 150 days varied between -7.25 and 20.01 kg,

growth rate at 100 days varied between -170.09 and 547.70 g/day, and growth rate at

150 days varied between -170.23 and 393.49 g/day and across these ranges myoglobin

concentration reduced by 1.47, 0.91 and 0.72 mg/g tissue. These associations varied

between sire types and are presented in Table 7-5.

7.3.2.5 Effect of adiposity

When correcting the growth models for short loin fat weight (P < 0.01) the magnitude

of the growth covariates diminished by a third to a half. When correcting the growth

models for intramuscular fat (P < 0.01) the magnitude of the growth covariates was

diminished by less than a tenth.

195

7.3.2.6 Effect of ASBVs

When the ASBV model was corrected for weight at 150 days or for growth rate at 150

days the magnitude of the breeding value effects reported by Kelman et al. (2014) did

not change, with the exception of PWWT which was no longer significant in either

model.

7.3.3 Iron Concentration

7.3.3.1 Base model

The base model for this study (not shown) utilised iron concentration for 8,434 animals,

and has previously been published by Pannier et al. (2014b) who analysed a subset of

this data (n = 5,625). There were minor variations to the Pannier et al. (2014b) model

with the inclusion of an effect for dam breed within sire type and the absence of a birth

type-rear type or birth type-rear type by sire type effect.

7.3.3.2 Birth weight

Birth weight demonstrated an association with iron concentration reducing it by 0.03

mg/100g for each kg of weight, which was equivalent to 0.25 mg/100g across the 7.9 kg

weight range (P < 0.01) (Table 7-4). This association varied between sites and sire

types. Between sites, iron concentration increased by as much as 0.06x10-2

mg/100g and

reduced by as much as 0.05 mg/100g for each kg of weight. The variation with sire type

is detailed in Table 7-5.

196

7.3.3.3 Effect of weight at 100 and 150 days

Weight at 100 days was associated with an increase in iron concentration of 0.45x10-2

mg/100g for each kg of weight, which was equivalent to 0.26 mg/100g across the 57 kg

weight range (P < 0.01). This association varied between sire types. Weight at 150 days

demonstrated an association with iron concentration, increasing it by 0.23 mg/100g

across the 57.45 kg weight range (P < 0.01). This association varied between sire types.

The variation with sire type at 100 and 150 days is detailed in Table 7-5.

7.3.3.4 Effect of change in weight between 100 and 150 days and growth rate at 100

and 150 days

Growth rate measured at 100 and 150 days or the change in weight measured between

these time points were all associated with reducing iron concentration (P < 0.01).

Change in weight between 100 and 150 days varied between -5.01 and 20.01 kg, growth

rate at 100 days varied between -170.09 and 547.70 g/day, and growth rate at 150 days

varied between -143.78 and 393.49 g/day and across these ranges iron concentration

reduced by 0.26, 0.17 and 0.22 mg/100g. The association between change in weight

between 100 and 150 days or growth rate and iron concentration varied at some sites

and between sire types. The variation at one site reduced iron concentration by 0.65x10-

2 mg/100g for each kg of weight change between 100 and 150 days. For growth rate at

100 days iron concentration reduced by as much as 0.06x10-2

mg/100g and as little as

0.01x10-2

mg/100g between sites for each g of growth rate. The variation with sire type

with change in weight between 100 and 150 days and growth rate at 100 days is detailed

in Table 7-5.

197

7.3.3.5 Effect of adiposity

When correcting the growth models for short loin fat weight (P < 0.05) the magnitude

of the growth covariates diminished by a third to a half. When correcting the growth

models for intramuscular fat (P < 0.01) the magnitude of the growth covariates did not

diminish.

7.3.3.6 Effect of ASBVs

When the ASBV model was corrected for weight at 150 days or growth rate at 150 days

the magnitude of the effect of the breeding values reported by Pannier et al. (2014b) did

not change, with the exception of PWWT which became significant with the inclusion

of weight at 150 days.

7.3.4 Zinc Concentration

7.3.4.1 Base model

The base model for this study (not shown) utilised zinc concentration for 7,236 animals,

and has previously been published by Pannier et al. (2014b) who analysed a subset of

this data (n = 5,625). There were no variations to the Pannier et al. (2014b) base model.

7.3.4.2 Birth weight

Birth weight demonstrated no association with zinc concentration.

7.3.4.3 Effect of weight at 100 and 150 days

Weight at 100 days was associated with an increase in zinc concentration of 0.23x10-2

mg/100g for each kg of weight, which was equivalent to 0.13 mg/100g across the 55 kg

198

weight range (P < 0.05) (Table 7-4). Weight at 150 days demonstrated no association

with zinc concentration.

7.3.4.4 Effect of change in weight between 100 and 150 days and growth rate at 100

and 150 days

Growth rate measured at 100 and 150 days or the change in weight measured between

these time points were all associated with reducing zinc concentration. Change in

weight between 100 and 150 days varied between -7.25 and 20.01 kg, growth rate at

100 days varied between -170.09 and 547.70 g/day, and growth rate at 150 days varied

between -170.23 and 393.49 g/day and across these ranges zinc concentration reduced

by 0.35, 0.35 and 0.32 mg/100g. The association between growth rate at 100 days and

zinc concentration varied between sire types and the magnitude of the variations are

detailed in Table 7-5.

7.3.4.5 Effect of adiposity

When correcting the growth models for short loin fat weight (P < 0.05) the magnitude

of the effect of growth rate at 100 days did not diminish, short loin fat weight was no

longer significant with weight change between 100 and 150 days or growth rate at 150

days and did not diminish the association between growth rate and zinc concentration,

and the association between weight at 100 days and zinc concentration was no longer

significant. When correcting the growth models for intramuscular fat (P < 0.01) the

magnitude of the growth covariates did not diminish except for weight at 100 days

which no longer had a significant association with zinc concentration.

7.3.4.6 Effect of ASBVs

When the ASBV model was corrected for weight at 150 days the magnitude of the

effect of the breeding values reported by Pannier et al. (2014b) did not change. When

199

the ASBV model was corrected for growth rate at 150 days PWWT became significant,

increasing zinc concentration by 0.16 mg/100g across the 23 kg PWWT range.

7.3.5 Meat Browning

7.3.5.1 Base model

The base model for this study (not shown) utilised meat colour for 5,033 animals, and

has previously been published by Calnan et al. (2014) who analysed a subset of this

data (n = 4,238). There were minor variations to the Calnan et al. (2014) model with the

inclusion of a site by year effect.

7.3.5.2 Birth weight

Birth weight varied between 1.70 and 10.10 kg and across this range was associated

with a reduction in meat browning of 0.14 units (P < 0.05) (Table 7-4).

7.3.5.3 Effect of weight at 100 and 150 days

Weight at 100 days demonstrated a curvilinear association with meat browning reducing

it by 0.02 units for each kg of weight, which was equivalent to 0.86 units across the 53

kg weight range (P < 0.05). Weight at 150 days was associated with a reduction in meat

browning of 0.02 units for each kg of weight, which was equivalent to 0.88 units across

the 55 kg weight range (P < 0.01).

7.3.5.4 Effect of change in weight between 100 and 150 days and growth rate at 100

and 150 days

Growth rate measured at 150 days and the change in weight between 100 and 150 days

were both associated with an increase in meat browning (P < 0.01), while there was no

200

association with growth rate at 100 days. Change in weight between 100 and 150 days

varied between -7.25 and 19.31 kg and growth rate at 150 days varied between -170.23

and 393.49 g/day and across these ranges meat browning increased by 0.60 and 0.67

units. The effect of growth rate at 150 days varied between sites and years. Between

sites, meat browning increased by as much as 0.23x10-2

and as little as 0.10x10-2

units

for each g of growth rate at 150 days. Between years, meat browning increased by as

much as 0.24x10-2

and as little as 0.11x10-2

units for each g of growth rate at 150 days.

7.3.5.5 Effect of adiposity

When correcting the growth models for short loin fat weight (P < 0.05) the

magnitude of the growth covariates diminished by around a third. When correcting the

growth models for intramuscular fat (P < 0.01) the magnitude of the growth covariates

did not diminish.

7.3.5.6 Effect of ASBVs

When the ASBV model was corrected for weight at 150 days or growth rate at 150 days

the magnitude of the breeding value effects reported by (Calnan et al., 2014) did not

change.

7.3.6 Phenotypic correlations

The phenotypic correlations of the meat traits with weight and growth rate are presented

in Table 7-6. For most traits there is a negative correlation with birth weight followed

by a positive correlation with weight at 100 days. The phenotypic correlations generally

aligned well with the phenotypic associations.

.

201

97

Table 7-6 Phenotypic correlations of isocitrate dehydrogenase (ICDH) activity, myoglobin concentration, iron concentration, zinc concentration and meat browning with lamb weight

and growth rate. Standard error values in brackets. Significant correlations are in bold.

Birth weight Weight 100 days Weight 150 days Growth rate 100 days Growth rate 150 days

ICDH Activity -0.05 (0.00) 0.00 (0.97) -0.04 (0.01) -0.05 (0.00) -0.05 (0.00)

Myoglobin Concentration -0.08 (0.00) 0.11 (0.00) 0.11 (0.00) 0.02 (0.09) -0.07 (0.00)

Iron Concentration -0.07 (0.00) 0.06 (0.00) 0.07 (0.00) 0.03 (0.01) -0.04 (0.00)

Zinc Concentration -0.03 (0.03) 0.03 (0.04) 0.00 (0.98) -0.04 (0.00) -0.05 (0.00)

Meat Browning -0.02 (0.12) -0.12 (0.00) -0.10 (0.00) 0.03 (0.04) 0.09 (0.00)

202

7.4 Discussion

7.4.1 Isocitrate dehydrogenase activity

Contrary to our hypothesis, the oxidative indicator enzyme ICDH showed a negative

association with lamb growth. This association was consistent with both weight and

growth rate measures including weight at birth, 100 and 150 days, weight change

between 100 to 150 days, and instantaneous growth rate at 100 and 150 days. The

negative association between lamb growth and ICDH activity is supported by the

negative phenotypic correlations between ICDH activity and both weight and growth

rate (Table 7-6).

The strength of the association between weight and ICDH activity increased between

100 and 150 days as indicated by the magnitude of change in ICDH activity across the

range of weights, with this magnitude doubling from 0.39 µmol/min/g tissue at 100

days to 0.78 µmol/min/g tissue at 150 days (Table 7-4). Supporting this change was the

strong association between growth rate at 100 days and ICDH activity (i.e. 0.78) which

would have gradually increased the strength of the association between ICDH activity

and live weight. Therefore the window for growth rate to elicit the greatest change to

oxidative capacity lies in the period between 100 and 150 days from the weaning to post

weaning phase.

ICDH activity has a strong positive association with intramuscular fat (Kelman et al.,

2014) which is similar in magnitude to the association between ICDH activity and

weight at 150 days. Given that intramuscular fat increases with live weight (Chapter 6)

the change in ICDH activity with growth may be associated with intramuscular fat or

whole carcass adiposity. However, the correction of the growth models for

intramuscular fat or short loin fat weight (indicating whole carcass fatness) had little

203

impact on the association between growth and ICDH, suggesting that the effect of

growth on ICDH activity is independent of intramuscular fat. Bernard et al. (2009)

showed a similar disassociation between muscle metabolism with growth and fat

deposition in cattle. When short loin fat weight was included in the growth models it

was not significant, and therefore did not diminish the association between growth and

ICDH activity. This is not surprising given that short loin fat weight alone had no

impact on ICDH activity (Kelman et al., 2014). Thus, it is can be concluded that growth

drives reduced oxidative capacity and this is not simply a reflection of associated

changes in fatness.

Although the hypothesis for an increase in ICDH activity with lamb growth was based

on a positive association between oxidative capacity and hot carcass weight there is

evidence suggesting the reverse can occur. Selection for increased growth rate has been

shown to shift muscle metabolism toward a more glycolytic and less oxidative type

(Rahelic and Puac, 1981). This shift may occur as animals selected for growth tend

toward a larger mature size (Huisman and Brown, 2008) having heavier carcass weights

(Sharaby and Suleiman, 1988) but being proportionately less of their mature size at a

given age. Thus they have more glycolytic fibres and lower levels of ICDH activity.

Likewise, the impact of selection for leanness in lambs has also been shown to increase

weight at 150 days (Chapter 4) and reduce ICDH activity (Gardner et al., 2007; Gardner

et al., 2006b). Hence there is some basis for our results in the literature.

What is less clear is why there was no phenotypic association (Kelman et al., 2014) or

phenotypic correlation (Mortimer et al., 2014) between hot carcass weight at slaughter

and ICDH activity. While Kelman et al. (2014) indicated increased oxidative capacity

with heavier hot carcass weights, this was largely based on myoglobin concentration not

ICDH activity. ICDH activity varied with hot carcass weight at two of the eight sites in

204

the study, increasing at one site and reducing at the other so the effect was quite

inconsistent. With weight change increasingly dominated by fat deposition as animals

approach slaughter, we speculate that there may a gradual uncoupling of the association

between early life growth and muscle oxidative type, resulting in no hot carcass weight

association at slaughter. Given the lack of consistent phenotypic association with ICDH

activity, hot carcass weight alone should not be used as a proxy for growth at all stages

of development when assessing its impact on muscle metabolic type.

7.4.2 Myoglobin concentration

The association between lamb growth and myoglobin concentration varied throughout

the different phases of lamb growth. Myoglobin concentration had a negative

association with weight at birth, suggesting an impact of in-utero growth rate, after

which the association inverted and myoglobin concentration had a strong positive

association with weight at 100 days (Table 7-4). These results align well with the

phenotypic correlations between growth indicators and myoglobin concentration during

this period (Table 7-6).

While the association with weight from weaning onwards aligned with our hypothesis,

and has been demonstrated previously (Beriain et al., 2000), the inversion between birth

and weaning was unexpected. It did however highlight that the window for the greatest

change to myoglobin concentration was between birth and weaning, with the magnitude

of the increase in myoglobin concentration during this period being as large as 3.52

mg/g tissue.

The association between myoglobin concentration and growth rate was negative and

strengthened between 100 and 150 days. This may explain the diminishing association

between myoglobin concentration and live weight which decreased by about 20% at

205

150 days compared to 100 days. Nonetheless, the magnitude of the association between

myoglobin concentration and slaughter weight, as indicated by the association with hot

carcass weight in the Kelman et al. (2014) study, was a similar magnitude to that seen

for weight at 150 days. This suggests that the negative association between myoglobin

concentration and growth rate following weaning was not enough to uncouple the

overall positive association with weight. During this same time period there is a shift in

muscle fibre type with increasing hot carcass weight from a glycolytic to an oxidative

metabolism (Greenwood et al., 2007). As myoglobin concentration is increased with

oxidative capacity the association between myoglobin concentration and weight may

stabilise post weaning due to a shift in muscle fibre type.

As with ICDH activity, myoglobin concentration has a positive association with

intramuscular fat (Kelman et al., 2014). To determine whether the change in myoglobin

concentration with growth was simply a reflection of its association with adipose tissue

the growth models were corrected separately for intramuscular fat and short loin fat

weight. The inclusion of intramuscular fat had little impact on the magnitude of the

association between myoglobin concentration and growth while the inclusion of short

loin fat weight reduced the association by a third to a half. These results suggest that the

impact of growth on myoglobin concentration is independent of intramuscular fat

although not completely independent from whole carcass adiposity. Nonetheless growth

still describes a significant amount beyond just that which is correlated with whole

carcass adiposity.

Myoglobin concentration is an indirect indicator of oxidative muscle type however the

association with growth does not perfectly align with the results of the more direct

indicator of oxidative metabolism in ICDH activity. This was also true in the Kelman et

al. (2014) study where hot carcass weight impacted myoglobin concentration however

206

had no overall impact on ICDH activity, so the association between hot carcass weight

and oxidative capacity was inconsistent. The lack of alignment between the two

indicators is also evident in the window of growth during which both can be influenced,

being during early pre weaning growth for myoglobin concentration and post weaning

for ICDH activity.

7.4.3 Iron and zinc concentration

In general, the association between growth and iron and zinc concentration largely

reflected the results of myoglobin concentration. This was particularly true for iron

concentration, which had a negative association with weight at birth followed by a

strong positive association with weight at 100 days that diminished slightly by 150 days

in response to an strengthening negative association with growth rate between 100 and

at 150 days (Table 7-4). Again, these associations align well with the phenotypic

correlations for iron which were negative with birth weight and positive with weight at

100 and 150 days (Table 7-6). The phenotypic correlations between zinc concentration

and growth were generally weaker, reflecting a lack of association with birth weight and

weight at 150 days, a weak positive association with weight at 100 days and negative

associations with growth rate at 100 and 150 days. Based on the phenotypic associations

and correlations the window for greatest change to iron concentration is between birth

and weaning, as was the case for the myoglobin concentration, with the magnitude of

the increase in iron concentration during this period being as large as 0.51 mg/100g.

The alignment in responses of myoglobin and iron concentrations to growth is not

unexpected given myoglobin is an iron rich protein resulting in a high phenotypic

correlation of 0.97 (Mortimer et al., 2014) between these two measurements. In Chapter

6, the response of intramuscular fat to growth paralleled the responses of myoglobin and

iron concentrations, including a window for greatest change during early life pre

207

weaning growth. This is likely to be due to the close association of these traits, each

being linked to muscle fibre type and oxidative capacity (Hocquette et al., 2003; Jurie et

al., 2007; Kelman et al., 2014). Given lambs are reliant on ewes for nutrition prior to

weaning, and that the greatest window for change for these traits is pre weaning, there

may be scope to influence intramuscular fat, myoglobin concentration and iron

concentration in lambs by manipulating feed to ewes.

Iron and zinc concentration are linked to oxidative capacity (Pannier et al., 2014b) and

are therefore linked to oxidative fibre types (Gardner et al., 2007) so an increase in

mineral content with increased weight was expected, however iron concentration was

found to be more sensitive to changes in weight than zinc concentration. This aligns

with the findings of Pannier et al. (2014b) of a larger increase in iron concentration with

increasing hot carcass weight. In the current study, the total magnitude of change in zinc

concentration due to weight at 100 days was equivalent to approximately 5% of the

average zinc concentration, indicating a relatively small impact of weight on zinc

concentration.

Despite a negative phenotypic correlation for iron concentration and no significant

correlation for zinc concentration (Mortimer et al., 2014) the impact of hot carcass

weight on both iron and zinc concentration was positive, as reported by Pannier et al.

(2014b). The magnitude of the association between iron concentration and slaughter

weight, as indicated by the association with hot carcass weight in Pannier et al. (2014b)

was a similar magnitude to that seen for weight at 150 days so the negative association

between iron concentration and growth rate was not enough to uncouple the positive

association with weight. Similarly for zinc concentration, where the association with hot

carcass weight was about 50% less (Pannier et al., 2014b) than with weight at 100 days,

208

the negative association between zinc concentration and growth rate did not completely

uncouple the positive association with weight.

Iron and zinc concentrations increase with intramuscular fat (Pannier et al., 2014b) and

intramuscular fat increases with weight (Chapter 6) so the growth models were

corrected separately for intramuscular fat and short loin weight to determine if the

impact of growth was a reflection of increased adiposity. The results largely reflected

those seen for myoglobin concentration, in that the inclusion of intramuscular fat

generally had no impact on the magnitude of the association between growth and iron or

zinc concentration while the inclusion of short loin fat weight reduced the associations

by around a third to a half. The only exception was weight at 100 days, where both

measures completely accounted for the association with zinc concentration. These

results suggest that the impact of growth on iron and zinc concentration describes a

significant amount beyond just that which is correlated with whole carcass adiposity.

The impact of growth on myoglobin concentration, iron concentration and zinc

concentration varied between sire types (Table 7-5). The effect on myoglobin

concentration was strongest in Merinos at most times measured and did not change with

correction of the model for short loin fat weight. This indicates a higher level of

myoglobin in progeny of Merino sires at maturity, independent of whole body

adiposity. Similarly, correction of the iron and zinc models for whole carcass adiposity

did not change the variation of the impact of growth between sire types.

7.4.4 Meat browning

The reduction in meat browning with increasing weight was contrary to expectation but

aligns well with the association between ICDH activity and weight. As oxidative

metabolism is a key driver of meat browning (Calnan et al., 2014) it makes sense that a

209

reduction in oxidative metabolism, as indicated by ICDH activity, with increasing

weight will result in a reduction in meat browning. Furthermore, these results align well

with the negative phenotypic correlations between meat browning and weight (Table

7-6). The lack of association between meat browning and growth rate at 100 days

explains why there is no change in the magnitude of the association between meat

browning and weight at 100 and 150 days (Table 7-4). The difference in association

between meat browning and weight is greatest between birth and weaning indicating the

period where the greatest change can be made.

Meat browning is reduced with increased hot carcass weight at slaughter (Calnan et al.,

2014). The magnitude of the association is about two thirds the magnitude of the

association with weight at 100 or 150 days in this study. This reduction in magnitude is

due to the positive association between meat browning and growth rate starting to

uncouple the negative association with weight.

Meat browning has a positive association (Calnan et al., 2014) and phenotypic

correlation (Mortimer et al., 2014) with intramuscular fat. Therefore, increases in

intramuscular fat with increased live weight (Chapter 6) may increase browning. When

the growth models were corrected for intramuscular fat the association with meat

browning did not diminish. This suggests that growth impacts meat browning

independently of associated changes in intramuscular fat, which aligns with the findings

for ICDH activity. Alternatively, correction of the growth models for short loin fat

weight reduced the association with meat browning by about one third. This indicates

that the impact of growth on meat browning is not completely independent of whole

carcass adiposity, although growth still describes a significant amount beyond its

correlation with whole carcass adiposity.

210

The published impact of PWWT, PFAT and PEMD on ICDH activity and myoglobin

concentration (Kelman et al., 2014), iron and zinc concentration (Pannier et al., 2014b)

and meat browning (Calnan et al., 2014) did not change when corrected for weight at

150 days or growth rate at 150 days. The effects of PFAT and PEMD did not change

with correction for weight or growth rate at 150 days indicating that they are not

mechanistically driven through the impact of growth, as evidenced by the lack of a

PWWT effect on mineral content and meat browning.

211

7.5 Conclusion

This study demonstrated that lamb growth impacts on muscle oxidative capacity,

mineral content and meat browning. The key period for modulation of myoglobin and

iron concentration was between birth and weaning with the magnitude of the responses

reducing after weaning. As iron is a key component of myoglobin the alignment of the

response was as expected. Birth to weaning was also the key period for modulation of

meat browning, although the magnitude of the response does not diminish after

weaning. They key period for modulation of the oxidative capacity indicator enzyme

isocitrate dehydrogenase is later than for other traits reported, being between weaning

and post weaning as the strength of the association between isocitrate dehydrogenase

and live weight increases. Zinc was less sensitive to changes in growth although the

responses which were present aligned with iron concentration. As for intramuscular fat

in Section 6.3.1.5, the response of myoglobin concentration, iron concentrations and

meat browning to lamb weight inverted between birth and weaning indicating they may

also be influenced by in-utero growth. In general, the impact of growth on oxidative

capacity, mineral content and meat browning were not due to local adiposity while up to

half the magnitude of the associations were due to a correlated increase in whole carcass

adiposity. However, there was still a significant variation that was not associated with

whole body fatness. Hot carcass weight, as a proxy for whole of life growth, provided a

poor estimate of the impact of growth frequently underestimating the impact of growth

on oxidative capacity, mineral content and meat browning. Hot carcass weight varies

with age at slaughter, while growth at weaning and post weaning do not, contributing to

the poor alignment of hot carcass weight and growth effects. Furthermore, hot carcass

weight was unable to describe the impact of growth during different phases of growth.

212

213

Chapter 8. General Discussion

8.1 Benefits of lamb growth to industry

The growth of lambs from birth to slaughter is of vital importance to producers in the

Australian sheep industry, while meat quality is of increasing interest to consumers

(Pethick et al., 2006a; Pethick et al., 2006b). Currently, producers use Australian Sheep

Breeding Values (ASBVs) to select for increased growth at birth, weaning and post

weaning time points (Hall et al., 2002). However, the impact of growth at these time

points on meat quality has not been well characterised.

Lamb growth can be tracked throughout life by measuring live weight at various

intervals as they grow. However, lamb weight at individual time points varies greatly

due to gut fill, time of day, length of curfew and other production factors. Furthermore,

when dealing with large populations, not all animals are weighed at the same time.

Hence there is a need to assign an accurate growth function through the available live

weight data so that growth at equivalent time points of interest can be estimated.

Three growth functions were fitted to the available live weight data collected on 18,185

lambs to determine the function with the best fitting characteristics (Chapter 3). These

functions were chosen to represent a difference in curve shape, linear or nonlinear, and a

difference in fitting type, individual or population based.

Subsequent to growth being estimated at weaning and post weaning time points

(Chapter 4), the impact of growth at these times on meat quality was examined

(Chapters 6 and 7). The indicators of meat quality examined were chosen due to their

importance in consumer acceptability of meat and their links to muscle fibre type.

214

Muscle fibre type is of particular interest because it is linked to important traits of

consumer acceptability such as fresh and retail meat colour, mineral content, ultimate

pH and rate of pH decline, intramuscular fat and ageing rate. Therefore, growth may

impact meat traits, and thus consumer acceptability, through its impact on fibre type.

Oxidative fibre type can be indicated by activity of the oxidative enzyme isocitrate

dehydrogenase (ICDH) and by myoglobin concentration so the impact of growth on

these markers was examined. The impact of growth on traits linked to eating quality,

retail shelf life and nutritional qualities were examined. These included:

Indicators of eating quality: intramuscular fat, shear force and consumer overall liking;

and

Indicators of meat nutrition: iron and zinc concentration; and

Indicators of retail shelf life: rate of meat browning.

8.2 Brody function

The Brody function is a nonlinear function which is fitted individually to each animal’s

weight records. It is commonly used due to its ease of estimation, with an ease of

computation and a limited number of curves which do not converge (Kaps et al., 1999;

Topal et al., 2004). Furthermore, the parameters have biological significance and are

simple to interpret, representing mature body weight and maximum growth rate

(Fitzhugh, 1976). However, the Brody function had the poorest fitting characteristics of

all three curves examined including the highest mean squared error indicating the

greatest deviation of the estimated weights from the recorded weights. A contributing

factor to this is the overestimation of weights during early life growth, a recognised

215

limitation of the Brody function (Beltrán et al., 1992). To rectify this, a two function

approach could be used, with a separate function fitted during early life growth and a

Brody function fitted after the point of inflection of the growth curve. However, this

unnecessarily increases the complexity of the curve fitting process and would also

require the collection of more data points during this early growth period pre weaning.

This was not done with this study due to the risk of mis-mothering of lambs, a practical

challenge likely to limit the application of this approach.

The scope to include further nonlinear growth functions in the comparison is virtually

unlimited. The Brody function was used due to its ease of computation, however other

nonlinear functions such as the Von Bertalanffy, Richards, Logistic and Gompertz

functions have each been used to model lamb growth and may have been acceptable

substitutes.

8.3 Cubic polynomial function

The cubic polynomial is a linear function fitting weight as a function of time only. As

for the Brody function, it is fitted on an individual basis. Polynomial functions are

attractive when modelling weight change due to the flexible nature of the curves

increasing as the degree of polynomial increases from linear to quadratic to cubic to

quartic and so on. While this appears advantageous there are limitations.

It is difficult to know when to stop increasing the polynomial since with every increase

in polynomial the curve will conform to the live weight data with greater accuracy. This

results in smaller residual weights and a lower mean square error than for other curves,

as seen in the first experiment. However this could result in the modelling of small

216

changes in weight including gut fill on a given day, which may not necessarily increase

the understanding of how weight changes with time on a macroscopic level. The shape

of the curve would fluctuate increasingly with the degree of polynomial and the

trajectory, from which instantaneous growth rates are calculated, would be influenced

by factors such as gut fill rather than whole of life trajectory.

Increasing the degree of polynomial also increases by one the number of weight

recordings needed for each animal. Thus, for a cubic polynomial to be fitted, an animal

needs to have at least four data points. Therefore as the degree of polynomial increases

the number of animals with sufficient weight recordings reduces, potentially limiting

the application of this technique. In comparison, a Brody function can be fitted to

animals with as little as three weight recordings.

Polynomial functions are frequently asymptotic. This can result in extreme estimates for

birth and adult weights. Brody curves may also have asymptotes however they are

closer to horizontal so the estimated adult weight is relatively stable.

Lastly, unlike the parameters of a Brody function, the parameters of polynomial

functions do not have a direct biological interpretation. Therefore it is difficult to

compare the results across studies or understand how they influence the fit of the

growth curve.

8.4 Random effects model

A random effects model function fits a growth curve to the entire population of animals,

with the growth curve of individual animals varying directly from the population

growth curve. As it is based on a polynomial function it has similar limitations to fitting

217

a cubic polynomial. However this method is not as subject to bias as an individual

Brody or cubic polynomial fit. This is because random effects models can accommodate

different durations of data recording. Where weight data is missing in some animals, it

is informed by other animals with similar production or genetic information. Based on

these properties, a random effects model can be fitted for an animal with as little as one

live weight recording.

The random effects model was hypothesised to have better curve fitting characteristics

than an individual animal function fitted on an individual animal basis. While this did

prove to be the case when compared to the Brody function, the fit was actually worse

than for the individually fitted cubic polynomial function. The specific reason for this is

unknown however the inclusion of animals with less than three or four weight

recordings, which was not possible with a Brody or cubic polynomial function, did not

significantly compromise the fit, increasing the MSE by around only 5%.

An alternative population based fit not examined in this thesis is the use of spline

functions, which do not require the growth function to follow a strict cubic polynomial

form (McPhee et al., 2012). Rather, spline functions can force the growth curve through

actual weight recordings. While this reduces the variance it also increases bias as the

value of the smoothing parameter lambda increases to facilitate the curvature. Due to

the large number of growth functions that exist, it was not possible to utilise each

method. The limitations of potential functions were considered and splines were not

used due to the bias they introduce. However, splines are not uncommonly used when

modelling animal growth data so the method cannot be entirely discounted.

From a pragmatic standpoint, each of the growth functions assessed gave a reasonable

estimation of growth. Ultimately the random effects model offers a more powerful and

flexible means of evaluating repeated live weight measures, particularly with respect to

218

accommodating animals with few live weight recordings, which is common with

industry data. Random effects models are also subject to less bias than the Brody

function, which is influenced when estimating weights near the edge of the available

data. Therefore, the random effects model was used to estimate growth at weaning and

post weaning time points, which were used to determine the impact of growth on meat

quality.

Based on the random effects model, weights and growth rates were estimated at key

points in production. These were 100 days, 150 days and 240 days which represent the

ages at weaning, post weaning and the average age at which post weaning ASBVs were

calculated.

8.5 Impact of environment on growth

Growth is a function of environment. Environmental and production factors within this

study included site of birth, year of birth, age, sex, birth type-rear type combination,

dam age, dam breed and sire type (Chapter 4).

Numerous studies have reported the impact of environmental factors on lamb growth

hence the impacts on growth were generally as expected (Afolayan et al., 2007;

Thatcher et al., 1991). However, this analysis represents a benchmarking study allowing

comparison of magnitudes of effect for each of these factors based on large numbers of

animals.

The site at which lambs were reared had the largest effect on growth while the effect

varied very little between years. The differences in growth between sites was partly due

to differences in nutrition. However nutrition was confounded by environmental factors

219

such as rainfall, pasture types and fertiliser application, and no single factor explains all

the variation in growth between sites. For example, annual rainfall at the Kirby site is

50% greater than at the Trangie site yet lamb growth at Kirby was consistently low and

at Trangie was consistently high.

The effect of birth type on lamb growth was almost as large in magnitude as the impact

of environmental factors at the site of rearing with multiple birth lambs having lower

weights and growth rates. However, from a practical standpoint it is difficult to select

against multiple births to ensure maximum lamb growth. Furthermore, the number of

lambs weaned and is also a profit driver in lamb production which may be an acceptable

trade-off for reduced growth.

Production effects which producers can target include dam breed, dam age and sire type

with lambs from dams less than four years of age and Terminal sires with Border

Leicester x Merino dams increasing lamb growth. Of these, the magnitude of the sire

type effect is greatest, and it represents as much as 80% of the magnitude of the effect

of site.

Production systems which are targeting accelerated lamb growth will increase success

by ensuring adequate nutrition, in male lambs born to Terminal sires and Border

Leicester x Merino which are less than 4 years of age.

Nutrition is a key driver of lamb growth however there are production factors which can

impact nutrition that were not directly captured by this experiment. For example, milk

volume is reduced in Merino dams leading to reduced growth in lambs compared to

Border Leicester-Merino dams. However the impact of colostrum volume and

composition is unknown. Colostrum volume and composition could be investigated,

220

particularly from single compared to multiple pregnancy ewes, to determine if

colostrum quality was a source of nutritional restriction in addition to milk volume.

While most of the results of this study are not new, the key outcome of this analysis was

to provide a good basis for the comparative magnitude of these effects, based on large

numbers of animals.

8.6 Impact of genetics on growth

The impact of selection for growth using ASBVs was quantified (Chapter 4). The

magnitude of the effect increased with time, being larger at 240 days than at birth.

However, as a proportion of the average live weight the effect diminished with time,

being strongest at birth. There was no effect of the breeding values on instantaneous

growth rate measured at 100, 150 and 240 days. The magnitude of the ASBV effects

compared to production effects also diminished with time. For example, at birth the site

of rearing (i.e. environmental effect) and the ASBV for birth weight had a similar

magnitude of effect however by 240 days the ASBV effect was half the effect of site.

While producers cannot control the environmental/site effects they are able to control

ASBVs and genotype. Therefore, the magnitude of these effects is a more fair

comparison. For example, the magnitude of the effect of the ASBV for birth weight is

twice that for dam breed or sire type, so if birth weight is critical, ASBVs will have a

larger effect. This effect is reversed at weaning and post weaning with the magnitude of

difference between sire types close to double the magnitude of the growth ASBV effect.

Therefore producers interested in higher growth at weaning and post weaning will have

more of an impact by choosing a Terminal sire, although use of a Terminal sire in

221

combination with growth ASBVs will have the greatest impact on lamb growth. That is

not to say that progeny of Terminal sires will always be heaviest as progeny from high

PWWT Maternal sires are on average heavier than progeny from low PWWT Terminal

sires. Thus expectations of weight must be tempered with knowledge of the interaction

between production factors, such as sire type, and potential weights based on PWWT.

The expression of growth breeding values, particularly sire PWWT, is limited by poor

nutrition (Hegarty et al., 2006b). Following on from these findings, our study

investigated different forms of nutritional restriction in dams or lambs due to siblings,

dam breed, dam age etc. on the expression of growth breeding values. Nutritional

restriction was found to moderate the full expression of growth breeding values.

The cause of reduced expression of growth breeding values is important for producers

to understand. It identified breeding objectives which may be in conflict. For example,

producers selecting for multiple birth and increased lamb growth can now expect less

genetic growth in multiples compared to singles.

These findings are extremely important from an industry perspective. Firstly, they

highlight nutrition as an important and powerful production tool for the manipulation of

lamb growth. Secondly, these results highlight that production factors which limit the

available nutrition, such as birth type, will depress the full expression of growth

breeding values. Quantification of the magnitude of these effects are provided at birth,

weaning and post weaning.

222

8.7 Fibre type and oxidative capacity

Muscle oxidative capacity was a specific focus of this thesis as it reflects fibre type

which is of central importance as it influences a number of meat traits of commercial

importance, including meat eating quality, mineral content, and meat browning (Calnan

et al., 2014; Pannier et al., 2014b; Pannier et al., 2014c). Therefore, factors that impact

on muscle oxidative capacity may lead to correlated changes in these meat traits.

Furthermore, a number of these factors are likely to be a reflection of the effect of

growth rate on muscle oxidative capacity and fibre type. Thus characterising them in

detail and then assessing their magnitude, after correcting for growth, enables us to

asses which are simply a reflection of growth.

Muscle oxidative capacity was indicated by isocitrate dehydrogenase activity and

myoglobin concentration in this thesis (Chapter 5). Due to the large number of animals

sampled in this experiment, and the costs and time associated with measurement, a

pragmatic decision was made to include only two indicators of muscle oxidative

capacity. This approach to assessing oxidative capacity has been taken previously in

lamb. There are a number of other indicator enzymes that could have been measured to

reflect oxidative (citrate synthase) or glycolytic (lactate dehydrogenase) capacity, as

well as other histochemical and immunohistochemical fibre typing techniques.

However, isocitrate dehydrogenase was chosen as it is a relatively consistent indicator

of muscle fibre type (Gardner et al., 2006b). Myoglobin concentration was also used,

not only for its links to aerobicity, but due to its phenotypic and genetic correlation with

other important meat traits investigated in this thesis. These include a high correlation

with iron concentration and meat browning (Mortimer et al., 2014).

Despite the links of ICDH activity and myoglobin concentration with muscle fibre type

and aerobicity their response to environmental and growth factors was not always

223

consistent. ICDH activity, which is linked to metabolic efficiency, was demonstrated to

be more responsive to the environment than myoglobin concentration with a greater

variation across sites and years. Conversely myoglobin concentration was less

responsive to environmental factors and heavily driven by age.

8.8 Impact of genetics on oxidative capacity

Oxidative capacity increased with selection for increased PWWT ASBV. Although this

was unexpected the shift in fibre type with growth could impact meat quality traits

linked to fibre type. The impact of PWWT on isocitrate dehydrogenase activity negated

through its association with whole body adiposity so the main impact of PWWT was on

myoglobin concentration. Therefore, the traits linked to myoglobin concentration,

namely zinc concentration and meat browning were likely to have similar responses to

growth. This was particularly evident for iron concentration, which is intrinsically

linked to myoglobin concentration (Hazell, 1982).

The environment in which the lamb was raised had the largest influence on both

indicators of oxidative capacity, being triple the magnitude of the effect of PWWT. The

impact of growth ASBVs on oxidative capacity was only assessed at the post weaning

time point. As myoglobin concentration changes significantly between birth and

weaning it would be interesting to see if growth ASBVs at birth and weaning were

impacted by weight at these times as they could be used by producers to manipulate

oxidative capacity and meat colour.

While there are currently no market price signals for eating quality that flow back to

producers this is likely to change in the future. The impact of selection for growth on

224

meat quality can be predicted and then tracked using gene markers and altered using

quantitative breeding values if necessary.

8.9 Impact of growth on oxidative capacity

Myoglobin concentration changed markedly with weight between birth and weaning

however the change in the negative association with weight at birth to a positive

association with weight at weaning is not well characterised as few weights were

recorded during this period (Chapter 7).

As fibre type is linked to meat quality traits it is relevant to know the impact that

selection for leanness and growth will have, and to know whether phenotypic growth

will have an impact in addition to genetic selection for growth. The impact of selection

using PEMD and PFAT ASBVs on oxidative capacity was generally unchanged with

correction for phenotypic growth, indicating that these breeding values impact oxidative

capacity through mechanisms not correlated with growth rate. This is not surprising,

particularly given the growth breeding value itself (i.e. PWWT) also had minimal

impact on oxidative capacity.

While intramuscular fat was linked to an increase in oxidative capacity (Chapter 5) it

was shown to be independent of its association with PFAT or whole body adiposity or

muscularity (Chapter 5). An increase in growth reduced isocitrate dehydrogenase

activity however this was independent of a correlated increase in adiposity with growth

(Chapter 6 and Chapter 7). Thus growth impacts oxidative capacity independently of

adiposity.

225

8.10 Impact of growth on objective and subjective measures of

eating quality

Intramuscular fat and shear force are important objective measures of eating quality

which influence the subjective measure of eating quality, consumer overall liking

(Chapter 6). Intramuscular fat in particular influences the organoleptic properties of

meat (Hocquette et al., 2010) and is clearly visible to consumers.

As for myoglobin concentration, intramuscular fat changed markedly with weight

between birth and weaning. The change was less dramatic for shear force, with no

impact of birth weight followed by a negative association at weaning. For both traits the

association had weakened by the post weaning time point (150 days) due to an

increasing positive association with growth rate. The magnitude of the effect at weaning

represented as much as a quarter the average total amount of intramuscular fat present at

slaughter, while for shear force it was two fifths, indicating the size of the impact

delivered through increased growth.

The magnitude of the effect of growth on intramuscular fat or shear force was in part

delivered through an associated increase in carcass adiposity. Within this study

adiposity accounted for about a third to a half of the magnitude of the growth effect.

Alternatively this still means that growth independent of its associated increase in

adiposity still accounted for the remaining half of the effect.

While there is currently no price premium for objective measures of eating quality in

lambs, this may change in the future. This could occur with the development of an

ASBV for intramuscular fat, which will allow intramuscular fat and carcass leanness to

be controlled independently thus uncoupling the biological association. While it is not

currently detectable, the impact of phenotypic growth could eventually impact overall

226

liking through its associations with intramuscular fat. Therefore it is relevant to know

whether phenotypic growth will have an impact on these traits in addition to genetic

selection for growth.

The impact of selection using PEMD and PFAT ASBVs on intramuscular fat, shear

force and overall liking was generally unchanged when corrected for phenotypic

growth, indicating that these breeding values impact these traits through mechanisms

not correlated with growth rate. This is not surprising, particularly given the growth

breeding value itself (i.e. PWWT) also had no impact on intramuscular fat, shear force

and overall liking.

A key hypothesis was based on hot carcass weight, which is used as a proxy for whole

of life growth rate (Pannier et al., 2014b; Pannier et al., 2014c) based on its high

correlation with pre slaughter live weight. This study demonstrated that the effect of hot

carcass weight on traits is a poor proxy for the effect of growth, overestimating the

effect for intramuscular fat and shear force. Obviously this measure cannot describe the

impact of growth during different phases of growth therefore the window for

intramuscular fat and shear force manipulation during pre weaning growth had not

previously been demonstrated.

8.11 Impact of growth on mineral content of meat

Iron and zinc concentration are important consumer health traits. As for myoglobin

concentration, intramuscular fat and shear force, iron concentration changed markedly

with weight between birth and weaning (Chapter 7). This is not surprising given that

iron is a key component of myoglobin (Hazell, 1982). The change was less dramatic for

227

zinc concentration which was less sensitive to lamb weight, as indicated by a reduced

association with hot carcass weight compared to zinc concentration. Arguably however,

the profile of iron in human health is more prominent than zinc so the stronger

association of lamb weight is of more commercial significance.

As for measures of oxidative capacity as well as subjective and objective measures of

meat quality, the magnitude of the effect of growth on iron and zinc concentration was

in part delivered through an associated increase in carcass adiposity. Within this study

adiposity accounted for about a third to a half of the magnitude of the growth effect on

iron and zinc concentration. This suggests that growth, independent of the associated

increase in adiposity, accounts for at least half the effect.

As for other traits, it is relevant to know what portion of the impact of PFAT and PEMD

ASBVs were delivered through their correlated effects on growth. The impact of

selection using PEMD and PFAT ASBVs on mineral content was unchanged with

correction for phenotypic growth, indicating that these breeding values impact iron and

zinc concentration through mechanisms not correlated with growth rate. This is not

surprising, particularly given the growth breeding value itself (i.e. PWWT) also had no

impact on mineral content.

As for other oxidative capacity as well as objective and subjective measures of eating

quality, the impact of hot carcass weight on iron and zinc concentration (Pannier et al.,

2014b) was shown to be a poor proxy for growth, underestimating the effect.

228

8.12 Impact of growth on meat browning

Meat browning is costly to retailers due to discounting of otherwise safe and saleable

meat. The rate of meat browning is slower in heavier lambs, with the magnitude of the

effect increasing in the birth to weaning period then reaching a plateau (Chapter 7). This

was contrary to expectations as increased growth is associated with increased

intramuscular fat, which accelerates the rate of browning due to the auto oxidation of

lipid. Alternatively, the results align well with the associated reduction in oxidative

capacity with weight during the same period.

When the growth models were corrected for intramuscular fat there was no change in

the effect of growth on intramuscular fat, further emphasising that intramuscular fat

isn’t driving more rapid browning. When the growth models were corrected for whole

body adiposity, the magnitude of the impact on meat browning reduced by one third.

Thus the impact of growth on browning is only partially explained by its correlation

with adiposity.

As for the traits mentioned previously, the impacts of PFAT and PEMD on meat

browning were not delivered through mechanisms correlated with growth rate since

they did not change with correction for phenotypic growth. This was as expected given

the lack of association between PWWT and meat browning.

The effect of hot carcass weight (Calnan et al., 2014) was found to be a poor proxy for

the effect of growth, underestimating the effect on meat browning.

229

8.13 Impact of growth on meat traits

Several meat traits had a similar period in which they were influenced by growth and a

similar pattern of change with growth. These are likely to be linked to fibre type.

Birth to weaning: Intramuscular fat, myoglobin concentration, iron concentration and

meat browning are impacted by lamb growth during this period. Although growth

between birth and weaning is now known to be a key driver for the manipulation of

these traits, it is still a three month window. Greater precision would allow for targeted

manipulation of ewe nutrition to impact the traits. These traits may also be influenced

by in-utero growth as the association with weight inverts between birth and weaning.

Pre weaning: Zinc concentration and shear force are impacted by lamb growth during

this period.

Post weaning: ICDH activity is impacted by lamb growth during this period.

Arguably increased pre weaning growth leading to heavier lamb weights has a positive

impact on traits impacted in this period resulting in higher levels of intramuscular fat,

myoglobin and iron concentration, a reduction in shear force and meat browning. While

increased intramuscular fat with growth may become a consumer health issue in the

future its uncoupling from whole carcass adiposity using an ASBV would avoid large

visible layers of subcutaneous fat in lambs with increased intramuscular fat, minimising

the impact on visual appeal.

Associations between meat traits and lamb growth beyond 150 days were not

investigated in this study as the window for greatest changed was in the pre weaning

period for most traits. Other traits related to meat quality which are likely to be

influenced later in life, such as muscle glycogen concentration, would benefit for

230

investigation of associations later in life. Due to a lack of bias, random effects models

can be used to estimate growth later in life, despite a reducing proportion of animals

having weights recorded after the post weaning time point.

Environment had the largest effect on most traits and in general this was the same or

larger than the impact of growth on traits. Therefore, growth impacted most traits

however was not a large driver of the traits relative to the magnitude of environment,

although it had a larger effect than some production factors such as sex or sire type.

This warrants further investigation.

Further investigation may also include the change in magnitude of the impact of

production factors on traits, e.g. birth type, after correction for growth.

8.14 Adjustments to experimental design

Information Nucleus Flock Design

This thesis is based on information collected as part of the Information Nucleus Flock

experiment. The comprehensive design allows the assessment of production, phenotypic

and genetic effects (Fogarty et al., 2007; van der Werf et al., 2010) however there are

areas of the design that could be addressed differently. Ewe lambs from Maternal and

Merino progeny were retained so no slaughter information is available for these

animals. As a result sire type comparisons are only possible in wether progeny of

Merino dams where Terminal, Maternal and Merino sire types were represented

equally. Similarly it was only possible to compare dam breed effects within the

Terminal sired lambs as Border-Leicester x Merino dams were not used in the Merino

and Maternal sire type groups.

231

Many analyses highlighted the large effect of site or rearing, reflecting environment, on

traits. Within the design of the Information Nucleus Flock it was not possible to

measure all factors contributing to the effect of environment however these could

include pasture availability and feed type, worm burden and weather conditions. If these

factors were known in detail the impacts of nutrition, disease and climate could be

teased apart from the generic environmental effect. Measures were taken to reduce bias

caused by these factors by balancing the sites for key factors of sire type and breeding

value impacts by ensuring a representation of sires across all sites.

Frequency of weight recording between birth and weaning

In this experiment live weight was infrequently measured between birth and weaning at

approximately 100 days. The lack of data available in this time reduced confidence in

the ability to make a good estimation of growth between these times so it was difficult

to characterise how the relationship between growth and traits such as intramuscular fat

and myoglobin concentration changed between birth and weaning. Therefore, a

mechanistic experiment where lambs on divergent growth paths were weighed

frequently, every 7-10 days, would help to characterise the associations between growth

and meat quality traits at slaughter.

Accounting for wool growth

When growth functions were fitted to lamb live weight data no adjustment was made

for the approximate wool weight at the time of measuring live weight. Likewise no

adjustment was made for the reduction in live weight at shearing due to the loss of

wood. The impact of wool weight on the fit of the growth function was expected to be

small, as wool weight is a small proportion of live weight, particularly when wool

length is short. Furthermore, the effect of wool weight would be captured, within small

232

groups, by the kill group effect. Therefore no adjustment was made in this study.

However, since it is possible to estimate the weight of wool during lamb growth, and

accounting for this may make a small improvement in the precision of the growth

function fit, this would be a consideration in future experiments.

8.15 Future studies

Curve fitting methodology

Random effects models offer significant advantages in terms of flexibility in the

estimation of lamb weights however it is based on linear functions. An improved

scenario would be a random effects model which utilises a nonlinear function so that the

curve parameters would have a simple biological interpretation. The usefulness of a

curve with this methodology has been expressed by various authors and to this end a

random effects model based on a nonlinear Brody function is being developed

subsequent to this thesis.

Estimated or recorded weights

In this thesis growth curves were used to estimate lamb weights rather than using the

measurements routinely collected in industry at weaning and post weaning. This was

undertaken as estimated weights are less influenced (within each animal) by factors that

cause large day to day fluctuations in weight, such as gut fill. However it would be

useful to compare the outcome of using estimated weights with routinely collected

measurements in industry. If the magnitude of the effect of growth on traits was similar

using routinely collected measurements then the additional effort required to derive

growth curves may be unnecessary.

233

Growth rates between birth and weaning

Growth had the greatest impact on several traits between birth and weaning. Isolating

the specific window during which growth most heavily impacts these traits will allow

producers to have more control over manipulation of these traits. For example, traits

which are heavily influenced by growth in early life could be manipulated through dam

nutrition or the use of the ASBV for birth weight, which has a magnitude of effect as

large, or larger, than most production effects. To achieve this it would be necessary to

perform more frequent weighing of lambs during the birth to weaning period.

234

References

Afolayan, R. A., Fogarty, N. M., Ingham, V. M., Gilmour, A. R., Gaunt, G. M., Cummins, L. J., and

Pollard, T. (2007). Genetic evaluation of crossbred lamb production. 3. Growth and carcass

performance of second-cross lambs. Australian Journal of Agricultural Research, 58(5), 457-

466.

Agabriel, J., Bony, J., and Micol, D. (1998). Le bison d'Amérique: élevage, production et qualité de la

viande: Institut national de la recherche agronomique.

Akbas, Y., and Yaylak, E. (2000). Heritability estimates of growth curve parameters and genetic

correlations between the growth curve parameters and weights at different age of Japanese quail.

Archiv für Geflügelkunde, 64(4), 141-146.

Alexander, G., Stevens, D., Kilgour, R., De Langen, H., Mottershead, B. E., and Lynch, J. J. (1983).

Separation of ewes from twin lambs: incidence in several sheep breeds. Applied animal ethology,

10(4), 301-317.

Ali, S., Hussain, A., Akhtar, P., Younas, M., and Khan, M. A. (2006). Relationship of birth weight of

lambs with age of ewes at service in rambouillet sheep Pakistan veterinary journal, 26(4), 187-

189.

AMSA. (2012). Colour Measurement Guidelines. Champaign, IL, USA: American Meat Science

Association.

Andersen, S., and Pedersen, B. (1996). Growth and food intake curves for group-housed gilts and

castrated male pigs. Animal Science, 63(03), 457-464.

Anonymous. (2005). Handbook of Australian Meat, International red meat manual: AUS-MEAT:

Sydney, Australia, 7th Edition.

ARC, A. R. C. (1980). The nutrient requirements of ruminant livestock. Wallingford: CAB International.

Arnold, A. M., and Meyer, H. H. (1988). Influences of gender, time of castration, genotype, birthtype and

feeding regimen on lamb longissimus fiber type proportions. Journal of Animal Science, 66(10),

2476-2483.

Ashmore, C. R. (1974). Phenotypic expression of muscle fiber types and some implications to meat

quality. Journal of Animal Science, 38(5), 1158-1164.

Ashmore, C. R., and Doerr, L. (1971). Comparative aspects of muscle fiber types in different species.

Experimental neurology, 31(3), 408-418.

Ashmore, C. R., Robinson, D. W., Rattray, P., and Doerr, L. (1972). Biphasic development of muscle

fibers in the fetal lamb. Experimental neurology, 37(2), 241-255.

Atkins, K. D. (1980). The comparative productivity of five ewe breeds. 1. Lamb growth and survival.

Australian Journal of Experimental Agriculture, 20(104), 272-279.

Atkins, K. D., and Gilmour, A. R. (1981). The comparative productivity of five ewe breeds. 4. Growth

and carcase characteristics of purebred and crossbred lambs. Australian Journal of Experimental

Agriculture, 21(109), 172-178.

Atkins, K. D., and Thompson, J. M. (1979). Carcass characteristics of heavyweight crossbred lambs. I.

Growth and carcass measurements. Australian Journal of Agricultural Research, 30(6), 1197-

1205.

235

Ayyangar, L. (2007). Skewness, Multicollinearity, Heteroskedasticity-You Name It, Cost Data Have It!

Solutions to Violations of Assumptions of Ordinary Least Squares Regression Models Using

SAS®. In.

Bailey, A. J., Enser, M. B., Dransfield, E., Restall, D. J., and Avery, N. C. (1982). Muscle and adipose

tissue from normal and double muscled cattle: collagen types, muscle fibre diameter, fat cell size

and fatty acid composition and organoleptic properties. In Muscle hypertrophy of genetic origin

and its use to improve beef production (pp. 178-202): Springer.

Ball, A. J., Thompson, J. M., and Pleasants, A. B. (1996). Seasonal changes in body composition of

growing Merino sheep. Livestock Production Science, 46(3), 173-180.

Banks, R. (2002). Enhancing the value of meat from wool sheep: Is there a need for specialisation? Wool

Technology and Sheep Breeding, 50(4).

Banks, R. G., Ball, A. J., and Field, S. R. (2002). Young sire programs in meat sheep breeds in Australia -

a vehicle for rapid genetic improvement and continuous reserach and development. In

Proceedings of Seventh World Congress of Genetics Applied Livestock Production. Montipellier,

France: CD-ROM Communication No. 23-15.

Bär, A., and Pette, D. (1988). Three fast myosin heavy chains in adult rat skeletal muscle. FEBS letters,

235(1), 153-155.

Barnard, J. R., Edgerton, R. V., Furukawa, T., and Peter, J. B. (1971). Histochemical, biochemical, and

contractile properties of red, white, and intermediate fibers. American Journal of Physiology,

220(2), 410-414.

Bathaei, S. S., and Leroy, P. L. (1996). Growth and mature weight of Mehraban Iranian fat-tailed sheep.

Small Ruminant Research, 22(2), 155-162.

Bathaei, S. S., and Leroy, P. L. (1998). Genetic and phenotypic aspects of the growth curve

characteristics in Mehraban Iranian fat-tailed sheep. Small Ruminant Research, 29(3), 261-269.

Beaulieu, A. D., Aalhus, J. L., Williams, N. H., and Patience, J. F. (2010). Impact of piglet birth weight,

birth order, and litter size on subsequent growth performance, carcass quality, muscle

composition, and eating quality of pork. Journal of Animal Science, 88(8), 2767-2778.

Beecher, G. R., Kastenschmidt, L. L., Cassens, R. G., Hoekstra, W. G., and Briskey, E. J. (1968). A

comparison of the light and dark portions of a striated muscle. Journal of food science, 33(1),

84-88.

Bekhit, A. E. D., and Faustman, C. (2005). Metmyoglobin reducing activity. Meat Science, 71(3), 407-

439.

Bekhit, A. E. D., Geesink, G. H., Ilian, M. A., Morton, J. D., and Bickerstaffe, R. (2003). The effects of

natural antioxidants on oxidative processes and metmyoglobin reducing activity in beef patties.

Food Chemistry, 81(2), 175-187.

Bekhit, A. E. D., Geesink, G. H., Morton, J. D., and Bickerstaffe, R. (2001). Metmyoglobin reducing

activity and colour stability of ovine longissimus muscle. Meat Science, 57(4), 427-435.

Bellof, G., Most, E., and Pallauf, J. (2007). Concentration of copper, iron, manganese and zinc in muscle,

fat and bone tissue of lambs of the breed German Merino Landsheep in the course of the

growing period and different feeding intensities. Journal of Animal Physiology and Animal

Nutrition, 91(3-4), 100-108.

Beltrán, J. J., Butts, W. T., Olson, T. A., and Koger, M. (1992). Growth patterns of two lines of Angus

cattle selected using predicted growth parameters. Journal of Animal Science, 70(3), 734-741.

Bérard, J., Kreuzer, M., and Bee, G. (2010). In large litters birth weight and gender is decisive for growth

performance but less for carcass and pork quality traits. Meat Science, 86(3), 845-851.

236

Berg, R. T., and Butterfield, R. M. (1968). Growth Patterns of Bovine Muscle, Fat and Bone. Journal of

Animal Science, 27(3), 611.

Beriain, M. J., Horcada, A., Purroy, A., Lizaso, G., Chasco, J., and Mendizabal, J. A. (2000).

Characteristics of Lacha and Rasa Aragonesa lambs slaughtered at three live weights. Journal of

Animal Science, 78(12), 3070-3077.

Bernard, C., Cassar-Malek, I., Renand, G., and Hocquette, J.-F. (2009). Changes in muscle gene

expression related to metabolism according to growth potential in young bulls. Meat Science,

82(2), 205-212.

Borg, R. C., Notter, D. R., and Kott, R. W. (2009). Phenotypic and genetic associations between lamb

growth traits and adult ewe body weights in western range sheep. Journal of Animal Science,

87(11), 3506-3514.

Bouton, P. E., Ford, A. L., Harris, P. V., Shorthose, W. R., Ratcliff, D., and Morgan, J. H. L. (1978).

Influence of animal age on the tenderness of beef: Muscle differences. Meat Science, 2(4), 301-

311.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of

protein utilizing the principle of protein-dye binding. Analytical Biochemistry, 72(1-2), 248-254.

Brandstetter, A. M., Picard, B., and Geay, Y. (1998a). Muscle fibre characteristics in four muscles of

growing bulls: I. Postnatal differentiation. Livestock Production Science, 53(1), 15-23.

Brandstetter, A. M., Picard, B., and Geay, Y. (1998b). Muscle fibre characteristics in four muscles of

growing male cattle: II. Effect of castration and feeding level. Livestock Production Science,

53(1), 25-36.

Briand, M. (1981). Metabolic types of muscle in the sheep: I. Myosin ATPase, glycolytic, and

mitochondrial enzyme activities. European journal of applied physiology and occupational

physiology, 46(4), 347-358.

Brody, S. (1945). Bioenergetics and growth. New York: Reinhold Publishing.

Brooke, M. H., and Kaiser, K. K. (1970). Muscle fiber types: How many and what kind? Archives of

Neurology, 23(4), 369-379.

Brown, D. J., Atkins, K. D., and Huisman, A. E. (2005). Expression of body weight, fleece weight and

fibre diameter in across flock genetic evaluation. In Proceedings of the Association for the

Advancement of Animal Breeding and Genetics (Vol. 16:84).

Brown, D. J., Ball, A., Huisman, A. E., Swan, A. A., Atkins, K. D., Graser, H. U., Banks, R., Swan, P.,

and Woolaston, R. (2006). Sheep Genetics Australia: a national genetic evaluation system for

Australian sheep. In Proceedings of the 8th World Congress on Genetics Applied to Livestock

Production (pp. 05-03). Minas Gerais, Brazil: Instituto Prociência.

Brown, D. J., Huisman, A. E., Swan, A. A., Graser, H. U., Woolaston, R. R., Ball, A. J., Atkins, K. D.,

and Banks, R. G. (2007). Genetic evaluation for the Australian sheep industry. In Proceedings

of the Association for the Advancement of Animal Breeding and Genetics (Vol. 17, pp. 187-194).

Brown, D. J., Jones, R. M., and Hinch, G. N. (2014). Genetic parameters for lamb autopsy traits. Animal

Production Science, 54(6), 736-744.

Brown, D. J., and Reverter, A. (2002). A comparison of methods to pre-adjust data for systematic

environmental effects in genetic evaluation of sheep. Livestock Production Science, 75(3), 281-

291.

Brown, D. J., Savage, D. B., Hinch, G. N., and Hatcher, S. (2015). Monitoring liveweight in sheep is a

valuable management strategy: a review of available technologies. Animal Production Science,

55(4), 427-436.

237

Brown, J. E. (1970). A comparison of five stochastic models on their ability to describe the weight-age

relationship in cattle. Ph.D. Dissertation, Texas A&M Univeristy, College Station.

Brown, J. E., Fitzhugh, H. A., and Cartwright, T. C. (1976). A Comparison of Nonlinear Models for

Describing Weight-Age Relationships in Cattle. Journal of Animal Science, 42(4), 810-818.

Buckley, D. J., Morrissey, P. A., and Gray, J. I. (1995). Influence of dietary vitamin E on the oxidative

stability and quality of pig meat. Journal of Animal Science, 73(10), 3122-3130.

Burris, M. J., and Baugus, C. A. (1955). Milk Consumption and Growth of Suckling Lambs. Journal of

Animal Science, 14(1), 186-191.

Butterfield, R. M. (1988). New concepts of sheep growth: Department of Veterinary Anatomy, University

of Sydney.

Butterfield, R. M., Griffiths, D. A., Thompson, J. M., Zamora, J., and James, A. M. (1983). Changes in

body composition relative to weight and maturity in large and small strains of Australian Merino

rams 1. Muscle, bone and fat. Animal Science, 36(01), 29-37.

Butterfield, R. M., Zamora, J., Thompson, J. M., Reddacliff, K. J. R., and Griffiths, D. A. (1984).

Changes in body composition relative to weight and maturity of Australian Dorset Horn rams

and wethers. 1. Carcass muscle, fat and bone and body organs. Animal Science, 39(02), 251-258.

Cafe, L. M., Hennessy, D. W., Hearnshaw, H., Morris, S. G., and Greenwood, P. L. (2006). Influences of

nutrition during pregnancy and lactation on birth weights and growth to weaning of calves sired

by Piedmontese or Wagyu bulls. Animal Production Science, 46(2), 245-255.

Calkins, C. R., Dutson, T. R., Smith, G. C., Carpenter, Z. L., and Davis, G. W. (1981). Relationship of

fiber type composition to marbling and tenderness of bovine muscle. Journal of food science,

46(3), 708-710.

Calkins, C. R., Seideman, S. C., and Crouse, J. D. (1987). Relationships between rate of growth, catheptic

enzymes and meat palatability in young bulls. Journal of Animal Science, 566.

Calnan, H. B., Jacob, R. H., Pethick, D. W., and Gardner, G. E. (2014). Factors affecting the colour of

lamb meat from the longissimus muscle during display: the influence of muscle weight and

muscle oxidative capacity. Meat Science, 96(2 Pt B), 1049-1057.

Čandek-Potokar, M., Lefaucheur, L., Zlender, B., and Bonneau, M. (1999). Effect of slaughter weight

and/or age on histological characteristics of pig longissimus dorsi muscle as related to meat

quality. Meat Science, 52(2), 195-203.

Cassens, R. G., and Cooper, C. C. (1971). Red and white muscle. Advances in food research, 19, 1-74.

Cassens, R. G., Hoekstra, W. G., Faltin, E. C., and Briskey, E. J. (1967). Zinc content and subcellular

distribution in red vs. white porcine skeletal muscle (Vol. 212).

Channon, H. A., Reynolds, J., and Baud, S. (2001). Identifying pathways to ensure acceptable eating

quality of pork. In Final Report. Canberra: Australian Pork Limited.

Cho, M., Hughes, S. M., Karsch-Mizrachi, I., Travis, M., Leinwand, L. A., and Blau, H. M. (1994). Fast

myosin heavy chains expressed in secondary mammalian muscle fibers at the time of their

inception. Journal of cell science, 107(9), 2361-2371.

Choi, Y. M., and Kim, B. C. (2009). Muscle fiber characteristics, myofibrillar protein isoforms, and meat

quality. Livestock Science, 122(2), 105-118.

Clark, S. T., Odell, D. K., and Lacinak, C. T. (2000). Aspects of growth in captive killer whales (Orcinus

orca). Marine Mammal Science, 16(1), 110-123.

238

Clarke, B. E., Brown, D. J., and Ball, A. J. (2003). Preliminary genetic parameters for live weight and

ultrasound scan traits in Merinos. In Proceedings of the Association for the Advancement of

Animal Breeding and Genetics (Vol. 15, pp. 326-330).

Clinquart, A., Van Eenaeme, C., Van Vooren, T., Van Hoof, J., Hornick, J.-L., and Istasse, L. (1994).

Meat quality in relation to breed (Belgian Blue vs Holstein) and conformation (double muscled

vs dual purpose type). Sciences des aliments, 14(4), 401-407.

Condon, K., Silberstein, L., Blau, H. M., and Thompson, W. J. (1990). Development of muscle fiber

types in the prenatal rat hindlimb. Dev Biol, 138(2), 256-274.

Craigie, C. R., Lambe, N. R., Richardson, R. I., Haresign, W., Maltin, C. A., Rehfeldt, C., Roehe, R.,

Morris, S. T., and Bunger, L. (2012). The effect of sex on some carcass and meat quality traits in

Texel ewe and ram lambs. Animal Production Science, 52(7), 601-607.

Cullis, B. R., and McGilchrist, C. A. (1990). A model for the analysis of growth data from designed

experiments. Biometrics, 131-142.

Curtis, K., Dolling, M., Cronjé, P., and Maxwell, D. K. (2006). The sheep industry-Australia in the global

scene. In Proceedings of the 2006 Australian Sheep Industry Cooperative Research Centre

Conference, Orange, Australia, 22-23 February 2006. (pp. 16-25): Australian Sheep Industry

CRC.

Da Silva, P., Aitken, R. P., Rhind, S. M., Racey, P. A., and Wallace, J. M. (2001). Influence of

placentally mediated fetal growth restriction on the onset of puberty in male and female lambs.

Reproduction, 122(3), 375-383.

de Freitas, A. R. (2005). Curvas de crescimento na produção animal. R. Bras. Zootec, 34(3), 786-795.

Denise, R. S., and Brinks, J. S. (1985). Genetic and environmental aspects of the growth curve parameters

in beef cows. Journal of Animal Science, 61(6), 1431-1440.

Díaz, M. T., Alvarez, I., De la Fuente, J., Sañudo, C., Campo, M. M., Oliver, M. A., Font i Furnols, M.,

Montossi, F., San Julián, R., and Nute, G. R. (2005). Fatty acid composition of meat from typical

lamb production systems of Spain, United Kingdom, Germany and Uruguay. Meat Science,

71(2), 256-263.

Dikeman, M. E. (1987). Fat reduction in animals and the effects on palatability and consumer acceptance

of meat products. In Proceedings-Annual Reciprocal Meat Conference of the American Meat

Science Association (USA).

Doornenbal, H., and Murray, A. C. (1982). Effects of age, breed, sex and muscle on certain mineral

concentrations in cattle. Journal of food science, 47(1), 55-58.

Doren, P. E., Baker, J. F., Long, C. R., and Cartwright, T. C. (1989). Estimating parameters of growth

curves of bulls. Journal of Animal Science, 67(6), 1432-1445.

Dubowitz, V., and Pearse, E. A. G. (1960). A comparative histochemical study of oxidative enzyme and

phosphorylase activity in skeletal muscle. Histochemistry and Cell Biology, 2(2), 105-117.

Emmans, G. C. (1997). A Method to Predict the Food Intake of Domestic Animals from Birth to Maturity

as a Function of Time. Journal of Theoretical Biology, 186(2), 189-200.

Engel, W. K. (1962). The essentiality of histo-and cytochemical studies of skeletal muscle in the

investigation of neuromuscular disease. Neurology, 12, 778-784.

Essén-Gustavsson, B., Karlsson, A., Lundström, K., and Enfält, A. C. (1994). Intramuscular fat and

muscle fibre lipid contents in halothane-gene-free pigs fed high or low protein diets and its

relation to meat quality. Meat Science, 38(2), 269-277.

239

Essen-Gustavsson, B., Karlström, K., and Lundström, K. (1992). Muscle fibre characteristics and

metabolic response at slaughter in pigs of different halothane genotypes and their relation to

meat quality. Meat Science, 31(1), 1-11.

Everett-Hincks, J. M., Blair, H. T., Stafford, K. J., Lopez-Villalobos, N., Kenyon, P. R., and Morris, S. T.

(2005). The effect of pasture allowance fed to twin- and triplet-bearing ewes in late pregnancy

on ewe and lamb behaviour and performance to weaning. Livestock Production Science, 97(2–

3), 253-266.

Fahey, A. J., Brameld, J. M., Parr, T., and Buttery, P. J. (2005). The effect of maternal undernutrition

before muscle differentiation on the muscle fiber development of the newborn lamb. Journal of

Animal Science, 83(11), 2564-2571.

Faustman, C., and Cassens, R. G. (1990). The biochemical basis for discoloration in fresh meat: A

review. Journal of Muscle Foods, 1(3), 217-243.

Faustman, C., Cassens, R. G., Schaefer, D. M., Buege, D. R., Williams, S. N., and Scheller, K. K. (1989).

Improvement of pigment and lipid stability in Holstein steer beef by dietary supplementation

with vitamin E. Journal of food science, 54(4), 858-862.

Faustman, C., Liebler, D. C., McClure, T. D., and Sun, Q. (1999). α, β-Unsaturated aldehydes accelerate

oxymyoglobin oxidation. Journal of agricultural and food chemistry, 47(8), 3140-3144.

Faustman, C., Sun, Q., Mancini, R., and Suman, S. P. (2010). Myoglobin and lipid oxidation interactions:

Mechanistic bases and control. Meat Science, 86(1), 86-94.

Ferguson, N. S., and Gous, R. M. (1993a). Evaluation of pig genotypes 1. Theoretical aspects of

measuring genetic parameters. Animal Production, 56(02), 233-243.

Ferguson, N. S., and Gous, R. M. (1993b). Evaluation of pig genotypes 2. Testing experimental

procedure. Animal Production, 56(02), 245-249.

Fischer, T. M., and Van der Werf, J. H. J. (2002). Effect of data structure on the estimation of genetic

parameters using random regression. In Proceedings of the 7th World Congress on Genetics

Applied to Livestock Production, Montpellier, France, August, 2002. Session 17. (pp. 1-4):

Institut National de la Recherche Agronomique (INRA).

Fischer, T. M., Van der Werf, J. H. J., Banks, R. G., and Ball, A. J. (2004). Description of lamb growth

using random regression on field data. Livestock Production Science, 89(2-3), 175-185.

Fishell, V. K., Aberle, E. D., Judge, M. D., and Perry, T. W. (1985). Palatability and Muscle Properties of

Beef as Influenced by Preslaughter Growth Rate. Journal of Animal Science, 61(1), 151-157.

Fisher, A. V., Enser, M., Richardson, R. I., Wood, J. D., Nute, G. R., Kurt, E., Sinclair, L. A., and

Wilkinson, R. G. (2000). Fatty acid composition and eating quality of lamb types derived from

four diverse breed × production systems. Meat Science, 55(2), 141-147.

Fitzhugh, H. A. (1976). Analysis of Growth Curves and Strategies for Altering Their Shape. Journal of

Animal Science, 42(4), 1036-1051.

Fogarty, N. M., Banks, R. G., van der Werf, J. H. J., Ball, A. J., and Gibson, J. P. (2007). The information

nucleus-a new concept to enhance sheep industry genetic improvement. In Proceedings of the

Association for the Advancement of Animal Breeding and Genetics (Vol. 17, pp. 29-32).

Fogarty, N. M., Dickerson, G. E., and Young, L. D. (1984). Lamb Production and Its Components in Pure

Breeds and Composite Lines. II. Breed Effects and Heterosis. Journal of Animal Science, 58(2),

301-311.

Fogarty, N. M., Gilmour, A. R., and Hopkins, D. L. (1997). The relationship of crossbred lamb growth

and carcass traits with lambplan EBVs of sires. In Association for the Advancement of Animal

Breeding and Genetics. (Vol. 12, pp. 304-309). Dubbo, NSW: AAABG Distribution Service

AGBU.

240

Fogarty, N. M., Hall, D. G., and Holst, P. J. (1992). The effect of nutrition in mid pregnancy and ewe

liveweight change on birth weight and management for lamb survival in highly fecund ewes.

Australian Journal of Experimental Agriculture, 32(1), 1-10.

Fogarty, N. M., Hopkins, D. L., and van de Ven, R. (2000). Lamb production from diverse genotypes 1.

Lamb growth and survival and ewe performance. Animal Science, 70(1), 135-145.

Fox, J., and Jay, B. (1966). Chemistry of meat pigments. Journal of agricultural and food chemistry,

14(3), 207-210.

Friggens, N. C., Shanks, M., Kyriazakis, I., Oldham, J. D., and McClelland, T. H. (1997). The growth and

development of nine European sheep breeds. 1. British breeds: Scottish Blackface, Welsh

Mountain and Shetland. Animal Science, 65(03), 409-426.

Gagniere, H., Picard, B., Jurie, C., and Geay, Y. (1999). Comparison of foetal metabolic differentiation in

three cattle muscles. Reproduction Nutrition Development, 39(1), 105-112.

Gardner, G. E., Anderson, F., Williams, A., Kelman, K. R., Pannier, L., and Pethick, D. (2012). Growth

breeding value redistributes weight to the saddle region of lamb carcasses. In 63rd Annual

Meeting of the European Federation of Animal Science (pp. 302). Bratislava, Slovakia.

Gardner, G. E., Hopkins, D. L., Greenwood, P. L., Cake, M. A., Boyce, M. D., and Pethick, D. W.

(2007). Sheep genotype, age and muscle type affect the expression of metabolic enzyme

markers. Australian Journal of Experimental Agriculture, 47(10), 1180-1189.

Gardner, G. E., Kennedy, L., Milton, J. T. B., and Pethick, D. W. (1999). Glycogen metabolism and

ultimate pH of muscle in Merino, first-cross, and second-cross wether lambs as affected by stress

before slaughter. Australian Journal of Agricultural Research.

Gardner, G. E., Pethick, D., Hopkins, D. L., Hegarty, R. S., Cake, M. A., Boyce, M. D., and Allingham,

P. G. (2006a). The impact of carcase estimated breeding values on yield and quality of sheep

meat. In Wool meets Meat. Orange, NSW: Sheep CRC.

Gardner, G. E., Pethick, D. W., Greenwood, P. L., and Hegarty, R. S. (2006b). The effect of genotype and

plane of nutrition on the rate of pH decline in lamb carcasses and the expression of metabolic

enzymatic markers. Australian Journal of Agricultural Research, 57(6), 661-670.

Gardner, G. E., Williams, A., Ball, A. J., Jacob, R. H., Refshauge, G., Hocking Edwards, J., Behrendt, R.,

and Pethick, D. W. (2015). Carcase weight and dressing percentage are increased using

Australian Sheep Breeding Values for increased weight and muscling and reduced fat depth.

Meat Science, 99, 89-98.

Gardner, G. E., Williams, A., Siddell, J., Ball, A. J., Mortimer, S., Jacob, R. H., Pearce, K. L., Hocking

Edwards, J. E., Rowe, J. B., and Pethick, D. W. (2010). Using Australian Sheep Breeding Values

to increase lean meat yield percentage. Animal Production Science, 50(12), 1098-1106.

Gbangboche, A. B., Adamou-Ndiaye, M., Youssao, A. K. I., Farnir, F., Detilleux, J., Abiola, F. A., and

Leroy, P. L. (2006a). Non-genetic factors affecting the reproduction performance, lamb growth

and productivity indices of Djallonke sheep. Small Ruminant Research, 64(1–2), 133-142.

Gbangboche, A. B., Glele-Kakai, R., Salifou, S., Albuquerque, L. G., and Leroy, P. L. (2008).

Comparison of non-linear growth models to describe the growth curve in West African Dwarf

sheep. Animal: an International Journal of Animal Bioscience, 2(7), 1003-1003-1012.

Gbangboche, A. B., Youssao, A. K. I., Senou, M., Adamou-Ndiaye, M., Ahissou, A., Farnir, F., Michaux,

C., Abiola, F. A., and Leroy, P. L. (2006b). Examination of non-genetic factors affecting the

growth performance of djallonke sheep in soudanian zone at the Okpara breeding farm of Benin.

Tropical Animal Health and Production, 38(1), 55-64.

Giddings, G. G., and Solberg, M. (1977). The basis of color in muscle foods. C R C Critical Reviews in

Food Science and Nutrition, 9(1), 81-114.

241

Girard, B., Vanderstoep, J., and Richards, J. F. (1990). Characterization of the Residual Pink Color in

Cooked Turkey Breast and Pork Loin. Journal of food science, 55(5), 1249-1254.

Glitsch, K. (2000). Consumer perceptions of fresh meat quality: cross‐national comparison. British Food

Journal, 102(3), 177-194.

Gondret, F., Ferré, P., and Dugail, I. (2001). ADD-1/SREBP-1 is a major determinant of tissue

differential lipogenic capacity in mammalian and avian species. Journal of lipid research, 42(1),

106-113.

Gondret, F., Lefaucheur, L., Juin, H., Louveau, I., and Lebret, B. (2006). Low birth weight is associated

with enlarged muscle fiber area and impaired meat tenderness of the longissimus muscle in pigs.

Journal of Animal Science, 84(1), 93-103.

Gondret, F., Mourot, J., and Bonneau, M. (1998). Comparison of intramuscular adipose tissue cellularity

in muscles differing in their lipid content and fibre type composition during rabbit growth.

Livestock Production Science, 54(1), 1-10.

Goonewardene, L. A., Berg, R. T., and Hardin, R. T. (1981). A growth study of beef cattle. Canadian

Journal of Animal Science, 61(4), 1041-1048.

Gous, R. M., Moran, E. T., Stilborn, H. R., Bradford, G. D., and Emmans, G. C. (1999). Evaluation of the

parameters needed to describe the overall growth, the chemical growth, and the growth of

feathers and breast muscles of broilers. Poultry Science, 78(6), 812-821.

Greenwood, P. L., and Cafe, L. M. (2007). Prenatal and pre-weaning growth and nutrition of cattle: long-

term consequences for beef production. animal, 1(09), 1283-1296.

Greenwood, P. L., Cafe, L. M., Hearnshaw, H., Hennessy, D. W., and Morris, S. G. (2009).

Consequences of prenatal and preweaning growth for yield of beef primal cuts from 30-month-

old Piedmontese-and Wagyu-sired cattle. Animal Production Science, 49(6), 468-478.

Greenwood, P. L., Cafe, L. M., Hearnshaw, H., Hennessy, D. W., Thompson, J. M., and Morris, S. G.

(2006a). Long-term consequences of birth weight and growth to weaning on carcass, yield and

beef quality characteristics of Piedmontese- and Wagyu-sired cattle. Australian Journal of

Experimental Agriculture, 46(2), 257-269.

Greenwood, P. L., Davis, J. J., Gaunt, G. M., and Ferrier, G. R. (2006b). Influences on the loin and

cellular characteristics of the m. longissimus lumborum of Australian Poll Dorset-sired lambs.

Australian Journal of Agricultural Research, 57(1), 1-12.

Greenwood, P. L., Gardner, G. E., and Hegarty, R. S. (2006c). Lamb myofibre characteristics are

influenced by sire estimated breeding values and pastoral nutritional system. Australian Journal

of Agricultural Research, 57(6), 627-639.

Greenwood, P. L., Harden, S., and Hopkins, D. L. (2007). Myofibre characteristics of ovine longissimus

and semitendinosus muscles are influenced by sire breed, gender, rearing type, age and carcass

weight. Australian Journal of Experimental Agriculture, 47(10), 1137-1146.

Greenwood, P. L., Hunt, A. S., Hermanson, J. W., and Bell, A. W. (2000a). Effects of birth weight and

postnatal nutrition on neonatal sheep: II. Skeletal muscle growth and development. Journal of

Animal Science, 78(1), 50-61.

Greenwood, P. L., Slepetis, R. M., and Bell, A. W. (2000b). Influences on fetal and placental weights

during mid to late gestation in prolific ewes well nourished throughout pregnancy. Reproduction,

Fertility and Development, 12(4), 149-156.

Hall, D. G. (2000). Opportunities for meat from lambs and goats in Australia. Asian Australian Journal of

Animal Sciences, 13, 91-92.

Hall, D. G., Gilmour, A. R., Fogarty, N. M., and Holst, P. J. (2002). Growth and carcass composition of

second-cross lambs. 2. Relationship between estimated breeding values of sires and their

242

progeny performance under fast and slow growth regimes. Australian Journal of Agricultural

Research, 53(12), 1341.

Hancock, C. E., Bradford, G. D., Emmans, G. C., and Gous, R. M. (1995). The evaluation of the growth

parameters of six strains of commercial broiler chickens. British Poultry Science, 36(2), 247-

264.

Harper, G. S., and Pethick, D. W. (2004). How might marbling begin? Animal Production Science, 44(7),

653-662.

Harris, A. J., Fitzsimons, R. B., and McEwan, J. C. (1989). Neural control of the sequence of expression

of myosin heavy chain isoforms in foetal mammalian muscles. Development, 107(4), 751-769.

Hartley, H. O. (1961). The modified Gauss-Newton method for the fitting of non-linear regression

functions by least squares. Technometrics, 3(2), 269-280.

Hawkins, R. R., Moody, W. G., and Kemp, J. D. (1985). Influence of genetic type, slaughter weight and

sex on ovine muscle fiber and fat-cell development. Journal of Animal Science, 61(5), 1154-

1163.

Hazell, T. (1982). Iron and zinc compounds in the muscle meats of beef, lamb, pork and chicken. Journal

of the Science of Food and Agriculture, 33(10), 1049-1056.

Hegarty, P. V. J., and Allen, C. E. (1978). Effect of pre-natal runting on the post-natal development of

skeletal muscles in swine and rats. Journal of Animal Science, 46(6), 1634-1640.

Hegarty, R. S., Hopkins, D. L., Farrell, T. C., Banks, R., and Harden, S. (2006a). Effects of available

nutrition and sire breeding values for growth and muscling on the development of crossbred

lambs. 2: Composition and commercial yield. Australian Journal of Agricultural Research,

57(6), 617-626.

Hegarty, R. S., Shands, C., Marchant, R., Hopkins, D. L., Ball, A. J., and Harden, S. (2006b). Effects of

available nutrition and sire breeding values for growth and muscling on the development of

crossbred lambs. 1: Growth and carcass characteristics. Australian Journal of Agricultural

Research, 57(6), 593-603.

Hegarty, R. S., Warner, R. D., and Pethick, D. W. (2006c). Genetic and nutritional regulation of lamb

growth and muscle characteristics. Australian Journal of Agricultural Research, 57(6), 721-730.

Hendrick, H. B., Aberle, E. D., Forrest, J. C., Judge, M. D., and Merkel, R. A. (1989). Principles of Meat

Science. In: Kendall Hunt Publishing Company, Dubuque, IA.

Hernández, P., Aliaga, S., Pla, M., and Blasco, A. (2004). The effect of selection for growth rate and

slaughter age on carcass composition and meat quality traits in rabbits. Journal of Animal

Science, 82(11), 3138-3143.

Hocquette, J. F., Cassar-Malek, I., Listrat, A., Jurie, C., Jailler, R., and Picard, B. (2001). Some recent

advances in muscle biology and its regulation by nutrition: consequences on bovine meat

quality. Recent Advances in Animal Nutrition in Australia, 13, 135-143.

Hocquette, J. F., Cassar‐Malek, I., Jurie, C., Bauchart, D., Picard, B., and Renand, G. (2012).

Relationships between muscle growth potential, intramuscular fat content and different

indicators of muscle fibre types in young Charolais bulls. Animal Science Journal, 83(11), 750-

758.

Hocquette, J. F., Gondret, F., Baéza, E., Médale, F., Jurie, C., and Pethick, D. W. (2010). Intramuscular

fat content in meat-producing animals: development, genetic and nutritional control, and

identification of putative markers. animal, 4(02), 303-319.

Hocquette, J. F., Jurie, C., Ueda, Y., Boulesteix, P., Bauchart, D., and Pethick, D. (2003). The

relationship between muscle metabolic pathways and marbling of beef. In Proceedings of the

62nd annual meeting of the european federation of animal science (pp. 513-516).

243

Hocquette, J. F., Ortigues-Marty, I., Pethick, D., Herpin, P., and Fernandez, X. (1998). Nutritional and

hormonal regulation of energy metabolism in skeletal muscles of meat-producing animals.

Livestock Production Science, 56(2), 115-143.

Hoh, J. F., Hughes, S., Hale, P. T., and Fitzsimons, R. B. (1988). Immunocytochemical and

electrophoretic analyses of changes in myosin gene expression in cat limb fast and slow muscles

during postnatal development. J Muscle Res Cell Motil, 9(1), 30-47.

Holst, P. J., Fogarty, N. M., and Stanley, D. F. (2002). Birth weights, meningeal lesions, and survival of

diverse genotypes of lambs from Merino and crossbred ewes. Australian Journal of Agricultural

Research, 53(2), 175-181.

Hopkins, D. L., and Fogarty, N. M. (1998). Diverse lamb genotypes—2. Meat pH, colour and tenderness.

Meat Science, 49(4), 477-488.

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

age, intramuscular fat, cooking loss, pH, shear force and eating quality of aged meat from sheep.

Australian Journal of Experimental Agriculture, 46(7), 879-884.

Hopkins, D. L., and Mortimer, S. I. (2014). Effect of genotype, gender and age on sheep meat quality and

a case study illustrating integration of knowledge. Meat Science, 98(3), 544-555.

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

age effects on sheep meat production 3. Meat quality. Australian Journal of Experimental

Agriculture, 47(10), 1155-1164.

Hopkins, D. L., Toohey, E. S., Warner, R. D., Kerr, M. J., and van de Ven, R. (2010). Measuring the

shear force of lamb meat cooked from frozen samples: comparison of two laboratories. Animal

Production Science, 50(6), 382-385.

Hopkins, D. L., Walker, P. J., Thompson, J. M., and Pethick, D. W. (2005). Effect of sheep type on meat

and eating quality of sheep meat. Animal Production Science, 45(5), 499-507.

Hoppeler, H. (1985). Muscle aerobic potential in the animal kingdom. Biochemistry of Exercise VI,

International Series on Sport Sciences. Champaign, IL: Human Kinetics Publishers, 417-434.

Huffman, K. L., Miller, M. F., Hoover, L. C., Wu, C. K., Brittin, H. C., and Ramsey, C. B. (1996). Effect

of beef tenderness on consumer satisfaction with steaks consumed in the home and restaurant.

Journal of Animal Science, 74(1), 91-97.

Huisman, A. E., and Brown, D. J. (2008). Genetic parameters for bodyweight, wool, and disease

resistance and reproduction traits in Merino sheep. 2. Genetic relationships between bodyweight

traits and other traits. Australian Journal of Experimental Agriculture, 48(9), 1186-1193.

Huisman, A. E., Brown, D. J., Ball, A. J., and Graser, H. (2008). Genetic parameters for bodyweight,

wool, and disease resistance and reproduction traits in Merino sheep. 1. Description of traits,

model comparison, variance components and their ratios. Australian Journal of Experimental

Agriculture, 48(9), 1177-1185.

Hunt, M. C., Acton, J. C., Benedict, R. C., Calkins, C. R., Cornforth, D. P., Jeremiah, L. E., Olson, D. G.,

Salam, C. P., Savell, J. W., and Shivas, S. D. (2001). American meat science association

committee on guidelines for meat color evaluation. In Proceedings of the Reciprocal Meat

Conference (44).

Hunt, M. C., Mancini, R. A., Hachmeister, K. A., Kropf, D. H., Merriman, M., Lduca, G., and Milliken,

G. (2004). Carbon monoxide in modified atmosphere packaging affects color, shelf life, and

microorganisms of beef steaks and ground beef. Journal of food science, 69(1), FCT45-FCT52.

Hwang, I. H., Polkinghorne, R., Lee, J. M., and Thompson, J. M. (2008). Demographic and design effects

on beef sensory scores given by Korean and Australian consumers. Animal Production Science,

48(11), 1387-1395.

244

Hwang, Y.-H., Kim, G.-D., Jeong, J.-Y., Hur, S.-J., and Joo, S.-T. (2010). The relationship between

muscle fiber characteristics and meat quality traits of highly marbled Hanwoo (Korean native

cattle) steers. Meat Science, 86(2), 456-461.

Jacob, R. H., D'Antuono, M. F., Gilmour, A. R., and Warner, R. D. (2014). Phenotypic characterisation of

colour stability of lamb meat. Meat Science, 96(2), 1040-1048.

Jacob, R. H., D'Antuono, M. F., Smith, G. M., Pethick, D. W., and Warner, R. D. (2007). Effect of lamb

age and electrical stimulation on the colour stability of fresh lamb meat. Crop and Pasture

Science, 58(4), 374-382.

Jeong, J. Y., Hur, S. J., Yang, H. S., Moon, S. H., Hwang, Y. H., Park, G. B., and Joo, S. T. (2009).

Discoloration characteristics of 3 major muscles from cattle during cold storage. Journal of food

science, 74(1), C1-C5.

Jeyamkondan, S., Jayas, D. S., and Holley, R. A. (2000). Review of centralized packaging systems for

distribution of retail-ready meat. Journal of Food Protection®, 63(6), 796-804.

Joo, S. T., Kim, G. D., Hwang, Y. H., and Ryu, Y. C. (2013). Control of fresh meat quality through

manipulation of muscle fiber characteristics. Meat Science, 95(4), 828-836.

Jose, C. G. (2011). Anti-and Pro-oxidant interactions with myoglobin in Lamb meat, and the influence on

colour stability. Murdoch University.

Joubert, D. M. (1956). An analysis of factors influencing post-natal growth and development of the

muscle fibre. The Journal of Agricultural Science, 47(01), 59-102.

Juengel, J. L., Hudson, N. L., Whiting, L., and McNatty, K. P. (2004). Effects of Immunization Against

Bone Morphogenetic Protein 15 and Growth Differentiation Factor 9 on Ovulation Rate,

Fertilization, and Pregnancy in Ewes. Biology of Reproduction, 70(3), 557-561.

Jurie, C., Cassar-Malek, I., Bonnet, M., Leroux, C., Bauchart, D., Boulesteix, P., Pethick, D. W., and

Hocquette, J. F. (2007). Adipocyte fatty acid-binding protein and mitochondrial enzyme

activities in muscles as relevant indicators of marbling in cattle1. Journal of Animal Science,

85(10), 2660-2669.

Jurie, C., Martin, J.-F., Listrat, A., Jailler, R., Culioli, J., and Picard, B. (2005). Effects of age and breed

of beef bulls on growth parameters, carcass and muscle characteristics. Animal Science, 80(03),

257-263.

Kaps, M., Herring, W. O., and Lamberson, W. R. (1999). Genetic and environmental parameters for

mature weight in Angus cattle. Journal of Animal Science, 77(3), 569-574.

Kaps, M., Herring, W. O., and Lamberson, W. R. (2000). Genetic and environmental parameters for traits

derived from the Brody growth curve and their relationships with weaning weight in Angus

cattle. Journal of Animal Science, 78(6), 1436-1442.

Karlsson, A., Enfält, A.-C., Essén-Gustavsson, B., Lundström, K., Rydhmer, L., and Stern, S. (1993).

Muscle histochemical and biochemical properties in relation to meat quality during selection for

increased lean tissue growth rate in pigs. Journal of Animal Science, 71(4), 930-938.

Kelman, K. R., Pannier, L., Pethick, D. W., and Gardner, G. E. (2014). Selection for lean meat yield in

lambs reduces indicators of oxidative metabolism in the longissimus muscle. Meat Science,

96(2), 1058-1067.

Kemp, J. D., Mahyuddin, M., Ely, D. G., Fox, J. D., and Moody, W. G. (1980). Effect of feeding systems,

slaughter weight and sex on organoleptic properties, and fatty acid composition of lamb. Journal

of Animal Science, 51(2), 321-330.

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

consumer ranking of lamb colour and objective measures of colour. Meat Science, 85(2), 224-

229.

245

Kim, G.-D., Kim, B.-W., Jeong, J.-Y., Hur, S.-J., Cho, I.-C., Lim, H.-T., and Joo, S.-T. (2013).

Relationship of carcass weight to muscle fiber characteristics and pork quality of crossbred

(Korean native black pig× landrace) F2 pigs. Food and Bioprocess Technology, 6(2), 522-529.

Kim, G. D., Hur, S. J., and Jeon, J. T. (2010). The Relationship between Meat Color (CIE L and a),

Myoglobin Content, and Their Influence on Muscle Fiber Characteristics and Pork Quality.

Korean Journal for Food Science of Animal Resources, 30(4), 626-633.

King, D. A., Shackelford, S. D., Kuehn, L. A., Kemp, C. M., Rodriguez, A. B., Thallman, R. M., and

Wheeler, T. L. (2010). Contribution of genetic influences to animal-to-animal variation in

myoglobin content and beef lean color stability. Journal of Animal Science, 88(3), 1160-1167.

Kirchgessner, M., Heindl, U., and Schwarz, F. J. (1994). Gehalte und Ansatz von Spurenelementen in

Geweben und Ganzkörper von wachsenden Jungbullen der Rasse Fleckvieh. Journal of Animal

Physiology and Animal Nutrition, 72(1‐5), 260-271.

Kirkland, J. L., and Dobson, D. E. (1997). Preadipocyte function and aging: links between age-related

changes in cell dynamics and altered fat tissue function. Journal of the American Geriatrics

Society, 45(8), 959-967.

Kleemann, D. O., Dolling, C. H. S., and Ponzoni, R. W. (1984). Effect of breed of dam, type of birth and

sex of lamb on efficiency of conversion of food to lamb and wool in Merino, Poll Dorset x

Merino and Border Leicester x Merino ewes. Australian Journal of Agricultural Research,

35(4), 579-594.

Klont, R. E., Brocks, L., and Eikelenboom, G. (1998). Muscle fibre type and meat quality. Meat Science,

49, S219-S229.

Knap, P. W. (2000). Time trends of Gompertz growth parameters in'meat-type'pigs. Animal Science,

70(1), 39-49.

Kondo, H., Kimura, M., and Itokawa, Y. (1991). Manganese, Copper, Zinc, and Iron Concentrations and

Subcellular Distribution in Two Types of Skeletal Muscle. Experimental Biology and Medicine,

196(1), 83-88.

Kotula, A. W., and Lusby, W. R. (1982). Mineral composition of muscles of 1-to 6-year-old steers.

Journal of Animal Science, 54(3), 544-548.

Krausgrill, D. I., Tulloh, N. M., Shorthose, W. R., and Sharpe, K. (1999). Effects of weight loss in ewes

in early pregnancy on muscles and meat quality of lambs. The Journal of Agricultural Science,

132(01), 103-116.

LaFlamme, L. F., Trenkle, A., and Topel, D. G. (1973). Effect of castration or breed type on growth of

the longissimus muscle in male cattle. Growth, 37(3), 249.

Laird, A. K. (1965). Dynamics of relative growth. Growth, 29(3), 249-263.

Lambe, N. R., Navajas, E. A., Fisher, A. V., Simm, G., Roehe, R., and Bünger, L. (2009). Prediction of

lamb meat eating quality in two divergent breeds using various live animal and carcass

measurements. Meat Science, 83(3), 366-375.

Lambe, N. R., Navajas, E. A., Simm, G., and Bunger, L. (2006). A genetic investigation of various

growth models to describe growth of lambs of two contrasting breeds. Journal of Animal

Science, 84(10), 2642-2654.

Langlands, J. P. (1972). Growth and herbage consumption of grazing Merino and Border Leicester lambs

reared by their mothers or fostered by ewes of the other breed. Animal Science, 14(03), 317-322.

Larzul, C., Monin, G., Sellier, P., Ecolan, P., and Lefaucheur, L. (1999). Selection for reduced muscle

glycolytic potential in Large White pigs. II. Correlated responses in meat quality and muscle

compositional traits. Genetics selection evolution (Paris), 31(1), 61-76.

246

Lawrence, T. L. J., and Fowler, V. R. (2002). Growth of Farm Animals. Oxon, England: CABI

Publishing, pp. 347.

Lawrie, R. A. (1983). Aspects of the biochemistry of meat. International Journal of Biochemistry, 15(3),

233-242.

Ledward, D. A. (1992). Haemoproteins in Meat and Meat Products. In B. J. F. Hudson (Ed.),

Biochemistry of Food Proteins (pp. 197-234): Springer US.

Ledward, D. A., and Shorthose, W. R. (1971). A note on the haem pigment concentration of lamb as

influenced by age and sex. Animal Science, 13(01), 193-195.

Lee, G. J., Harris, D. C., Ferguson, B. D., and Jelbart, R. A. (1990). Growth and carcass fatness of ewe,

wether, ram and cryptorchid crossbred lambs reared at pasture: effects of weaning age.

Australian Journal of Experimental Agriculture, 30(6), 743-747.

Lee, S. H., Joo, S. T., and Ryu, Y. C. (2010). Skeletal muscle fiber type and myofibrillar proteins in

relation to meat quality. Meat Science, 86(1), 166-170.

Lefaucheur, L. (2010). A second look into fibre typing–Relation to meat quality. Meat Science, 84(2),

257-270.

Lefaucheur, L., Hoffman, R. K., Gerrard, D. E., Okamura, C. S., Rubinstein, N., and Kelly, A. (1998).

Evidence for three adult fast myosin heavy chain isoforms in type II skeletal muscle fibers in

pigs. Journal of Animal Science, 76(6), 1584-1593.

Lewis, R. M., and Brotherstone, S. (2002). A genetic evaluation of growth in sheep using random

regression techniques. Animal Science, 74, 63-70.

Lewis, R. M., Emmans, G. C., Dingwall, W. S., and Simm, G. (2002). A description of the growth of

sheep and its genetic analysis. Animal Science, 74, 51-62.

Lewis, R. M., Macfarlane, J. M., Simm, G., and Emmans, G. C. (2004). Effects of food quality on growth

and carcass composition in lambs of two breeds and their cross.

Lin, K. C., Cross, H. R., Johnson, H. K., Breidenstein, B. C., Randecker, V., and Field, R. A. (1988).

Mineral composition of lamb carcasses from the United States and New Zealand. Meat Science,

24(1), 47-59.

Linder, M. C. (1991). Nutritional biochemistry and metabolism: with clinical applications.

Livingston, D. J., and Brown, W. D. (1981). The chemistry of myoglobin and its reactions [Meat

pigments, food quality indices]. Food Technology (USA).

London, J. C., and Weniger, J. H. (1994). Investigations into traditionally managed Djallonké‐sheep

production in the humid and subhumid zones of Asante, Ghana. Journal of Animal Breeding and

Genetics, 111(1‐6), 314-336.

London, J. C., and Weniger, J. H. (1995). Investigations into traditionally managed Djallonké‐sheep

production in the humid and subhumid zones of Asante, Ghana III. Relationship between birth

weight, preweaning growth, and postweaning growth of lambs. Journal of Animal Breeding and

Genetics, 112(1‐6), 431-453.

López de Torre, G., Candotti, J. J., Reverter, A., Bellido, M. M., Vasco, P., García, L. J., and Brinks, J. S.

(1992). Effects of growth curve parameters on cow efficiency. Journal of Animal Science, 70(9),

2668-2672.

Lucero-Borja, J., Pouzo, L. B., de la Torre, M. S., Langman, L., Carduza, F., Corva, P. M., Santini, F. J.,

and Pavan, E. (2014). Slaughter weight, sex and age effects on beef shear force and tenderness.

Livestock Science, 163(0), 140-149.

247

Luno, M., Roncalés, P., Djenane, D., and Beltrán, J. A. (2000). Beef shelf life in low O2 and high CO2

atmospheres containing different low CO concentrations. Meat Science, 55(4), 413-419.

Lyons, G. E., Ontell, M., Cox, R., Sassoon, D., and Buckingham, M. (1990). The expression of myosin

genes in developing skeletal muscle in the mouse embryo. The Journal of cell biology, 111(4),

1465-1476.

Maier, A., McEwan, J. C., Dodds, K. G., Fischman, D. A., Fitzsimons, R. B., and Harris, A. J. (1992).

Myosin heavy chain composition of single fibres and their origins and distribution in developing

fascicles of sheep tibialis cranialis muscles. J Muscle Res Cell Motil, 13(5), 551-572.

Malhado, C. H. M., Carneiro, P. L. S., Affonso, P. R. A. M., Souza Jr, A. A. O., and Sarmento, J. L. R.

(2009). Growth curves in Dorper sheep crossed with the local Brazilian breeds, Morada Nova,

Rabo Largo, and Santa Inês. Small Ruminant Research, 84(1-3), 16-21.

Maltin, C. A., Sinclair, K. D., Warriss, P. D., Grant, C. M., Porter, A. D., Delday, M. I., and Warkup, C.

C. (1998). The effects of age at slaughter, genotype and finishing system on the biochemical

properties, muscle fibre type characteristics and eating quality of bull beef from suckled calves.

Animal Science, 66(02), 341-348.

Mancini, R. A., and Hunt, M. (2005). Current research in meat color. Meat Science, 71(1), 100-121.

McKenna, D. R., Mies, P. D., Baird, B. E., Pfeiffer, K. D., Ellebracht, J. W., and Savell, J. W. (2005).

Biochemical and physical factors affecting discoloration characteristics of 19 bovine muscles.

Meat Science, 70(4), 665-682.

McManus, C., Evangelista, C., Fernandes, L. A. C., de Miranda, R. M., Moreno-Bernal, F. E., and dos

Santos, N. R. (2003). Curvas de crescimento de ovinos Bergamácia criados no Distrito Federal.

Revista Brasileira de Zootecnia, 32(5), 1207-1212.

McPhee, M. J., Harden, S., Robinson, D. L., Dicker, R. W., and Oddy, V. H. (2012). Effects of

backgrounding and finishing growth rates on P8 fat and intramuscular fat in Bos taurus steers.

Animal Production Science, 52(5), 354-364.

McPhee, M. J., Hopkins, D. L., and Pethick, D. W. (2008). Intramuscular fat levels in sheep muscle

during growth. Australian Journal of Experimental Agriculture, 48(7), 904-909.

Meyer, K. (1999). Estimates of genetic and phenotypic covariance functions for postweaning growth and

mature weight of beef cows. Journal of Animal Breeding and Genetics, 116(3), 181-205.

Meyer, K. (2000). Random regressions to model phenotypic variation in monthly weights of Australian

beef cows. Livestock Production Science, 65(1-2), 19-38.

Meyer, K. (2002). Estimates of covariance functions for growth of Australian beef cattle from a large set

of field data. In Proceedings of the 7th World Congress on Genetics Applied to Livestock

Production (Vol. 100, pp. 0-4). Montpellier, France.

Miller, L. R., Garwood, V. A., and Judge, M. D. (1975). Factors Affecting Porcine Muscle Fiber Type,

Diameter and Number. Journal of Animal Science, 41(1), 66-77.

MLA. (2013). Australia's Sheepmeet Industry. In Fast facts 2013. In. Meat & Livestock Australia.

Sydney.

Moloney, A. P., Keane, M. G., Mooney, M. T., and Troy, D. J. (2000). Fat deposition and muscle

tenderness in steers with different patterns of concentrate consumption. Irish Journal of

Agricultural and Food Research, 39.

Moody, W. G., Kemp, J. D., Mahyuddin, M., Johnston, D. M., and Ely, D. G. (1980). Effect of Feeding

Systems, Slaughter Weight and Sex on Histological Properties of Lamb Carcasses. Journal of

Animal Science, 50(2), 249-256.

248

Moore, A. J. (1985). A mathematical equation for animal growth from embryo to adult. Animal Science,

40(03), 441-453.

Morgan, J. E., Fogarty, N. M., Nielsen, S., and Gilmour, A. R. (2006). Milk yield and milk composition

from grazing primiparous non-dairy crossbred ewes. Australian Journal of Agricultural

Research, 57(4), 377-387.

Mortimer, S. I., van der Werf, J. H., Jacob, R. H., Hopkins, D. L., Pannier, L., Pearce, K. L., Gardner, G.

E., Warner, R. D., Geesink, G. H., Edwards, J. E., Ponnampalam, E. N., Ball, A. J., Gilmour, A.

R., and Pethick, D. W. (2014). Genetic parameters for meat quality traits of Australian lamb

meat. Meat Science, 96(2 Pt B), 1016-1024.

Mortimer, S. I., van der Werf, J. H. J., Jacob, R. H., Pethick, D. W., Pearce, K. L., Warner, R. D.,

Geesink, G. H., Hocking Edwards, J. E., Gardner, G. E., Ponnampalam, E. N., Kitessa, S. M.,

Ball, A. J., and Hopkins, D. L. (2010). Preliminary estimates of genetic parameters for carcass

and meat quality traits in Australian sheep. Animal Production Science, 50(12), 1135-1144.

MSA. (2012). Meat Standards Australia Sheepmeat. In. Fortitude Valley, QLD.

Mulsant, P., Lecerf, F., Fabre, S., Schibler, L., Monget, P., Lanneluc, I., Pisselet, C., Riquet, J.,

Monniaux, D., Callebaut, I., Cribiu, E., Thimonier, J., Teyssier, J., Bodin, L., Cognié, Y.,

Chitour, N., and Elsen, J.-M. (2001). Mutation in bone morphogenetic protein receptor-IB is

associated with increased ovulation rate in Booroola Mérino ewes. Proceedings of the National

Academy of Sciences, 98(9), 5104-5109.

Murphy, T. A., Loerch, S. C., McClure, K. E., and Solomon, M. B. (1994). Effects of restricted feeding

on growth performance and carcass composition of lambs. Journal of Animal Science, 72(12),

3131-3137.

Murray, A., Pommier, S., Gibson, L., Robertson, W., and Johns, M. (2004a). Influence of intramuscular

fat (IMF) on pork quality. Advances in Pork Production, 15, 16.

Murray, R., Granner, D., Mayes, P., and Rodwell, V. (1996). Harpers biochemistry (24th ed.). New

Jersey: Prentice Hall International, Inc.

Murray, R. K., Granner, D. K., Mayes, P. A., and Rodwell, V. W. (2004b). Harper's Biochemistry. 2000

(Vol. 611). New York: McGrow-Hill.

Nelder, J. A. (1961). The fitting of a generalization of the logistic curve. Biometrics, 17(1), 89-110.

Nelson, D., and Cox, M. (2005). Lehninger principles of biochemistry (4th ed.). New York: WH Freeman

and Company.

Nissen, P. M., Danielsen, V. O., Jorgensen, P. F., and Oksbjerg, N. (2003). Increased maternal nutrition

of sows has no beneficial effects on muscle fiber number or postnatal growth and has no impact

on the meat quality of the offspring. Journal of Animal Science, 81(12), 3018-3027.

O'keeffe, M., and Hood, D. E. (1982). Biochemical factors influencing metmyoglobin formation on beef

from muscles of differing colour stability. Meat Science, 7(3), 209-228.

Ockerman, H. W., Jaworek, D., VanStavern, B., Parrett, N., and Pierson, C. J. (1984). Castration and sire

effects on carcass traits, meat palatability and muscle fiber characteristics in Angus cattle.

Journal of Animal Science, 59(4), 981-990.

Okeudo, N. J., and 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(2), 195-200.

Oldham, C. M., Thompson, A. N., Ferguson, M. B., Gordon, D. J., Kearney, G. A., and Paganoni, B. L.

(2011). The birthweight and survival of Merino lambs can be predicted from the profile of

liveweight change of their mothers during pregnancy. Animal Production Science, 51(9), 776-

783.

249

Ono, K., Berry, B. W., Johnson, H. K., Russek, E., Parker, C. F., Cahill, V. R., and Althouse, P. G.

(1984). Nutrient composition of lamb of two age groups. Journal of food science, 49(5), 1233-

1239.

Ozawa, S., Mitsuhashi, T., Mitsumoto, M., Matsumoto, S., Itoh, N., Itagaki, K., Kohno, Y., and Dohgo,

T. (2000). The characteristics of muscle fiber types of longissimus thoracis muscle and their

influences on the quantity and quality of meat from Japanese Black steers. Meat Science, 54(1),

65-70.

Pannier, L., Gardner, G. E., Pearce, K. L., McDonagh, M., Ball, A. J., Jacob, R. H., and Pethick, D. W.

(2014a). Associations of sire estimated breeding values and objective meat quality

measurements with sensory scores in Australian lamb. Meat Science, 96(2, Part B), 1076-1087.

Pannier, L., Pethick, D. W., Boyce, M. D., Ball, A. J., Jacob, R. H., and Gardner, G. E. (2014b).

Associations of genetic and non-genetic factors with concentrations of iron and zinc in the

longissimus muscle of lamb. Meat Science, 96(2), 1111-1119.

Pannier, L., Pethick, D. W., Geesink, G. H., Ball, A. J., Jacob, R. H., and Gardner, G. E. (2014c).

Intramuscular fat in the longissimus muscle is reduced in lambs from sires selected for leanness.

Meat Science, 96(2 Pt B), 1068-1075.

Pannier, L., Ponnampalam, E. N., Gardner, G. E., Hopkins, D. L., Ball, A. J., Jacob, R. H., Pearce, K. L.,

and Pethick, D. W. (2010). Prime Australian lamb supplies key nutrients for human health.

Animal Production Science, 50(12), 1115-1122.

Parks, J. R. (1982). Theory of feeding and growth of animals.

Pascual, M., and Pla, M. (2007). Changes in carcass composition and meat quality when selecting rabbits

for growth rate. Meat Science, 77(4), 474-481.

Pearce, K. L., Pannier, L., Williams, A., Gardner, G. E., Ball, A., and Pethick, D. W. (2009). Factors

affecting the iron and zinc content of lamb. Recent advances in animal nutrition–Australia, 17,

37-42.

Pearce, K. L., van de Ven, R., Mudford, C., Warner, R. D., Hocking-Edwards, J., Jacob, R., Pethick, D.

W., and Hopkins, D. L. (2010). Case studies demonstrating the benefits on pH and temperature

decline of optimising medium-voltage electrical stimulation of lamb carcasses. Animal

Production Science, 50(12), 1107-1114.

Perry, D., Shorthose, W. R., Ferguson, D. M., and Thompson, J. M. (2001). Methods used in the CRC

program for the determination of carcass yield and beef quality. Australian Journal of

Experimental Agriculture, 41(7), 953-957.

Perry, D., and Thompson, J. M. (2005). The effect of growth rate during backgrounding and finishing on

meat quality traits in beef cattle. Meat Science, 69(4), 691-702.

Peter, J. B., Barnard, R. J., Edgerton, V. R., Gillespie, C. A., and Stempel, K. E. (1972). Metabolic

profiles of three fiber types of skeletal muscle in guinea pigs and rabbits. Biochemistry, 11(14),

2627-2633.

Pethick, D. W., Banks, R. G., Hales, J., and Ross, I. R. (2006a). Australian prime lamb–a vision for 2020.

International Journal of Sheep and Wool Science, 54(1).

Pethick, D. W., Davidson, R., Hopkins, D. L., Jacob, R. H., D'Souza, D. N., Thompson, J. M., and

Walker, P. J. (2005a). The effect of dietary treatment on meat quality and on consumer

perception of sheep meat eating quality. Animal Production Science, 45(5), 517-524.

Pethick, D. W., Fergusson, D. M., Gardner, G. E., Hocquette, J. F., Thompson, J. M., and Warner, R.

(2005b). Muscle metabolism in relation to genotypic and environmental influences on consumer

defined quality of red meat. Indicators of milk and beef quality EAAP Publication, 112, 95-110.

250

Pethick, D. W., Harper, G. S., and Oddy, V. H. (2004). Growth, development and nutritional

manipulation of marbling in cattle: a review. Australian Journal of Experimental Agriculture,

44(7), 705-715.

Pethick, D. W., Hopkins, D. L., D’Souza, D. N., Thompson, J. M., and Walker, P. J. (2005c). Effects of

animal age on the eating quality of sheep meat. Australian Journal of Experimental Agriculture,

45(5), 491-498.

Pethick, D. W., Pleasants, A. B., Gee, A. M., Hopkins, D. L., and Ross, I. R. (2006b). Eating quality of

commercial meat cuts from Australian lambs and sheep (Vol. 66): New Zealand Society of

Animal Production.

Picard, B., and Cassar-Malek, I. (2009). Evidence for expression of IIb myosin heavy chain isoform in

some skeletal muscles of Blonde d’Aquitaine bulls. Meat Science, 82(1), 30-36.

Picard, B., Jurie, C., Duris, M. P., and Renand, G. (2006). Consequences of selection for higher growth

rate on muscle fibre development in cattle. Livestock Science, 102(1–2), 107-120.

Picard, B., Lefaucheur, L., Berri, C., and Duclos, M. J. (2002). Muscle fibre ontogenesis in farm animal

species. Reproduction Nutrition Development, 42(5), 415-432.

Picard, B., Robelin, J., Pons, F., and Geay, Y. (1994). Comparison of the foetal development of fibre

types in four bovine muscles. Journal of Muscle Research & Cell Motility, 15(4), 473-486.

Polkinghorne, R., Watson, R., Porter, M., Gee, A., Scott, J., and Thompson, J. (1999). Meat standards

Australia, APACCP'based beef grading scheme for consumers. 1) The use of consumer scores to

set grade standards. In International congress of meat science and technology (Vol. 45, pp. 14-

15).

Ponnampalam, E. N., Butler, K. L., Jacob, R. H., Pethick, D. W., Ball, A. J., Hocking Edwards, J. E.,

Geesink, G., and Hopkins, D. L. (2014a). Health beneficial long chain omega-3 fatty acid levels

in Australian lamb managed under extensive finishing systems. Meat Science, 96(2), 1104-1110.

Ponnampalam, E. N., Butler, K. L., Jacob, R. H., Pethick, D. W., Ball, A. J., Hocking Edwards, J. E.,

Geesink, G., and Hopkins, D. L. (2014b). Health beneficial long chain omega-3 fatty acid levels

in Australian lamb managed under extensive finishing systems. Meat Science, 96(2, Part B),

1104-1110.

Ponnampalam, E. N., Hopkins, D. L., Dunshea, F. R., Pethick, D. W., Butler, K. L., and Warner, R. D.

(2007). Genotype and age effects on sheep meat production. 4. Carcass composition predicted by

dual energy X-ray absorptiometry. Animal Production Science, 47(10), 1172-1179.

Portolano, B., and Todaro, M. (1997). Curves and biological efficiency of growth for 100-and 180-day-

old lambs of different genetic types. In Annales de Zootechnie (Vol. 46, pp. 245-253): Elsevier

Science.

Powell, S. R. (2000). The antioxidant properties of zinc. The Journal of nutrition, 130(5), 1447S-1454S.

Rahelic, S., and Puac, S. (1981). Fibre types in Longissimus dorsi from wild and highly selected pig

breeds. Meat Science, 5(6), 439-450.

Rehfeldt, C., Tuchscherer, A., Hartung, M., and Kuhn, G. (2008). A second look at the influence of birth

weight on carcass and meat quality in pigs. Meat Science, 78(3), 170-175.

Renand, G., Picard, B., Touraille, C., Berge, P., and Lepetit, J. (2001). Relationships between muscle

characteristics and meat quality traits of young Charolais bulls. Meat Science, 59(1), 49-60.

Renerre, M. (1990). Factors involved in the discoloration of beef meat. International Journal of Food

Science & Technology, 25(6), 613-630.

Renerre, M., and Labas, R. (1987). Biochemical factors influencing metmyoglobin formation in beef

muscles. Meat Science, 19(2), 151-165.

251

Resconi, V. C., Campo, M. M., Furnols, M., Montossi, F., and Sañudo, C. (2009). Sensory evaluation of

castrated lambs finished on different proportions of pasture and concentrate feeding systems.

Meat Science, 83(1), 31-37.

Richards, F. J. (1959). A flexible growth function for empirical use. Journal of experimental Botany,

10(2), 290-301.

Rickansrud, D. A., and Henrickson, R. L. (1967). Total Pigments and Myoglobin Concentration In Four

Bovine Muscles. Journal of food science, 32(1), 57-61.

Robelin, J., Lacourt, A., Bechet, D., Ferrara, M., Briand, Y., and Geay, Y. (1990). Muscle differentiation

in the bovine fetus: a histological and histochemical approach. Growth, development, and aging:

GDA, 55(3), 151-160.

Robelin, J., Picard, B., Listrat, A., Jurie, C., Barboiron, C., Pons, F., and Geay, Y. (1993). Myosin

expression in semitendinosus muscle during fetal development of cattle: immunocytochemical

and electrophoretic analyses. Reproduction Nutrition Development, 33(1), 25-41.

Rowe, R. W., and Goldspink, G. (1969). Muscle fibre growth in five different muscles in both sexes of

mice. Journal of anatomy, 104(Pt 3), 519.

Ruusunen, M., and Puolanne, E. (2004). Histochemical properties of fibre types in muscles of wild and

domestic pigs and the effect of growth rate on muscle fibre properties. Meat Science, 67(3), 533-

539.

Ryu, Y. C., Choi, Y. M., Lee, S. H., Shin, H. G., Choe, J. H., Kim, J. M., Hong, K. C., and Kim, B. C.

(2008). Comparing the histochemical characteristics and meat quality traits of different pig

breeds. Meat Science, 80(2), 363-369.

Ryu, Y. C., and Kim, B. C. (2005). The relationship between muscle fiber characteristics, postmortem

metabolic rate, and meat quality of pig longissimus dorsi muscle. Meat Science, 71(2), 351-357.

Safari, A., and Fogarty, N. M. (2003). Genetic parameters for sheep production traits. In Estimates from

the literature. (Vol. 2800, pp. 13-21). Orange, NSW: NSW Agriculture, Orange Agricultural

Institute.

Safari, E., Fogarty, N. M., Ferrier, G. R., Hopkins, L. D., and Gilmour, A. (2001). Diverse lamb

genotypes. 3. Eating quality and the relationship between its objective measurement and sensory

assessment. Meat Science, 57(2), 153-159.

Sammel, L. M., Hunt, M. C., Kropf, D. H., Hachmeister, K. A., and Johnson, D. E. (2002). Comparison

of assays for metmyoglobin reducing ability in beef inside and outside semimembranosus

muscle. Journal of food science, 67(3), 978-984.

Sanders, S. K., Morgan, J. B., Wulf, D. M., Tatum, J. D., Williams, S. N., and Smith, G. C. (1997).

Vitamin E supplementation of cattle and shelf-life of beef for the Japanese market. Journal of

Animal Science, 75(10), 2634-2640.

Sañudo, C., Santolaria, M. P., María, G., Osorio, M., and Sierra, I. (1996). Influence of carcass weight on

instrumental and sensory lamb meat quality in intensive production systems. Meat Science,

42(2), 195-202.

Savell, J. W., and Cross, H. R. (1988). The role of fat in the palatability of beef, pork, and lamb. In

Designing foods: Animal product options in the marketplace/Committee on technological

options to improve the nutritional attributes of animal products, Board of Agriculture, National

Research Council (pp. 345).

Scaramuzzi, R. J., and Radford, H. M. (1983). Factors regulating ovulation rate in the ewe. Journal of

Reproduction and Fertility, 69(1), 353-367.

252

Schiaffino, S., Gorza, L., Sartore, S., Saggin, L., Ausoni, S., Vianello, M., Gundersen, K., and LØmo, T.

(1989). Three myosin heavy chain isoforms in type 2 skeletal muscle fibres. J Muscle Res Cell

Motil, 10(3), 197-205.

Schiaffino, S., and Reggiani, C. (1996). Molecular diversity of myofibrillar proteins: gene regulation and

functional significance. Physiological reviews, 76(2), 371-423.

Schinckel, P. G., and Short, B. F. (1961). Influence of nutritional level during pre-natal and early post-

natal life on adult fleece and body characters. Australian Journal of Agricultural Research,

12(1), 176-202.

Schreurs, N. M., Garcia, F., Jurie, C., Agabriel, J., Micol, D., Bauchart, D., Listrat, A., and Picard, B.

(2008). Meta-analysis of the effect of animal maturity on muscle characteristics in different

muscles, breeds, and sexes of cattle. Journal of Animal Science, 86(11), 2872-2887.

Scott, N. (2014). On-Farm use of sheep electronic identification. In Lambex (pp. 41). Adelaide, SA.

Shackelford, S. D., Koohmaraie, M., Cundiff, L. V., Gregory, K. E., Rohrer, G. A., and Savell, J. W.

(1994). Heritabilities and phenotypic and genetic correlations for bovine postrigor calpastatin

activity, intramuscular fat content, Warner-Bratzler shear force, retail product yield, and growth

rate. Journal of Animal Science, 72(4), 857-863.

Sharaby, M. A., and Suleiman, I. O. (1988). Slaughter weights, carcass weights and dressing percentages

of five breeds of sheep slaughtered in Saudi Arabia. Australian Journal of Experimental

Agriculture, 28(5), 567-570.

Shorthose, W. R., and Harris, P. V. (1990). Effect of Animal Age on the Tenderness of Selected Beef

Muscles. Journal of food science, 55(1), 1-8.

Smerdu, V., Karsch-Mizrachi, I., Campione, M., Leinwand, L., and Schiaffino, S. (1994). Type 11x

myosin heavy chain transcripts are expressed in type IIb fibers of human skeletal muscle.

American Journal of Physiology, 267, C1723-C1728.

Smith, G. C., Belk, K. E., Sofos, J. N., Tatum, J. D., and Williams, S. N. (2000). Economic impiications

of improved color stability in beef. In Antioxidants in muscle foods: Nutritional strategies to

improve quality. Wiley, New York, NY (pp. 397-426).

Smith, G. M. (1977). Factors Affecting Birth Weight, Dystocia and Preweaning Survival in Sheep.

Journal of Animal Science, 44(5), 745-753.

Snowder, G. D., and Glimp, H. A. (1991). Influence of breed, number of suckling lambs, and stage of

lactation on ewe milk production and lamb growth under range conditions. Journal of Animal

Science, 69(3), 923-930.

Solomon, M. B., Campbell, R. G., and Steele, N. C. (1990). Effect of sex and exogenous porcine

somatotropin on longissimus muscle fiber characteristics of growing pigs. Journal of Animal

Science, 68(4), 1176-1181.

Solomon, M. B., and Lynch, P. G. (1988). Biochemical, Histochemical and Palatability Characteristics of

Young Ram Lambs as Affected by Diet and Electrical Stimulation. Journal of Animal Science,

66(8), 1955-1962.

Solomon, M. B., and West, R. L. (1985). Profile of fiber types in muscles from wild pigs native to the

United States. Meat Science, 13(4), 247-254.

Sørheim, O., Nissen, H., Aune, T., and Nesbakken, T. (2001). Use of carbon monoxide in retail meat

packaging. In Proceedings of the 54th reciprocal meat conference (pp. 47-51).

Souza, J. C., and Bianchini Sobrinho, E. (1994). Estimativa do peso de bovinos de corte, aos 24 meses, da

raça Nelore, usando curvas de crescimento. Revista Brasileira de Zootecnia, 23(01), 85-91.

253

Speck, P. A., Clarke, J. N., and Morris, C. A. (1997). Effects of strain and selection for growth on

tenderness in sheep and beef meats. In Proceedings of the Association for the Advancement of

Animal Breeding and Genetics 12, 567 (Vol. 570).

Spiess, A.-N., and Neumeyer, N. (2010). An evaluation of R2 as an inadequate measure for nonlinear

models in pharmacological and biochemical research: a Monte Carlo approach. BMC

Pharmacology, 10(1), 6.

Staron, R. S., and Johnson, P. (1993). Myosin polymorphism and differential expression in adult human

skeletal muscle. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry,

106(3), 463-475.

Staun, H. (1963). Various factors affecting number and size of muscle fibers in the pig. Acta Agriculturae

Scandinavica, 13(3), 293-322.

Stephens, M. A. (1974). EDF statistics for goodness of fit and some comparisons. Journal of the

American statistical Association, 69(347), 730-737.

Stockdale, F. E. (1992). Myogenic cell lineages. Dev Biol, 154(2), 284-298.

Strohecker, M. G., Faustman, C., Furr, H., Hoagland, T. A., and Williams, S. N. (1997). Vitamin E

supplementation effects on color and lipid stability of whole and ground lamb. Journal of Muscle

Foods, 8(4), 413-426.

Suman, S. P., Hunt, M. C., Nair, M. N., and Rentfrow, G. (2014). Improving beef color stability: Practical

strategies and underlying mechanisms. Meat Science, 98(3), 490-504.

Suman, S. P., and Joseph, P. (2013). Myoglobin chemistry and meat color. Annual review of food science

and technology, 4, 79-99.

Suzuki, A., and Cassens, R. G. (1983). A histochemical study of myofiber types in the serratus ventralis

thoracis muscle of sheep during growth. Journal of Animal Science, 56(6), 1447-1458.

Tanabe, R.-i., Muroya, S., and Chikuni, K. (1998). Sequencing of the 2a, 2x, and slow isoforms of the

bovine myosin heavy chain and the different expression among muscles. Mammalian genome,

9(12), 1056-1058.

Taylor, S. C. S. (1978). Methods of Quantifying Growth and Development: General Review. In H. De

Boer and J. Martin (Eds.), Patterns of Growth and Development in Cattle (Vol. 2, pp. 625-638):

Springer Netherlands.

Teixeira, A., Batista, S., Delfa, R., and Cadavez, V. (2005). Lamb meat quality of two breeds with

protected origin designation. Influence of breed, sex and live weight. Meat Science, 71(3), 530-

536.

Tejeda, J. F., Peña, R. E., and Andrés, A. I. (2008). Effect of live weight and sex on physico-chemical and

sensorial characteristics of Merino lamb meat. Meat Science, 80(4), 1061-1067.

Thatcher, L. P., Warren, B., and Nicholls, P. J. (1991). Effects of sex and year on growth and live

assessment of carcass characteristics of lambs grazing annual pastures. Australian Journal of

Experimental Agriculture, 31(3), 307-314.

Thompson, J. M. (2004). The effects of marbling on flavour and juiciness scores of cooked beef, after

adjusting to a constant tenderness. Animal Production Science, 44(7), 645-652.

Thompson, J. M., Butterfield, R. M., and Perry, D. (1985a). Food intake, growth and body composition in

Australian Merino sheep selected for high and low weaning weight. 2. Chemical and dissectible

body composition. Animal Science, 40(01), 71-84.

Thompson, J. M., Butterfield, R. M., and Perry, D. (1987). Food intake, growth and body composition in

Australian Merino sheep selected for high and low weaning weight 4. Partitioning of dissected

and chemical fat in the body. Animal Production, 45(01), 49-60.

254

Thompson, J. M., Gee, A., Hopkins, D. L., Pethick, D. W., Baud, S. R., and O’Halloran, W. J. (2005).

Development of a sensory protocol for testing palatability of sheep meats. Australian Journal of

Experimental Agriculture, 45(5), 469-476.

Thompson, J. M., and Parks, J. R. (1983). Food intake, growth and mature size in Australian Merino and

Dorset Horn sheep. Animal Science, 36(03), 471-479.

Thompson, J. M., Parks, J. R., and Perry, D. (1985b). Food intake, growth and body composition in

Australian Merino sheep selected for high and low weaning weight. 1. Food intake, food

efficiency and growth. Animal Science, 40(01), 55-70.

Thornton, R. F., Hood, R. L., Jones, P. N., and Re, V. M. (1979). Compensatory growth in sheep.

Australian Journal of Agricultural Research, 30(1), 135-151.

Topal, M., Ozdemir, M., Aksakal, V., Yildiz, N., and Dogru, U. (2004). Determination of the best

nonlinear function in order to estimate growth in Morkaraman and Awassi lambs. Small

Ruminant Research, 55(1-3), 229-232.

Trout, G. R. (1991). A rapid method for measuring pigment concentration in porcine and other low

pigmented muscles. In Proceedings of the 37th international congress of meat science and

technology (pp. 1198-1201).

Troy, D. J., and Kerry, J. P. (2010). Consumer perception and the role of science in the meat industry.

Meat Science, 86(1), 214-226.

Underwood, E. (2012). Trace Elements in Human and Animal Nutrition 4e: Elsevier.

USEPA, U. E. P. A. (1991). Methods for the determination of metals in environmental samples. In (Vol.

EPA/600/4-91-010).

Valin, C. (1988). Différenciation du tissu musculaire. Conséquences technologiques pour la filière viande.

Reproduction Nutrition Developpement, 28(3B), 845-856.

Valin, C., Touraille, C., Vigneron, P., and Ashmore, C. R. (1982). Prediction of lamb meat quality traits

based on muscle biopsy fibre typing. Meat Science, 6(4), 257-263.

van der Werf, J. H. J., Kinghorn, B. P., and Banks, R. G. (2010). Design and role of an information

nucleus in sheep breeding programs. Animal Production Science, 50(12), 998-1003.

van der Werf, J. H. J., and Schaeffer, L. R. (1997). Random regression in animal breeding. In. University

of New England.

van der Werf, J. H. J., and Wheaton, T. (1999). Estimation of genetic parameters for live weight, fat and

muscle measurements in seven sheep breeds based on LAMBPLAN data. In MLA Report (pp. 1-

21). Sydney, Australia.

van Laack, R. L., Stevens, S. G., and Stalder, K. J. (2001). The influence of ultimate pH and

intramuscular fat content on pork tenderness and tenderization. Journal of Animal Science,

79(2), 392-397.

Vergara, H., Molina, A., and Gallego, L. (1999). Influence of sex and slaughter weight on carcass and

meat quality in light and medium weight lambs produced in intensive systems. Meat Science,

52(2), 221-226.

Vivarelli, E., Brown, W. E., Whalen, R. G., and Cossu, G. (1988). The expression of slow myosin during

mammalian somitogenesis and limb bud differentiation. The Journal of cell biology, 107(6),

2191-2197.

von Bertalanffy, L. (1957). Quantitative laws in metabolism and growth. Quarterly Review of Biology,

217-231.

255

Wallace, W. J., Houtchens, R. A., Maxwell, J. C., and Caughey, W. S. (1982). Mechanism of

autooxidation for hemoglobins and myoglobins. Promotion of superoxide production by protons

and anions. Journal of Biological Chemistry, 257(9), 4966-4977.

Wang, Y. H., Bower, N. I., Reverter, A., Tan, S. H., De Jager, N., Wang, R., McWilliam, S. M., Cafe, L.

M., Greenwood, P. L., and Lehnert, S. A. (2009). Gene expression patterns during intramuscular

fat development in cattle. Journal of Animal Science, 87(1), 119-130.

Warner, R. D., Greenwood, P. L., Pethick, D. W., and Ferguson, D. M. (2010). Genetic and

environmental effects on meat quality. Meat Science, 86(1), 171-183.

Warner, R. D., Pethick, D. W., Greenwood, P. L., Ponnampalam, E. N., Banks, R. G., and Hopkins, D. L.

(2007). Unravelling the complex interactions between genetics, animal age and nutrition as they

impact on tissue deposition, muscle characteristics and quality of Australian sheep meat.

Australian Journal of Experimental Agriculture, 47(10), 1229-1238.

Watson, R., Gee, A., Polkinghorne, R., and Porter, M. (2008). Consumer assessment of eating quality –

development of protocols for Meat Standards Australia (MSA) testing. Australian Journal of

Experimental Agriculture, 48(11), 1360-1367.

Wegner, J., Albrecht, E., Fiedler, I., Teuscher, F., Papstein, H.-J., and Ender, K. (2000). Growth- and

breed-related changes of muscle fiber characteristics in cattle. Journal of Animal Science, 78(6),

1485-1496.

Weiler, U., Appell, H. J., Kremser, M., Hofäcker, S., and Claus, R. (1995). Consequences of selection on

muscle composition. A comparative study on gracilis muscle in wild and domestic pigs.

Anatomia, histologia, embryologia, 24(2), 77-80.

White, N. A., McGavin, M. D., and Smith, J. E. (1978). Age-related changes in percentage of fiber types

and mean fiber diameters of the ovine quadriceps muscles. American Journal of Veterinary

Research, 39(8), 1297-1302.

Wigmore, P. M., and Stickland, N. C. (1983). Muscle development in large and small pig fetuses. Journal

of anatomy, 137(Pt 2), 235-245.

Williams, S., Frye, T., Frigg, M., Schaefer, D. M., Scheller, K. K., and Liu, Q. (1992). Vitamin E. Meat

International, 3, 853.

Wilson, S. J., McEwan, J. C., Sheard, P. W., and Harris, A. J. (1992). Early stages of myogenesis in a

large mammal: formation of successive generations of myotubes in sheep tibialis cranialis

muscle. J Muscle Res Cell Motil, 13(5), 534-550.

Wittenberg, J. B., and Wittenberg, B. A. (2007). Myoglobin-enhanced oxygen delivery to isolated cardiac

mitochondria. Journal of Experimental Biology, 210(12), 2082-2090.

Wood, J. D., Enser, M., Fisher, A. V., Nute, G. R., Sheard, P. R., Richardson, R. I., Hughes, S. I., and

Whittington, F. M. (2008). Fat deposition, fatty acid composition and meat quality: A review.

Meat Science, 78(4), 343-358.

Yapi-Gnaorè, C. V., Oya, A., Rege, J. E. O., and Dagnogo, B. (1997). Analysis of an open nucleus

breeding programme for Djallonkè sheep in the Ivory Coast. 1. Examination of non-genetic

factors. Animal Science, 64(02), 291-300.

Young, O. A., and Bass, J. J. (1984). Effect of castration on bovine muscle composition. Meat Science,

11(2), 139-156.

Zygoyiannis, D., Kyriazakis, I., Stamataris, C., Friggens, N. C., and Katsaounis, N. (1997). The growth

and development of nine European sheep breeds. 2. Greek breeds: Boutsko, Serres and

Karagouniko. Animal Science, 65(03), 427-440.

256