. . 9.71.1
“1.2“ film. 9;:

1.3.
.2. 3 l...‘ :55...

.
. :9...
1.153.»...
.2”; 3.... .
Inlllo: Bo, . 25")
3.3;}...

J11:

, . .. Iii
.. It: .vt....ak. 8. Va...
:1. . 3 131 1‘3:

. . u . . a .

.5
a {lawn
.

.
9.. .4908

. .$wu.§.
. I’AA") ..
.3:

{F ..
‘ilgslll
‘OIOlthin’ub'. .
g I; 1 {<1
5,. i
1 21.3.? $1!
«.39... unnatural?» 1.. . .r ..
I: atlflw...‘ “thunk? .
1.3.2... .. 3.9:.
a .n 4.4.1.3.
RIG”
» 2.1:; in,
z. .5 ii; .1
n....!.....€~.:.£2... !
v9.Mv..y.$’: a.) ..av¢c4¢ot
:13... .32-
.2. .
. 1...?
‘3.

.1 70:11.9
grufim, “153 u
. 92"- IL.» 7.
“1.!!r... !
.‘D’

2,... i
. Hahn?
{Ban

1118? r.

I a
a.» 5:? q
A3...)

I
2.? 54:3. .
.1 1V!) 11:,51.
. :19: .!.lz2.ll.>'| .
$1.351? .L a,

.1: .1 .J- 9|

 

.30!
info...
v.5»... .
, :2-
nl r _ ,7 z .I
. .tvw .L.E.v,.:.§ rig . my“:
. . L ziLu. c5). L.
. a .
ctr. $531.3
.2: . . . .

 

.48.? Q... . . . ,
A . 1?, . .

. e, an, €533 Liamwfifl ...n...z..
y.....~....... . .. .

#223... +.....%.
3...); 21...;
iii...

This is to certify that the
dissertation entitled

ANALYSIS OF A DUROC x PIETRAIN F2 PIG RESOURCE
POPULATION FOR QUANTITATIVE TRAIT LOCI AFFECTING
GROWTH, BODY COMPOSITION, AND MEAT QUALITY TRAITS

presented by

David Bowen Edwards

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Animal Science

 

 

flaw Mae

Major Professors Sighatu re
/U a u e M [04 / 7 W

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

.. . - . . . . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-- o c - . -n-o-o-c-o-o-o------ - n u o n a . . s s n u . o a o u o o u o o o o o a o c - c . o c o o . - n o 0-. - . n

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

NOV 1 92003
@525? .08

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DateDue.indd-p.1

ANALYSIS OF A DUROC x PIETRAIN F2 PIG RESOURCE POPULATION FOR
QUANTITATIVE TRAIT LOCI AFFECTING GROWTH, BODY COMPOSITION,
AND MEAT QUALITY TRAITS
By

David Bowen Edwards

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Animal Science

2005

ABSTRACT
ANALYSIS OF A DUROC x PIETRAIN F2 PIG RESOURCE POPULATION FOR
QUANTITATIVE TRAIT LOCI AFFECTING GROWTH, BODY COMPOSITION,
AND MEAT QUALITY TRAITS
By
David Bowen Edwards
A Duroc x Pietrain F 2 pig resource population was created to discover quantitative
trait loci (QTL) affecting growth, body composition, and meat quality traits. These pigs
(1259 born) were finished in either a Modified Open Front (MOF) or a Test Station (TS)
building. Body weight and ultrasound estimates of tenth rib backfat, last rib backfat, and
longissimus muscle area were serially measured throughout development. Random
regression analyses were performed to evaluate body weight gain and its components
over time. Carcass and meat quality data collection included primal cut weights, backfat
thickness, muscle pH, objective and subjective color information, marbling and firmness
scores, and drip loss of boneless Iongissimus muscle chops. Additionally, chops were
analyzed for moisture, protein, and fat composition as well as cook yield and shear force
measurements. Palatability of chops was determined by a trained sensory taste panel.
Models that included genetic, permanent environment, and residual error variance
components were used to evaluate the influence of finisher facilities on these traits. Pigs
finished in the MOF were heavier at harvest and had more backfat at 22 wk of age and at
harvest at the tenth and last rib than pigs raised in the TS. Body weight random
regression analysis revealed that pigs reared in the TS grew more slowly at first, but then
grew more quickly later in the finisher phase for the same overall weight gain from 10 to

22 wk of age as pigs in the MOF. Pigs raised in the MOF had a greater backfat accretion

rate from 10 to 22 wk of age than pigs raised in the TS. Additionally, pigs raised in the
MOF had greater decline in pH from 45 min to 24 h postmortem and had lower Warner-
Bratzler shear force measurements than pigs raised in the TS. Thus, animals of similar
genetic merit can Show differences in phenotypes as influenced by finisher facilities.

A total of 510 F2 animals were genotyped for 124 microsatellite markers evenly
spaced across the entire genome. Data were analyzed with line cross least squares
regression interval mapping methods using sex and litter as fixed effects with covariates
of carcass weight or harvest age for specific carcass and meat quality traits. Significance
thresholds of the F -statistic for additive, dominance, and imprinted QTL were determined
on chromosome- and genome-wise levels by permutation tests.

A total of 54 QTL for 22 of the 29 measured growth traits, 33 QTL for 15 of the
16 animal random regression terms, and 94 QTL for 35 of the 38 carcass merit and meat
quality traits were found to be significant at the 5% chromosome-wise level. Growth and
body composition putative QTL were discovered for tenth and last rib backfat on SSC 6,
body composition traits on SSC 9, backfat and lipid composition traits on SSC 11, tenth
rib backfat and total body fat tissue on SSC 12, and linear regressions of body weight,
longissimus muscle area, and tenth rib backfat on SSC 18. Carcass merit and meat
quality putative QTL were discovered for 45 min pH and pH decline on SSC 3, marbling
score and carcass backfat on SSC 6, carcass length and number of ribs on SSC 7,
marbling score on SSC 12, and color measurements and tenderness score on SSC 15.
These results will facilitate fine mapping efforts to identify genes controlling grth and
body composition of pigs that can be incorporated into marker-assisted selection

programs to accelerate genetic improvement in pig populations.

This dissertation is dedicated to my family, who have fostered my learning and

stimulated my thinking from day one, and to Christy, who has made my life whole.

iv

ACKNOWLEDGMENTS

The opportunities I have been afforded in my graduate work are greatly
appreciated. Dr. Ron Bates has been the ideal mentor and has given me the support to
complete both my M.S. and Ph.D. He has provided me the opportunity to perform
cutting edge research, attend scientific meetings, and take full advantage of all the
opportunities I have desired to complete while in graduate school. Dr. Cathy Ernst has
provided a great sounding board for ideas about molecular genetics and gracefully served
as my second reader of my dissertation. Dr. Rob Tempelman, Dr. Matt Doumit, and Dr.
Guilherme Rosa have served on my Ph.D. guidance committee and given me ideas along
the way that have improved my education and research.

The creation of the Michigan State University Duroc x Pietrain F2 pig resource
population would not have occurred without the assistance of many units and persons.
Financial support has been provided by the Department of Animal Science when we
needed to start the project and keep it going. Many thanks to the Department for
believing in us and our project, and it will see the benefits and results for years to come.
Financial support also was provided by the Michigan Agricultural Experiment Station, a
Michigan Animal Initiative Coalition Grant, and USDA-CSREES NR] Award 2004-
35604-14580. Many people have been involved in the collection of data: Mark Hoge,
Nancy Raney, Emily Helman, Al Snedegar, Lance Kirkpatrick, the swine farm crew,
Tom Forton, Jennifer Dominguez, the MSU meats lab crew, Valencia Rilington, Luke
Bates, Kwan-Suk Kim, Lan Xiao, A’Lana Bates, Tobin Bates, and Marco Noventa.

Without these people, this project would not have been as successful as it has become.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES ............................................................................................................ x
LIST OF ABBREVIATIONS ........................................................................................... xii
INTRODUCTION .............................................................................................................. 1
CHAPTER 1. LITERATURE REVIEW ............................................................................. 4
Growth and Composition Prediction .............................................................................. 4
Growth Modeling ............................................................................................................ 6
Random Regression ........................................................................................................ 7
Meat Quality ................................................................................................................. 10
Impact of Management on Growth and Carcass Quality .............................................. 11
Duroc and Pietrain Breeds ............................................................................................ 12
Duroc vs. Pietrain Studies ............................................................................................. 14
Swine Resource Populations ......................................................................................... 16
Growth QTL .................................................................................................................. 20
Carcass Merit QTL ....................................................................................................... 22
Meat Quality QTL ......................................................................................................... 24
F2 Population QTL Analysis Procedures ...................................................................... 28
CHAPTER II. INFLUENCE OF FIN ISHER FACILITIES ON PIG GROWTH
PERFORMANCE ............................................................................................................. 31
Abstract ......................................................................................................................... 31
Introduction ................................................................................................................... 32
Materials and Methods .................................................................................................. 33
Population Development ........................................................................................... 33
Animal Management ................................................................................................. 34
Trait Collection ......................................................................................................... 35
Trait Analysis ............................................................................................................ 38
Serial Data Analysis .................................................................................................. 39
Results and Discussion ................................................................................................. 41
Off-Test, Carcass Composition, and Meat Quality ................................................... 41
Serial Data Results .................................................................................................... 43
Serial Heritability Results ......................................................................................... 44
Implications ................................................................................................................... 45

vi

CHAPTER III. QTL MAPPING IN AN F2 DUROC x PIETRAIN RESOURCE

POPULATION: 1. GROWTH TRAITS ........................................................................... 61
Abstract ......................................................................................................................... 61
Introduction ................................................................................................................... 62
Materials and Methods .................................................................................................. 63

Population Development ........................................................................................... 63
Animal Management ................................................................................................. 63
Phenotype and Genotype Collection ......................................................................... 64
Random Regression .................................................................................................. 66
QTL Analysis ............................................................................................................ 67
Results and Discussion ................................................................................................. 69
QTL Analysis ............................................................................................................ 69
Body Weight ............................................................................................................. 71
Backfat ...................................................................................................................... 72
Longissimus Muscle Area ......................................................................................... 73
Composition Traits .................................................................................................... 74
Confidence Intervals ................................................................................................. 76
Implications ................................................................................................................... 76

CHAPTER IV. QTL MAPPING IN AN F2 DUROC x PIETRAIN RESOURCE

POPULATION: II. CARCASS AND MEAT QUALITY TRAITS ................................. 91
Abstract ......................................................................................................................... 91
Introduction ................................................................................................................... 92
Materials and Methods .................................................................................................. 93

Population Development ........................................................................................... 93
Phenotype Collection ................................................................................................ 93
Trained Sensory Panel Evaluation ............................................................................ 95
Genotype Collection ................................................................................................. 96
QTL Analysis ............................................................................................................ 97
Results and Discussion ................................................................................................. 98
QTL Analysis ............................................................................................................ 98
Carcass Measurements ............................................................................................ 100
Primal Cut Weights ................................................................................................. 103
Meat Quality ........................................................................................................... 104
Confidence Intervals ............................................................................................... 107
Implications ................................................................................................................. 107
SUMMARY AND CONCLUSIONS ............................................................................. 121
LITERATURE CITED ................................................................................................... 127

vii

LIST OF TABLES

Table 1.1. Summary of F2 resource populations including original QTL manuscript for
each population, founder breeds, number of offspring, number of markers used, and types

of phenotypic data analyzed for QTL analysis. ................................................................ 17
Table 1.2. Summary of reported QTL chromosomal locations for growth traits. ............ 21
Table 1.3. Summary of reported QTL chromosomal locations for carcass traits. ............ 23

Table 1.4. Summary of reported QTL chromosomal locations for meat quality traits. 26

Table 11.1. Covariates for carcass and meat quality trait analyses. .................................. 47

Table 11.2. Order of polynomial on week of age terms and other significant terms for
random regression analyses of serial growth data. ........................................................... 48

Table 11.3. Number of observations, least squares means, standard error of the mean, and
P-value of the difference between finisher means for 22 wk of age off-test traits. .......... 49

Table 11.4. Number of observations, least squares means, standard error of the mean, and
P-value of the difference between finisher means for carcass and meat quality traits. 50

Table 111.1. Markers used in the QTL analysis, map positions determined for the F2
Duroc x Pietrain resource population, number of alleles segregating for each marker, and
number of missing genotypes for each marker. Distances (in Kosambi cM) are relative to
position of first marker on each chromosome in this population ...................................... 78

Table 111.2. Order of polynomial on week of age terms and other significant terms for
random regression analyses of serial growth data. ........................................................... 80

Table 111.3. Number of records, means, and standard deviations for growth traits
measured. .......................................................................................................................... 81

Table 111.4. Position and significance levels of single point QTL significant at 5%
chromosome-wise level with additive, dominance, and imprinting effects and standard
errors of the QTL. ............................................................................................................. 82

Table 111.5. Position and significance levels of random regression QTL significant at 5%
chromosome-wise level with additive effects and standard errors of QTL at those
positions. ........................................................................................................................... 83

Table 111.6. Position and 95% confidence interval lower and upper limits of growth QTL
significant at the 5% genome-wise level. ......................................................................... 84

viii

Table 1V.l. Fixed effects and covariates for carcass and meat quality trait QTL analyses.
......................................................................................................................................... 1 09

Table IV.2. Number of records, means, and standard deviations for carcass and meat
quality traits measured. ................................................................................................... 110

Table IV.3. Position and significance levels of carcass and meat quality QTL significant
at the 5% chromosome-wise level with additive, dominance, and imprinting effects and
standard errors. ................................................................................................................ 111

Table 1V.4. Position and 95% confidence interval lower and upper limits of carcass merit
QTL significant at 5% genome-wise level ...................................................................... 114

ix

LIST OF FIGURES

Figure 11.]. Body weight means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 51

Figure 11.2. Tenth rib backfat means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 52

Figure 11.3. Longissimus muscle area means with standard errors from 10 to 22 wk of
age for Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

........................................................................................................................................... 53
Figure 11.4. Last rib backfat means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 54
Figure 11.5. Fat-free total lean means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 55
Figure 11.6. Total body fat tissue means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 56
Figure 11.7. Empty body protein means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 57
Figure 11.8. Empty body lipid means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs. .......... 58
Figure 11.9. Heritability for measured traits from 10 to 22 wk of age. ............................ 59
Figure 11.10. Heritability for calculated traits from 10 to 22 wk of age. ......................... 60
Figure 111.]. F -ratio plots versus relative positions on SSC 16 ....................................... 85
Figure 111.2. F-ratio plots versus relative positions on SSC 6 ......................................... 86
Figure 111.3. F-ratio plots versus relative positions on SSC 11 ....................................... 87
Figure 111.4. F -ratio plots versus relative positions on SSC 18 ....................................... 88
Figure 111.5. F -ratio plots versus relative positions on SSC X ........................................ 89
Figure 111.6. F-ratio plots versus relative positions on SSC 9 ......................................... 90

Figure 1V.l. F-ratio plots versus relative positions on SSC 3. ...................................... 115

Figure IV.2. F-ratio plots versus relative positions on SSC 7. ...................................... 116
Figure IV.3. F-ratio plots versus relative positions on SSC 6 ....................................... 117
Figure 1V.4. F-ratio plots versus relative positions on SSC 15 ..................................... 118
Figure IV.5. F-ratio plots versus relative positions on SSC 12 ..................................... 119
Figure 1V.6. F-ratio plots versus relative positions on SSC 15 ..................................... 120

xi

LIST OF ABBREVIATIONS

ADG = average daily gain

BFlO = tenth rib backfat

BW = body weight

DNA = deoxyribonucleic acid
EBLIPID = empty body lipid
EBPRO = empty body protein

FF TOLN = fat-free total lean
LMA = longissimus muscle area
LRF = last rib backfat

MOF = Modified Open Front building
QTL = quantitative trait loci
RYR-l = ryanodine receptor gene
SSC = Sus scrofa chromosome
TOFAT = total body fat tissue

TS = Test Station building

WBS = Warner-Bratzler shear force

xii

INTRODUCTION

Genetic progress in livestock has been based upon direct selection for phenotype
or on predicted breeding values based upon phenotype. These breeding values are
predicated on accurate measurements of the traits and proper accounting for
environmental influences. Knowledge gained on the effects of certain genes can augment
selection programs and increase the rate of genetic progress. Molecular genetics enables
the identification of regions of the genome known as quantitative trait loci (QTL) that can
affect many economically important traits. These QTL can be incorporated into selection
programs and used to enhance genetic progress by allowing selection for traits in both
sexes, on all animals, earlier in life, and with less phenotypic data collection. Dekkers
(2003) identified four key steps of development for the eventual application of these
molecular genetics techniques in livestock breeding programs. The first area is the
development of molecular markers and genetic maps. Creation of resource populations
and analysis for QTL is the second major step. Thirdly, once these QTL are identified,
these genes or markers linked to the genes can be used in genetic evaluation, and, finally,
these data can be utilized in selection.

In order to attain these goals, researchers have developed anonymous markers that
are located across the entire genome in many species of interest. Dinucleotide
microsatellite markers have been developed for use in swine maps and QTL
identification. Combining these markers with phenotypic information from well
characterized swine populations has allowed the detection of QTL for various classes of
traits including growth, composition, meat quality, and reproduction. A valuable source

of phenotypic data is a resource population since the known pedigrees allow tracking of

alleles through the generations; these populations also can be used to study management
and other environmental effects on these phenotypes. Many previously reported swine
resource populations have used rustic breeds crossed with commercial breeds to study
QTL segregating between diverse populations (e. g. Andersson et al., 1994; Rohrer and
Keele, 1998a). While analysis of these populations was able to reveal QTL for important
phenotypes, a resource population generated from breeds already utilized in commercial
populations would allow more rapid introduction of beneficial QTL into the breeding
program.

Enhancement of production efficiency and improvement of product quality are
major concerns for producers of food animals. Unfortunately, selection for rapid lean
grth rate in swine frequently results in production of animals that yield inferior quality
meat. Locating and utilizing specific favorable genes for lean growth and meat quality
will help overcome this natural antagonistic relationship and allow improvement to be
realized for both efficient production and product quality. A major step towards this goal
can be achieved through the development and analysis of a genetic resource population
that exhibits variation for lean growth and meat quality traits. Duroc and Pietrain breeds,
used in many modern commercial populations and divergent for many growth,
composition, and meat quality traits (Kanis et al., 1990; Affentranger et al., 1996; Ellis et '
al., 1996; Garcia-Macias etal., 1996; Edwards et al., 2003; Edwards et al., 2006), were
selected as the basis for the resource population developed at Michigan State University.

Discovery of previously unreported QTL can lead to further understanding of the
influence of growth QTL on meat quality. Furthermore, traits such as meat quality are

difficult and expensive to measure and require slaughter of the animal. Discovering QTL

for meat quality traits and then implementing these QTL into breeding systems will allow
the pork industry to more rapidly improve production efficiency and profitability and at
the same time provide consumers a better quality product. Implementing marker-assisted
selection using information gained from QTL discovery can augment traditional selection
strategies.

A Duroc x Pietrain F2 pig resource population was developed at Michigan State
University. Analysis of the phenotypic and genotypic data from this population for the
discovery of QTL was the main emphasis of this dissertation. Additionally, an analysis
of the data to determine influence of management practices on phenotypes was
undertaken. Three objectives were evaluated in this research program:

1. To determine the influence of finisher facilities on pig growth, composition, and
meat quality traits.

2. To discover positions and effects of QTL affecting pig growth and body
composition traits.

3. To discover positions and effects of QTL affecting pig carcass merit and meat

quality traits.

CHAPTER 1
LITERATURE REVIEW
Growth and Composition Prediction

Breeds and lines used in commercial pork production each have their unique
patterns of growth. "Growth may be considered from at least two different aspects: 1) an
increase in body mass with time and 2) changes in form or composition resulting from
different growth rates of component parts," (Robison, 1976). Robison (1976) further
stated that efforts in the past to change the growth curve have been directed at simply
changing rate of body mass growth.

Change in body mass can be partitioned into its component parts. This
partitioning provides more information concerning biological differences between
different animals within populations as well as further description of different
populations. In order to study these compositional changes, animals have been dissected
or chemically analyzed at different slaughter weights (Siemens et al., 1989; Schinckel
and de Lange, 1996; Wagner et al., 1999). The high cost of these procedures makes them
prohibitive to be routinely conducted on a large number of pigs (Schinckel and de Lange,
1996). In addition, animals slaughtered would not be available for breeding purposes,
which allows only information on relatives to be used for predicting genetic merit.

Methods to predict these components of growth on live animals are desirable and
have been developed. One method is ultrasound technology. Terry et a1. (1989) used
ultrasound estimates of backfat and loin muscle area to attempt to predict the four lean
cuts as a percent of side weight. Targeting protein accretion specifically with attention

paid to an adequate amount of fat within pork is a production goal that can be achieved.

Several studies have reported non-invasive measures to predict body protein and fat
composition at various stages of the growth phase (McLaren etal., 1989; Siemens et al.,
1989; Houghton and Turlington, 1992; Schinckel and de Lange, 1996; NPPC, 2000).
The ability to estimate composition on breeding animals allows optimization of breeding
schemes and production practices for desired end products.

A review by Houghton and Turlington (1992) showed that ultrasound could
accurately assess body measurements of swine. Overall accuracy of composition
measurements (e. g., backfat at first rib, tenth rib, last rib, and last lumbar and loin muscle
depth, width, and area), as assessed by correlation coefficients, was approximately 0.9.
These high correlations to carcass measures allow for non-invasive measures of an
animal's phenotype.

Ultrasound technology has been used to characterize growth and composition
throughout the life of the pig. Serial real-time ultrasound measures have been taken on
pigs at regular intervals to predict body composition measures at market weight and to
model growth. McLaren et al. (1989) took ultrasound measures starting at 42 d of age
and every 2 wk thereafter until slaughter (mean of 98.5 kg, 170 d). Both measures of
backfat and loin muscle area were obtained and used to predict lean gain per day. They
reported that serial measures appeared promising, and cost-effective, as a technique for
monitoring composition of growth in the pig. Johnson et al. (2004) compared several
methods to estimate fat-free lean in swine including carcass measurements, ultrasound
measurements, and automated collection systems of composition estimates utilized in
packing plants. In order to compare these systems, physical and chemical dissection of

the carcass was undertaken. To calculate fat-free lean an adjustment has to be made for

the weight of dissected lean tissue to a fat-free basis by using the ratio of percentage of
lipid in lean tissue to percentage of lipid in fat tissue. An assumption held is that the lipid
percentages in dissected fat tissue and in the fat depots remaining within dissected lean
tissue are equal (Johnson et al., 2004). Another procedure to estimate lipid-free lean
from weight of dissected lean and the percentage of lipid in the tissue involves
calculation of the percentage of lipid-free lean from the weight of dissected lean and the
percentage of lipid in the tissue (Johnson et al., 2004). These two procedures produce
different estimates because, besides lipid, fat contains cytoplasmic fluids, protein, and ash
and the percentages of lipid in the dissected fat tissue and different fat depots within the
carcass vary (Higbie et al., 2002). Although these two measures can be substantially
different (Schinckel et al., 2001), the correlation between them is high (Schinckel et al.,
2003). This high correlation results in only minor differences in ranking of animals
between the two procedures (Schinckel et al., 2003) allowing either to be used in
selection with no detriment to rate or direction of improvement.
Growth Modeling

Successful simulation of growth performance of an individual pig depends not
only on the correct parameterization of its genotypic parameters but also on a detailed
description of its environment (Knap, 2000). Growth curves are an effective way to
summarize measurement information into only a few parameters (Mignon-Grasteau et al.,
1999). Many different approaches have been used to summarize growth parameters in
different species. These include changes in body mass as well as changes in composition
of body mass. Change in body mass has been an important growth characterization in

meat animal species. Whittemore et a1. (2003) used Landrace, Pietrain, and Meishan

crossbred pigs to estimate growth phenotypes between pigs with genetic predisposition to
different body types. Serial slaughter and dissection was carried out to estimate carcass
composition and describe these changes using regression over time. Estimation of
grth curves in chickens (Mignon-Grasteau et al., 1999) has shown that parameters of
the growth curve are heritable. Four lines of chickens were selected for change of body
weight at 8 or 36 weeks of age. One line was selected for high body weight at both ages.
Another was selected for low body weight at both ages. A third was selected for high
body weight at 8 weeks and low body weight at 36 weeks. The fourth was selected for
low body weight at 8 weeks and high body weight at 36 weeks. Offspring from these
selected animals were measured for body weight growth and had growth patterns similar
to their parents. Thus, patterns of growth as described by parameters of the functions, not
just overall growth, are heritable (Mignon-Grasteau et al., 1999).

A further refinement of growth is to measure the change in mass of its
components. In meat animals the relationship of lean growth (protein) to fat is especially
important. Knap (2000) fitted body protein and lipid mass growth functions and assumed
that the rate parameter would be the same for both components. Knap concluded that
selection between 'meat-type' pig populations has greatly reduced mature body lipid mass
while leaving mature body protein mass practically unchanged. Also, the growth rate of
both body fractions had substantially increased and the peak of the protein accretion
curve had shifted towards more mature stages of development.

Random Regression
Random regression analysis is a method to fit longitudinal data, such as body

weight, where the trait of interest is changing, gradually but continually, over time.

Random regression allows regression of a trait on age to model growth and allows
separation of between and within animal variation. Henderson Jr. (1982) first considered
random regression coefficients in a linear mixed model context. Recent applications
include evaluation of dairy cattle using test day records (J amrozik and Schaeffer, 1997;
Jamrozik et al., 1997), and description of grth curves in beef cattle (Varona et al.,
1997; Meyer, 2004), lambs (Fischer et al., 2004), and pigs (Andersen and Pedersen,
1996; Huisman et al., 2002, Edwards et al., 2006).

Random regression, as opposed to multiple trait analysis, can be used as an
effective tool to reduce the number of traits to analyze, such as body weight or body
component weights, for those traits measured over time. Traits can be described by
parameters of a function, instead of by many measures taken over a time period. The
fixed regression parameter of the function can model population trajectory, while the
random regression coefficients represent individuals' deviation from this curve (Meyer,
1998). Random coefficients allow a (co)variance structure to be specified for related
factors (Fischer et al., 2004). Additionally, Fischer et al. (2004) also fit a heterogeneous
residual error variance over age, which is common amongst growth data. Inclusion of
random genetic components of regressions in animal models may be applicable to any
situation in which repeated time-dependent observations are taken on a given animal, and
when the time-dependent response function exhibits genetic variation (Schaefi'er and
Dekkers, 1994).

Andersen and Pedersen (1996) applied random regression methodology to grth
rate and daily food intake in pigs. Gilts and barrows were on test from 30 to 115 kg live

weight. They postulated that the growth curve of each individual followed a fourth

degree polynomial. The (co)variance parameters were estimated using REML, and then
fixed effects were estimated and random effects were predicted assuming the estimated
(co)variance parameters as the true parameters. The difficulty in introducing random
effects or variance components in non-linear functions was the reason that linear
polynomial models were chosen. When animals are not taken to mature weights, such as
growth evaluation to standard or contemporary market weights, it is not necessary to
consider curves which approach asymptotic values (Andersen and Pedersen, 1996). This
is an advantage of the random regression approach over non-linear models, in which an
asymptotic value is estimated by extrapolating from collected data. Huisman et a1.
(2002) also utilized random regression to model body weight in pigs. They used a sire
model and accounted for repeated measures on each animal, but did not allow the random
residual variance to vary over time. This restriction may have led to incorrect estimation
of some of their parameters. Overall, Huisman et a1. (2002) concluded that random
regression models had better log likelihood values than spline or multivariate models and
were valuable to model body weight growth data. Fischer et a1. (2004) did fit a
heterogeneous measurement error variance structure to growth data in lambs and
investigated the opportunity to select on a phenotype derived from random regression
analyses. They also mentioned the flexibility inherent to random regression to account
for differences in measurement dates across experiments or management structures.
Meyer (2004) applied random regression to beef cattle and reported that accuracy of
genetic evaluation for growth can be improved by 5% by replacing a multi-trait model
with a random regression model. Most of this improvement came from more appropriate

modeling of variances and genetic parameters and relied on the assumption that the

random regression model correctly described the covariance structure in the data (Meyer,
2004). Random regression allows the advantage of individually modeled random terms
that more accurately account for environmental variation and compute estimates of
genetic merit for selection and evaluation of breeding animals.
Meat Quality

While growth and composition are major concerns in the swine industry, meat
quality of pork products is also affected by genetic influences. Selection for pigs that
have a higher proportion of muscle and reduced amounts of fat may negatively affect
meat quality characteristics. Wood (1985) reported that work with leaner pigs (below 10
mm P2 fat thickness) suggested increased occurrence of slightly less juicy pork products.
Genetic correlations of carcass leanness to ultimate pH (-0.13), reflectance (0.16), and
drip loss (0.05) (Sellier, 1998) suggest decreased meat quality with leaner pigs. Brewer
et a1. (2002) characterized quality attributes of pork derived from pigs of divergent
genetic background in closed terminal sire lines. These lines were Duroc, Pietrain (Hal-
1843 normal), Pietrain (Hal-1843 positive), Berkshire, Hampshire (rn+), Hampshire
(RN'), and a synthetic line. They reported differences in meat quality color
characteristics, cooking loss, and shear force between these breeds, but did not find
differences in pH or flavor characteristics. Breeds or lines introduced into commercial
application should improve these meat quality characteristics or at least not detract from
current levels of meat quality.

Heavier slaughter-weight pigs can also be a concern for meat quality. Slaughter
weights for swine in the United States have been increasing steadily over recent years

from a mean of 108 kg in 1977 to 121 kg in 2004 (USDA, 1998; USDA, 2005a).

10

Cisneros et a1. (1996) reported significant linear regression coefficients on slaughter
weight (in kg) for lighter color (-0.006), less firmness (-0.009), a lower 24-h pH (-0.002),
higher drip loss (0.029), and decreased tenderness (-0.015) in heavier pigs. All these
measures led to a decrease in overall pork quality as market weight increased. In the
same study, non-significant effects of slaughter weight on growth rate and feed efficiency
were reported. As breeding scheme choices are made for grth and composition goals,
meat quality must also be considered.
Impact of Management on Growth and Carcass Quality

Growth performance can be dependent on environmental and housing differences
with animals of similar genetic merit showing differing patterns of growth performance
due to differences in management (Hamilton et al., 2003). Gentry et a1. (2002) reported
few differences in grth and meat quality traits between pigs reared in a deep bedded
semi-open building and those reared on a slatted floor in a conventional building.
Although only 46 and 56 pig carcasses from deep bedded versus slatted floored pens,
respectively, were tested for differences in carcass characteristics, some measures were
significantly different. Pigs from deep bedded environments had a heavier cold carcass
weight, more backfat at the first rib, last rib, and last lumbar positions, shorter carcasses,
and higher firmness scores on longissimus muscle chops. No differences were reported
for 24 h pH, longissimus muscle area, color score, marbling score, or objective color
scores of L*, a*, or b*. Edwards (2003) stated that energy requirements of outdoor pigs
in Northern Europe were generally higher because of increased climatic energy demand,
while protein requirements were relatively unaffected. These climatic factors would be

similar for pigs raised in the upper Midwest of the United States. Additionally, Edwards

11

(2005b) reviewed studies measuring meat quality attributes in indoor versus outdoor
systems and concluded that no difference in many meat quality attributes were
discovered, including measures of juiciness, tenderness, or meat flavor, but pigs reared in
outdoor systems had a reduced muscle pH at harvest. While these diverse environmental
conditions changed phenotypes of genetically similar pigs, research remains to be done to
determine if phenotype differences will occur between different types of indoor finisher
facilities. Knowledge of how these different systems affect phenotypes can affect
management and marketing practices and decisions for pork producers.
Duroc and Pietrain Breeds

Throughout the years, the Duroc breed has served as a terminal sire population
and as a reference sire breed in many research evaluations. Duroc animals and their
progeny have been compared in nation-wide breed comparisons in different countries
(Kennedy et al., 1996; Moeller et al., 1998), as reference sires to newly imported breeds
(Young, 1992a,b), and in studies of heterosis and mating schemes (McLaren et al.,
1987a,b; Langlois and Minvielle, 1989a,b; Kuhlers et al., 1994; Blanchard et al., 1999).
In general, Duroc pigs and their offspring have been found to grow faster, but also have
more backfat than other breeds (Blasco et al., 1994; Kennedy et al., 1996; Moeller et al.,
1998; Blanchard et al., 1999). At the same time, Duroc animals tend to have greater rates
of lean gain because of their faster overall rate of gain. Durocs tended to have heavier
ham and shoulder weight, but similar loin weight compared to other domestic breeds of
Hampshire, Landrace, and Yorkshire (Langlois and Minvielle, 1989b). One study has
reported that Duroc pigs were more efficient converters of feed to gain (McLaren et al.,

1987a), while another study reported Durocs to be less efficient (Mrode and Kennedy,

12

1993). One area in which Duroc pigs excel is meat quality. Pork from Duroc and Duroc
sired pigs tends to have lower shear force and cooking loss, better color and marbling
scores, and higher pH than other breeds (Langlois and Minvielle, 1989b; Oliver et al.,
1994; Blanchard et al., 1999; Jeremiah et al., 1999). Duroc pigs have been characterized
as fast growing, slightly less lean, but favorable for meat quality.

The Pietrain breed has been used in European production systems, but little
comparative data have been reported regarding their merit in US. production systems.
Lean et al. (1972) conducted an early study of purebred Pietrain pigs in Europe and found
they were lower in fat content, similar in feed efficiency, but had more meat quality
defects than Landrace pigs. McKay et al. (1985) reported Pietrains grew slower, but had
larger hams than Yorkshire or Minnesota No.1 pigs. McKay also reported that slower
body weight growth combined with less backfat allowed Pietrains to have similar lean
tissue grth rates to Yorkshire pigs. F ortin et a1. (1987) noted that Pietrains
demonstrated early maturing characteristics with leaner carcasses compared with Large
Whites. Quiniou and Noblet (1995) used Pietrain boars in their study of equations to
predict composition because of their propensity towards leanness, and found them to be
leaner than either Large White or Meishan pigs (P < 0.05), but similar in leanness to a
synthetic line used in the study. Whittemore et a1. (2003) conducted a serial slaughter
experiment using crossbred pigs that were sired by Landrace, Pietrain, or 50%
Meishan/25% Large White/25% Landrace boars. These sires were mated to J SR
Genepacker 90 primiparous females to generate types with tendencies to be ‘lean’,
‘blocky’, or ‘fatty’ (Whittemore et al., 2003). The Pietrain crossbred pigs were the

leanest and had a slower rate of fatty tissue deposition while also having the largest

l3

Iongissimus muscle area. Estimates of whole body protein and lipid content as a function
of pig live weight also indicated that the Pietrain pigs were the leanest of the three
groups. Purebred Pietrain and Pietrain crosses are generally leaner animals with a higher
proportion of muscle in valuable wholesale cuts than other breeds.
Duroc vs. Pietrain Studies

Few studies have been undertaken to compare Duroc and Pietrain animals for
growth, composition, and meat quality. Those that have been reported have conflicting
results for the traits studied, but varying ending weights across the studies may have
contributed to these observed differences. Kanis et a1. (1990) used Duroc and Pietrain
animals to study the effects of recombinant porcine somatotropin (rpST). Among control
animals (those not receiving rpST), Pietrain animals had better feed efficiency and lean
growth rate in early grth and were leaner at all weights, but had similar feed efficiency
and lean growth rate over the entire growth period to Durocs. Affentranger et al. (1996)
reported faster growth rate with more backfat, but worse feed efficiency for Duroc
animals as compared to Pietrains. Meat quality measures of pH and water holding
capacity were better for Duroc pigs in this study. Average daily gain was similar for
Duroc and Pietrain influenced animals in a study by Ellis et al. (1996) with Pietrain
influenced pigs having less backfat and a larger loin muscle area. Meat quality measures
again favored Duroc animals for marbling and Warner-Bratzler shear force. No
differences were reported for color score or cooking loss. Garcia-Macias et a1. (1996)
also reported less backfat and larger loin muscle area for Pietrain animals as compared to
Durocs. Similar weight of ham, loin, and shoulder primal cuts were reported for both

Duroc and Pietrain progeny with larger belly cuts in Pietrain progeny in this study.

14

Again, Duroc progeny had better 24-h pH, but no difference was discovered for
subjective or objective color scores.

Discrepancies reported in these studies may be due to different end weights used.
Many studies found Pietrain-sired pigs to have favorable characteristics at lighter
slaughter weights, but these differences may not be as apparent at heavier weights
typically seen in US. production systems. These studies also used Pietrain animals that
carried the malignant hypertherrnia mutation in the RYR-l gene. Pietrain animals that do
not carry this allele are now available for use in the US. pork industry. Edwards et a1.
(2006) compared growth traits between pigs that did not carry the mutation in the RYR-l
gene sired by Duroc or Pietrain boars and grown to a common age. Duroc-sired pigs
grew faster, but had more backfat than Pietrain-sired pigs. The combination of these
differences led to similar rates of fat-free lean accretion from 10 to 26 wk of age. Using
the same experimental animal group, Edwards et a1. (2003) also reported on carcass
composition and meat quality differences. Pietrain-sired pigs were leaner and had larger
longissimus muscle area than Duroc-sired animals, but Duroc-sired animals excelled in
meat quality measures of color, marbling, firmness, pH 24 h postmortem, and drip loss
percentage. Similar results were also reported by Cassady et a1. (2002) in a study of
heterosis and recombination effects on pig growth and carcass traits. Purebred Duroc
pigs had faster ADG from 10 to 26 wk of age and more backfat than Pietrain pigs, but
Duroc pigs also had smaller loin muscle area than purebred Pietrain pigs. Rauw et a1.
(2003) compared pigs sired by Duroc or Large White/Pietrain boars and reported no
difference in growth for progeny, but did report that Duroc-sired pigs were mostly fatter

and had more marbling. Duroc- versus Pietrain-sired crossbred pigs from F1 German

15

Landrace by Large White and Leicoma by (German Landrace by Large White) females
were compared for carcass composition and meat quality traits (Kuhn et al., 2005).
Again, Duroc-sired animals had lower lean percentage, but better meat quality
measurements of color, drip loss, and intramuscular fat. Thus, Duroc and Pietrain
animals can both be used in selection programs to obtain phenotypes that match
commercial production objectives.
Swine Resource Populations

While individual breeds can be studied for improvement of traits, the creation of
resource populations allows researchers to discover the underlying alleles that influence
these traits. Development of backcross or F2 resource populations for the discovery of
QTL has occurred at several research institutions worldwide. These populations have
involved many rustic or indigenous breeds as well as commercial populations, and a
summary of the worldwide reported populations is presented in Table 1.1. The
publications listed in Table 1.1 represent the initial genome scan publication for each
resource population. Additional phenotypic data and marker genotypes have been
collected for many of these populations and reported in subsequent publications which
are not listed in Table 1.1. Publications that divided phenotypes into multiple

publications but that were published simultaneously are listed together.

16

Table 1.1. Summary of F2 resource populations including original QTL manuscript for
each population, founder breeds, number of offspring, number of markers used, and types
of phenotypic data analyzed for QTL analysis.

 

 

Original QTL manuscript F0 sire F o dam F1 F I Trait
and year breeda breeda Mb Fb Pc Gd classesc

Andersson et al., 1994 EWB, 2 LW, 8 4 22 193 117 G, C

Rathje et al., 1997f 1,5; C,4 C,14; 1,12 10 50 114 55 R

Rohrer and Keele, 1998a,b M,5; WC,5 WC,5; M,5 BCg 41 540 156 C

Walling etal., 1998 LW,2; M,2 M,2; LW,2 7 25 390 9 G

Wang et al., 1998“ Meishan, 2 Duroc, 2 7 99 15 G, C, MQ

2

(330 F2 analyzed jointly) Meishan, 2 Hampshire, 2 3 5 111
Meishan, 2 Landrace, 2 2 4 46
Minzhu, 2 Hampshire, 2 2 4 69
Minzhu, 2 Landrace, 2 2 4 5

de Koning etal., 1999 Meishan, 19 LW&DL, 126 19 131 619 127 G, R
Paszek et al., 1999 Meishan, 3 Yorkshire, 7 (18 total) 298 119 G
Pérez-Enciso etal., 2000 Iberian, 3 Landrace, 31 6 73 250 7 C
Wada et al., 2000 Gottingen, 1 Meishan, 2 2 19 265 318 G, C
Bidanel et al., 2001 LW, 6 Meishan, 6 6 23_ 1090 137 G
Grindflek et al., 2001 Duroc, 5 Landrace, 5 5 8' 305 29 C, MQ
Malek et al., 2001a,b Berkshire, 2 Yorkshire, 9 8 26 512 125 G, C, MQ
Nezer et al., 2002 Pietrain, 27 LW, 20 31 82 528 137 G, C
Su et al., 2002 LW, 3 Meishan, 7 5 23 66 48 G
Geldermann et al., 2003 Meishan, 1 Pietrain, 8 3 19 316 185 G, C, MQ
EWB, 1 Pietrain, 9 2 26 315
EWB, 1 Meishan, 4 2 21 335
Lee et al., 2003 KN, 5 Landrace, 9 11 36 240 24 G
Sato et al., 2003 Duroc, 1 Meishan, 1 4 24 864 180 G, C, MQ
Zuo et al., 2003 LW, 3 Meishan, 7 5 23 140 24 MQ
Stearns et al., 2005 Berkshire, 3 Duroc, 18 6 56 806 30 G, C

 

3 Breed abbreviations: DL = Dutch Landrace, EWB = European wild boar, KN = Korean
native, LW = Large White, M =Meishan, WC = White Composite

b M = Males, F = Females

° Largest number of animals for any one trait in each manuscript

d Number of genetic markers used in genotyping in each manuscript

e Traits analyzed in the manuscript
G = Growth, C = Carcass characteristics, MQ = Meat quality, R = Reproduction

fI = Increased ovulation rate and embryonal survival line, C = Random selection control
line (Neal et al., 1989)

3 BC = Backcross design

T‘ Family structures from Yu et a1. (1995)

‘Norwegian Slaughter Pig cross sows (50% Norwegian Landrace-50% Yorkshire)

17

The seminal F2 resource population was created by Andersson et al. (1994) and
involved breeding two European wild boars with eight Large White sows. This cross of
an undomesticated breed with a commercially used breed resulted in 200 F2 animals
which exhibited variation in traits of growth and backfat thickness. Although the small
number of F2 animals prohibited finding QTL with smaller effects, QTL affecting 7.5 to
18.7 % of F2 variance were discovered. This population spawned other efforts to create
resource populations from several other breeds.

A breed utilized in many resource populations is the Meishan breed from China,
which excels for reproductive traits (Rohrer and Keele, 1998a,b). These animals were
often crossed with commercially used breeds, and traits of growth, composition, meat
quality, and reproduction were studied. Of the 25 populations in Table 1.1, 14 included
Meishan germplasm. Recently, resource populations have been created from breeds
utilized in commercial production (Grindflek et al., 2001; Malek et al., 2001a,b; Nezer et
al., 2002; Stearns et al., 2005). The advantage of these populations is the ability to
quickly incorporate QTL results into breeding schemes that are already in use in
commercial production.

The number of F2 animals generated in the populations listed in Table 1.1 varied
from 5 to 1090 animals with many of the populations containing 200 to 350 animals and
three of the populations having more than 800 animals included in the initial publications.
As more animals are included in the analyses, QTL that control a smaller percentage of
phenotypic variation can be discovered. A sample size of 1050 animals is required to
find a QTL that controls 1% of phenotypic variation at an error rate of 0.05 with 90%

power (Weller, 2001). Although the populations with 100 to 200 animals have identified

18

putative QTL (e.g. Andersson et al., 1994), more F2 animals allows more recombination
events to take place, so more QTL can also be found in larger populations.

These populations also differed greatly in the number of genetic markers
analyzed. The studies that genotyped more than 115 markers were able to obtain a full
scan of the genome, while those studies that genotyped fewer than this number of
markers chose to target specific chromosomes with their genotyping efforts. As more
markers are genotyped on each animal, the distance between markers decreases and the
confidence interval for the estimated position of putative QTL is reduced. The ultimate
goal of a resource population is to identify a chromosomal region containing a gene that
controls a significant portion of the variation for a trait of interest. Further saturation of
potential positions of QTL on chromosomes with additional linked markers allows a
more accurate description of the QTL position (e. g. Rattink et al., 2000; Thompsen et al.,
2004) and facilitates fine mapping to possibly identify the causative gene at the QTL.

An ideal resource population would be derived from breeds that are
phenotypically distinct, but with alleles in both breeds that are segregating and could be
identified in commercial breeds used in pork production. If these breeds are used in
current breeding schemes, it is a straightforward task to incorporate selection for
beneficial QTL into breeding objectives and ultimately influence commercial production.
Ideally, resource populations would be fixed for alternative alleles at each marker, which
would allow more power of test at each marker position (Weller, 2001). The creation of
a population should involve the generation of enough F2 animals and possibly subsequent
later generations to discover QTL that affect a smaller percentage of phenotypic

variation, and the measurement of phenotypes should be numerous enough to allow

19

discovery of QTL that affect traits important to pork production. Finally, the markers
used should cover the entire genome, as traits of interest have been found on every
autosome and the X chromosome (Hu et al., 2005). Proper design and creation of a
resource population will lead to discovery of important QTL and will allow for their
implementation into breeding programs.
Growth QTL

Many recent studies have undertaken the task of determining QTL for various
growth phenotypic traits. Description of growth can involve weights at different ages,
gain in weight between ages, and weight gain partitioned into protein and fat
components. Difficulties in comparison of traits from one population to another arise as
phenotypes are defined in many different ways across populations. Table 1.2 represents
QTL for birth weight and ADG during the finishing phase although ADG is reported in
many different formats in different studies. A summary of QTL studies published
through mid-2004 (Hu et al., 2005) attempted to combine some, but not all, similar
phenotypes. Comparison of off-test or harvest weight QTL across studies is difficult due
to the many different protocols for growth evaluation and marketing pigs for harvest.
These QTL for growth have been discovered on most of the autosomes except 5, 11, 15,

16,17, and 18.

20

Table 1.2. Summary of reported QTL chromosomal locations for grth traits.

 

 

Traita Manuscript Chromosome
Birth weight Knott et al., 1998 1, 12, 13
Paszek et al., 1999 4
Wada et al., 2000 1
Bidanel et al., 2001 4, 7
Malek et al., 2001a 3
Knott et al., 2002 4
Quintanilla et al., 2002 3
Sato et al., 2003 7
ADG on test Knott et al., 1998 2, 4, 10
Malek et al., 2001a 2, 4, 8, 9
ADG, 10 to 22 wk of age Bidanel et al., 2001 3, 4, 6, 7
Knott et al., 2002 4
Quintanilla et al., 2002 8, 9
Growth rate, 25-90 kg de Koning et al., 2001 1, 2, 4, 6, 7, 8, 12, 13, 14
Sato et al., 2003 6

 

a Hu et al. (2005) summary is the basis for trait names

21

Carcass Merit QTL

Backfat thickness (in varying locations) is the most commonly reported class of
QTL for carcass traits. One reason that so many backfat QTL have been reported is the
possible pleiotropic effects of these QTL that affect fat deposition at many body
locations. Regions that have been reported to contain QTL in several studies include
SSC 1, 4, 6, and 7 (Table 1.3). Not all studies have conducted entire genome scans when
searching for QTL. Several studies have targeted these previously mentioned
chromosomes, which may have disproportionately increased the number of reported QTL
on them.

Average backfat has been measured in several studies because of its significant
QTL reported in Andersson et al. (1994). However, the location of measurement and the
number of measurements combined to calculate average backfat has not been consistent
across studies, so the trait is not summarized here. Instead, locations of QTL for
individual points of measurement for backfat traits have been reported in Table 1.3.
While some QTL for the backfat thickness at the first rib, tenth rib, last rib, and last
lumbar vertebra share the same location, individual studies have reported unique QTL for
each trait. Another aspect of carcass composition is the size of the longissimus muscle
area. This important muscling measure has QTL that have been identified in several
studies, but generally with only one QTL location in each study. Nevertheless, QTL for
Iongissimus muscle area have been reported on SSC 2, 4, and 6, and these locations were

identified in eight of the ten publications in Table 1.3.

22

Table 1.3. Summary of reported QTL chromosomal locations for carcass traits.

Trait

Manuscript

Chromosome

 

First rib backfat

Tenth rib backfat

Last rib backfat

Last lumbar vertebra backfat

Longissimus muscle area

Rohrer and Keele, 1998a
Wang et al., 1998
Varona et al., 2002
Stearns et al., 2005

Rohrer and Keele, 1998a
Malek et al., 2001a
Stearns et al., 2005

Rohrer and Keele, 1998a
Wang et al., 1998

Malek et al., 2001a
Milan et al., 2002
Varona et al., 2002
Stearns et al., 2005

Rohrer and Keele, 1998a
Wang et al., 1998

Malek et al., 2001a
Stearns et al., 2005

Andersson-Eklund et al., 1998
Rohrer and Keele, 1998b
Jeon et al., 1999

Pe’rez-Enciso et al., 2000
Malek et al., 2001a

Ovilo et al., 2002b

Varona et al., 2002

Wimmers et al., 2002

Sato et al., 2003

Stearns et al., 2005

7,10,X

1, 5, 7,14, X
7

1, 4, 5, 7

2

,8, 11,14,X

A

NO-h'NQF-‘hNt—‘w
”-1:-
0\

ON

 

23

Meat Quality QTL

While growth and body composition measurements can be obtained on live
animals, other traits can only be measured after harvest. The inability to collect meat
quality phenotypes on the live animal increases the importance of discovering meat
quality QTL for use in selection of prospective parents of the next generation. Table 1.4
lists reported locations of QTL for meat quality traits grouped by similar trait
classification. The trait of 24 h pH is an important indication of meat quality in the
carcass and QTL for this trait have been reported on 13 different chromosomes. Only
SSC 6, 14, and 15 contain QTL that have been reported in multiple studies. Subjective
color, marbling, and firmness scores are fresh meat quality attributes that have QTL that
have been reported in two studies (Malek et al., 2001b; Stearns et al., 2005). A QTL has
been reported on SSC 2 in both studies for color. In addition, Stearns et a1. (2005)
reported a QTL on SSC 2 for marbling. Objective color scores measure reflectance (L*),
redness, (a*), and yellowness (b*) and have been measured across several resource
populations. While all autosomes except 9, 10, 11, and 12 have QTL reported for
objective color scores, regions on SSC 4, 7, and 14 have been reported in multiple studies
to affect these traits.

Specific QTL have been identified for water retention parameters, shear force
measurements, and sensory taste panel attributes. The analysis of water retention
measurements (drip loss, cook loss, and water holding capacity) is important to
processors of pork products and can determine how well these products will perform in
further value-added processing. Five studies have reported putative QTL for these water

retention traits. Shear force has been measured with differing procedures (e.g. Warner-

24

Bratzler or Instron), but they attempt to quantify a similar attribute of cooked products.
Although significant QTL for shear force have been reported in three studies, none of the
QTL positions have been replicated across the studies. An additional tool to measure
cooked product palatability is the trained sensory panel. While individual sensory panels
evaluate meat products with different organoleptic descriptions on differing scales, QTL
for the traits of juiciness, tenderness, and off-flavor were discovered in Malek et al.
(2001b) and Stearns et al. (2005). Similar to subjective fresh meat scores, QTL were
found on SSC 2 for two of these three traits in both studies. Only Stearns et al. (2005)
reported a QTL on SSC 2 for juiciness. Intramuscular fat percentage can be directly
determined through chemical extraction (AOAC, 2000) and has been reported in several
studies (Table 1.4). Most studies have reported just one QTL for intramuscular fat in
their results, and multiple studies have reported the location of this putative intramuscular

fat QTL on either SSC 4 or SSC 6 (Table 1.4).

25

Table 1.4. Summary of reported QTL chromosomal locations for meat quality traits.

 

 

Trait Manuscript Chromosome

24 h pH Bertram et al., 2000 15
de Koning et al., 2001 4, 9, 11, 14, 18, X
Malek et al., 2001b 5, 6, 14, 15
Ovilo et al., 2002a 3
Geldermann et al., 2003 6, X
Su et al., 2004 2, 6, 7

Color Malek et al., 2001b 2, 12, 17
Stearns et al., 2005 2

Marbling Malek et al., 2001b 1, 8, 10
Stearns et al., 2005 2

Firrnness Malek et al., 2001b 2
Stearns etal., 2005 2

L* de Koning et al., 2001 1, 3, 4, 14
Malek et al., 2001b 2, 4, 5, 7, 14, 15, 17, 18
Ovilo et al., 2002a 4, 7
Sato et al., 2003 3

a* Bertram et al., 2000 15
de Koning et al., 2001 13, 14, 15
Ovilo et al., 2002a 4, 7, 8
Geldermann et al., 2003 6
Stearns et al., 2005 2, 13

b* de Koning et al., 2001 13, 14
Stearns et al., 2005 2

Drip loss Bertram et al., 2000 15
de Koning et al., 2001 4, 6, 14, 18
Malek et al., 2001b 1, 2, 11
Stearns et al., 2005 2

Water holding capacity Malek et al., 2001b 2, 13
Su et al., 2004 1, 4, 6

Cook loss de Koning et al., 2001 7, 18
Malek et al., 2001b 14
Stearns et al., 2005 13

 

26

Table 1.4 (cont’d).

 

Shear force

J uiciness

Tenderness

Off-flavor

Intramuscular fat

de Koning et al., 2001
Malek et al., 2001b
Stearns et al., 2005

Malek et al., 2001b
Stearns et al., 2005

Malek et al., 2001b
Stearns et al., 2005

Malek et al., 2001b

de Koning et al., 1999
Harlizius et al., 2000
Rattink et al., 2000

de Koning et al., 2001
Grindflek et al., 2001
Ovilo et al., 2002b
Szyda et al., 2002
Zuo et al., 2003

Su et al., 2004
Stearns et al., 2005

 

27

F2 Population QTL Analysis Procedures

Methodology to analyze F2 or backcross populations for putative QTL has
evolved as more emphasis on error structures and accounting for environmental variance
components have become increasingly more complex. Initially, line cross analyses
focused on associating single markers with phenotypes, but were soon replaced by
methods to estimate likelihood of any region of the genome controlling part of the
variance of the trait. This method is called interval mapping and involves calculating the
probability of recombination between markers and the QTL position and how this QTL
relates to the phenotype of interest (Lander and Botstein, 1989). Further simplification of
this methodology used least squares regression to regress the phenotypes against the
probabilities that an individual has a certain genotype at each position in the genome
(Haley and Knott, 1992; Haley et al., 1994). This analysis includes probabilities for an F2
individual to inherit a particular allele from which parent at points throughout the
genome. In most analyses this evaluation is performed at continuous 1 cM intervals
across the entire region with segregating markers in the population. The full model that
includes the QTL effects is tested against the reduced model without the effects to obtain
an F-test and that is compared to significance threshold levels. The development of an
extension of this regression analysis to account for QTL that segregate in both breeds at
similar frequency was reported in Knott et al. (1996). This extension involves the
analysis of groups of progeny from one sire or dam and is referred to as the half-sib
analysis. This half-sib analysis can detect QTL that are missed in the line cross analysis,
and can be utilized in populations with outbred founders that have the same alleles

present in both founder breeds. Dekkers et al. (2003) proposed combined regression

28

interval mapping, in which line cross and half-sib analyses are performed jointly. The
QTL discovered in the Dekkers et al. (2003) joint analyses were classified as significant
for line cross only, significant for half-sib only, or significant for the combined analysis,
but not significant for either type of analysis individually. Since significance of a QTL
cannot be obtained from the standard F -tables because of the large number of correlated
tests conducted from the same marker data, permutation tests are typically used to derive
proper significance thresholds (Churchill and Doerge, 1994). Additionally, the accuracy
of the position of QTL can be described through the use of confidence intervals.
Bootstrap methods have been developed to estimate these intervals (Visscher et al.,
1996). The separate line cross and half-sib analyses have been combined with
permutation tests for significance thresholds and bootstrap methods for confidence
intervals into a web interface to create the QTL Express software (Seaton et al., 2002).
Although the regression interval mapping procedures have been used in many
analyses and are a good beginning step to find QTL that have been subsequently
reconfirmed in other experiments (e. g. Walling etal., 1998), it does inherently contain
only fixed effects and can be improved upon for analysis of populations derived from
outbred populations. To improve the estimation of QTL effects, Fernando and Grossman
(1989) proposed modeling the QTL effect as a normally distributed random variable with
mean zero and variance to be estimated. This variance is estimated by assessing the
degree of phenotypic similarity between relatives according to the probability of sharing
identical by descent alleles at specified positions. An alternative analysis scheme has
been proposed by Pe'rez-Enciso and Varona (2000) in which a mixed-model approach

allows for QTL segregation within lines as well as for differences in mean QTL effects

29

between lines. This method allows the estimation of differences in additive variances
between the parental lines. Pe'rez-Enciso and Varona (2000) simulated data to exhibit its
implementation; however, no studies have been reported that have used this method of
analysis. QxPak, a software package that has recently been released, incorporates these
methods into QTL estimation (Pe'rez-Enciso and Misztal, 2004). Most reported F2 animal

studies have implemented line cross least squares regression analysis for QTL discovery.

30

CHAPTER II
INFLUENCE OF FIN ISHER FACILITIES ON PIG GROWTH PERFORMANCEl
Abstract

Pigs from the F2 generation of a Duroc by Pietrain resource population were
finished in either a Modified Open Front or a Test Station building with bedded, solid
floors to evaluate finisher influence on growth, composition, and meat quality traits.
Serial data of body weight, tenth rib backfat, longissimus muscle area, and last rib fat
were obtained at three week intervals from 10 to 22 wk of age. From these
measurements, estimates of fat-free total lean, total body fat, empty body protein, and
empty body lipid were calculated. At harvest, carcass composition traits, carcass
temperature, and longissimus muscle pH were obtained. Primal cut weights, meat quality
evaluation traits, proximate analysis measurements, and sensory taste panel phenotypes
were also recorded. Models that included genetic, permanent environment, and residual
error variance components were used to evaluate these traits. Many of the traits did not
differ significantly between pigs raised in the two building types. Pigs finished in the
Modified Open Front building were heavier at harvest, had more backfat at 22 wk of age
at the tenth rib and last rib, and had more backfat at harvest at the tenth rib, last rib, and
last lumbar vertebra. Serial data analyses revealed differences in patterns of weight,
leanness, and fatness accretion not apparent in single time point measurements. Although
body weight did not differ at 10 or 22 wk of age, the pigs reared in the Test Station

building grew more slowly at first, but then grew more quickly later in the finisher phase

 

' Research for this project was financially supported by the Michigan State University
Department of Animal Science, the Michigan Agricultural Experiment Station, and the
Michigan Animal Initiative Coalition.

31

to achieve the same overall weight gain from 10 to 22 wk of age. Pigs raised in the
Modified Open Front building had a greater backfat accretion rate from 10 to 22 wk of
age than pigs raised in the Test Station building. Additionally, these pigs had a greater
decline in pH from 45 min to 24 h after harvest and were more tender based on Warner-
Bratzler shear force measurements. Heritability of the serial traits generally increased for
all traits from 10 to 22 wk of age. Body weight and Iongissimus muscle area had similar
increasing heritability patterns, whereas tenth rib backfat and last rib backfat had similar
heritability patterns, increasing from 10 to 16 wk of age and remaining constant
thereafter. Thus, animals of similar genetic merit can show differences in growth
patterns as influenced by differing finisher facilities.
Introduction

Growth performance can be dependent on environmental and housing differences
with animals of similar genetic merit showing differing patterns of growth performance
due to differences in management (Hamilton et al., 2003). Gentry et al. (2002) reported
differences in growth and meat quality traits between pigs reared indoors or outdoors.
While these diverse environmental conditions changed phenotypes of genetically similar
pigs, further research is needed to determine if differences in growth will occur when
pigs are housed in different types of enclosed facilities. Knowledge of how these
different systems affect phenotypes can affect management and marketing practices and
decisions for pork producers. To determine the influence that differing facility types may
have on pig growth, a genetic resource population with known pedigrees can be an ideal

study group. With known pedigrees and similar genetic merit amongst animals in a

32

resource population, those genetic factors can be accounted for and actual differences in
management practices can be tested.

Traits that are measured once in the lifetime of an animal do not represent how
the phenotype of that animal may change as the animal matures. Phenotypes such as
body weight and composition traits can be measured serially over time and modeled
through the use of random regression (Meyer, 2004). These procedures allow
(co)variance matrices to be estimated that model the relationship of the trait at different
points throughout the measurement period. Previous reports have demonstrated the
usefiilness of random regression modeling of weight data in pigs (Huisman et al., 2002;
Edwards et al., 2006). Evaluation of serially measured traits allows more in depth
characterization of phenotypic differences. The objective of this study was to compare
growth, composition, and meat quality traits for pigs of similar genetic merit finished in
two different types of finishing buildings.

Materials and Methods
Population Development

A three-generation resource population was developed at Michigan State
University and used to study traits of growth, body composition, and meat quality.
Semen from four F 0 Duroc sires from a closed unselected control population (Kuhlers et
al., 2003) and sixteen F0 Pietrain dams from a closed herd propagated the F 1 generation.
All F0 animals were determined homozygous normal for the RYR-l gene by a DNA test
(Fujii et al., 1991). All animals were produced through artificial insemination at the
Michigan State University Swine Teaching and Research Farm. From F1 progeny, 51

females and six males (sons of three F0 sires) were retained to produce 1259 F2 pigs born

33

in 142 litters across 11 farrowing groups. Females were retained across multiple parities
to produce F2 progeny.
Animal Management

Sows were placed into farrowing crates one week prior to farrowing. Baby pigs
were processed (individually identified by ear tag, given 0.5 ml penicillin and 1 ml iron
dextran subcutaneously, and tails clipped) at approximately 1 d of age. Pigs were weaned
at 16-25 (mean of 19.8) d of age and were sorted into nursery pens by sex and weight.
All pigs were managed similarly in farrowing and nursery stages. At 10 wk of age, F2
pigs were placed into one of two finishing facilities at the Michigan State University
Swine Teaching and Research Farm. Farrowing groups 1, 3, 5, 7, 9, and 11 (n = 521
pigs) were placed into a Modified Open Front (MOF) building with eight pens, in which
four pens differed in size from the other four. Four larger pens (2.03 m by 6.91 m) with
two-space feeders were targeted to contain 16 pigs. Four smaller pens (1.42 m by 6.91
m) with one-space feeders were targeted to contain 12 pigs. Each pen had two-thirds
solid, one-third slatted floors and wet-dry feeders. Farrowing groups 2, 4, 6, 8, and 10 (n
= 465 pigs) were placed into a test station (TS) facility in which pens had solid floors and
were bedded with straw or wood shavings and had single-space dry feeders and cup
drinkers. All 25 pens utilized (1.42 m by 4.93 m) were targeted to contain four pigs. All
diets fed were Michigan State University standard swine farm diets that met or exceeded
NRC (1998) requirements for all nutrients at each production stage. Pigs in both

facilities had ad libitum access to feed and water.

34

Trait Collection

Live animal traits collected on F2 animals included BW at birth, weaning, 6, 10,
13, 16, 19, and 22 wk of age. An ADG firom 10 to 22 wk of age was calculated.
Additionally, B-mode ultrasound (Pie Medical ZOOSLC, Classic Medical Supply, Inc.,
Tequesta, FL) estimates of tenth rib backfat (BFlO), last rib backfat (LRF), and
Iongissimus muscle area (LMA) were recorded at 10, 13, 16, 19, and 22 wk of age. At
each of these time points, estimates of fat-free total lean (FF TOLN), total body fat tissue
(TOF AT), empty body protein (EBPRO), and empty body lipid (EBLIPID) were
calculated using equations similar to those used by Wagner et a1. (1999). Animals were
also weighed prior to leaving the farm for harvest.

At harvest, pigs were transported to one of two abattoirs. A total of 176 pigs were
harvested at the Michigan State University Meats Laboratory (East Lansing, MI), and the
remainder of the pigs were transported to a small federally inspected plant in western
Michigan (DeVries Meats, Coopersville, MI). Both groups were fasted and allowed to
rest overnight with access to water. Ear tag and tattoo numbers were collected at
slaughter to maintain identity of each carcass. Carcass traits that were collected included
hot carcass weight and pH and temperature in longissimus dorsi at 45 min and 24 h
postmortem. After overnight chilling, measurements taken according to National Pork
Producers Council guidelines (N PPC, 2000) included midline first rib backfat, last rib
backfat, last lumbar backfat, and carcass length. Weights of primal cuts of ham, closely
trimmed loin, picnic shoulder, Boston shoulder, belly, and spareribs were recorded.
During carcass fabrication, measurements of tenth rib backfat and longissimus muscle

area were also recorded. A section of loin from the tenth rib to the last rib was returned

35

to Michigan State University for further meat quality analysis. All measurements were
taken from the left side of each carcass.

Boneless longissimus dorsi were removed from loin sections and external fat
removed. A small portion of longissimus dorsi was diced and frozen for proximate
analysis. Two 2.54 cm thick chops were cut from the anterior end for fresh meat quality
analysis. The two chops were allowed to bloom for a minimum of 10 minutes and
evaluated for subjective scores of color and marbling (N PPC, 2000) and firmness (NPPC,
1991). The color score scale ranged from 1 (pale pinkish gray) to 6 (dark purplish red).
The marbling score scale was 1 to 10 (closely approximating fat percentage). The
firmness score scale was 1 (very soft and watery) to 5 (very firm and dry). Additionally,
objective color scores of CIE L* (lightness), a* (redness), and b* (yellowness) were
obtained using a Minolta CR-310 colorimeter (Ramsey, NJ) with a D65 illuminant and a
two-degree standard observer. Chops were weighed, hung in sealed plastic bags for 24 h
at 4°C, and then weighed again for drip loss measurement. The remaining section of the
longissimus dorsi was vacuum packaged, aged 7 d at 4°C, and frozen for further meat
quality tests of cook yield, shear force, and sensory taste panel analysis.

From frozen loin sections, two 2.54 cm thick chops were cut for cook yield and
Warner-Bratzler shear force (WBS) analysis. For cook yield measurements, each chop
was thawed, weighed, cooked to 71°C internal temperature on a Taylor clamshell grill
(Model Q824, Taylor Co., Rockton, IL), cooled to room temperature, and weighed again.
From these chops six cores (three cores from each chop) were taken parallel to the
muscle fiber direction using a drill press-mounted corer. Cores were sheared

perpendicular to muscle fibers using a Warner-Bratzler head on a TA-HDi texture

36

analyzer (Texture Technologies Corp., Scarsdale, NY). The cross-head speed was 3.30
mm/s. Samples for proximate analysis were ground using dry ice and measured for
moisture (oven drying), fat (soxhlet ether extraction), and protein (nitrogen combustion,
Model FP-2000, Leco Inc., St. Joseph, MI) following AOAC procedures (2000).

A trained panel of seven healthy adults (ages 20-65) was utilized to determine
specific sensory attributes of each longissimus chop. The sensory panel was trained
according to Meilgaard et al. (1991) and AMSA (1995). All panelists had experience in
sensory evaluation and were previously trained to evaluate various meat products. Each
sample was evaluated for juiciness, muscle fiber and overall tenderness, connective
tissue, and off-flavor using an 8 point hedonic scale. Higher scores were more favorable
in each of the first four categories and indicated extremely juicy, extremely tender, or no
connective tissue for each of these attributes, respectively, while lower scores for off-
flavor were indicative of less off-flavor.

Frozen chops were thawed for 24 h at 26°C and then cooked on a Taylor
clamshell grill (Model QS24, Taylor Co., Rockton, IL). The upper plate was set to
104.4°C and the bottom plate was set to 102.8°C with a 2.16 cm gap between plates.
Temperature was monitored by inserting a copper constantan thermocouple (0.051 cm
diameter, 15.2 cm length, Omega Engineering Inc., Stamford, CT) into the geometric
center of the pork chop. Chops were cooked to a final internal temperature of 71°C.
Sample preparation included cutting 1.27 cm cubes from the center portion of each chop,
and two cubes were placed in 2 oz. souffle cups and covered with a lid. Souffle cups

were placed in a Pyrex two quart bowl with a lid, and the bowl was covered with warm

37

towels to keep the samples warm. The insulated bowl was placed in an insulated
container and transferred to the sensory evaluation room.

Testing took place in climate controlled, partitioned booths with cool
incandescent light. The order of sample preparation was randomized within each session
to minimize positional bias and a 3 digit random code was used to label the samples. The
samples were picked up with a toothpick, chewed with the molars, and evaluated.
Expectorant cups were provided to prevent taste fatigue and distilled, deionized water
was used to clean the palate between samples. The panelists were standardized each day
by evaluating a warm-up sample and discussing the results. A total of 18-24 samples
were evaluated on each day, and the day was divided into three sessions with a 15 min
break between each session. A total of 958 animals had carcass and meat quality traits
measured.

Trait Analysis

The 22 wk of age off-test traits, carcass phenotypes, and meat quality traits were
analyzed with a mixed model that included the fixed effects of sex and finisher and the
random effect of farrowing group for off-test traits or harvest group for carcass and meat
quality traits. An animal random effect augmented with a (co)variance matrix that
accounted for genetic relationships among animals was also included in all models. For
off-test traits, age at measurement was included as a covariate in the analysis. Table 11.1
lists whether carcass weight, harvest age, or neither covariate was used for carcass and
meat quality trait models. All covariates included in each model were those that were

found to be important to the model (P < 0.20) when each trait was analyzed by ordinary

38

least squares analysis of variance without random effects included. The following model

was used for analysis:

Yiljlklm = H + SCXi + finj + grp(fin)jk + Pen(gfp)kl + gm + Bxijklm + eijklm
w ere
Yijklm = record on the rnth pig within ith sex, jth finisher, kth group, and 1th pen,
u = overall mean of trait,
sex; = fixed effect of sex of animal i (Barrow or Gilt),
finj = fixed effect of finisher j (MOF or TS),
grp(fin)jk = random effect of farrowing group (1-11) or harvest date (1-33) nested
within finisher where grp ~N(0, Iogmz),
pen(grp)kl = random effect of pen (1-25) nested within farrowing group where
pen ~N(0, Iopcnz),
gm = random effect for animal m,
ijk|m= covariate appropriate to each trait such that B is the partial regression
coefficient on Xijkhm and
eijk|m = random error ~N(0, 16,2).

Here, g = {g m} ~ N (0, A03) where A is the numerator relationship matrix among

animals and 0'32 is the additive genetic variance. The relationship matrix accounted for
relationships among progeny. plus the three generations of sire and darn ancestors.
Serial Data Analysis

Serial BW and ultrasound estimates from 10 to 22 wk of age were used to
generate random regression equations to model pig BW, BF10, LMA, LRF, FF TOLN,
TOFAT, EBPRO, and EBLIPID on age at measurement. Age at measurement was
modeled as week on-test, calculated as age in weeks minus nine (i.e. 1, 4, 7, 10, and 13 as
distinct covariate values used in the analysis). A random intercept for each animal and a
linear regression on age for each animal was included in each model. Table 11.2 lists the
polynomial order of week of age and interactions used in models for these eight traits as
determined by log likelihood tests of significance. During preliminary analysis of these

traits a trend for increasing residual variance as age increased was noted. To account for

39

heterogenous residual variances across serial measurement and the relationship between

time points, thee = {eijklm } was specified as normally distributed with a general

(co)variance structure calculated from the data and specified within and across weeks
with five variance and 10 covariance terms. The (co)variances of the random animal
intercept as well as the animal linear by week term were also modeled. The following
model was used:

Yijklmn = u + Zweeki‘l + SCXj + 2(sex*week¢')j + fink + 2(fin*week¢)k + grp(fin)k1 +
pen(grp)lm + gn + an*Zi + eijklmn
where
Yijklmn = record on the nth pig within j‘h sex, kth finisher, lth group, and mth pen
regressed on om polynomial week i,
u = overall mean of trait,
week;¢ = fixed regression coefficients for polynomial terms (1) (1-4) of week i,
sexj = fixed effect of sex of animal j (Barrow or Gilt),
fink = fixed effect of finisher k (MOF or TS),
grp(fin)k. = random effect of farrowing group 1 (1-11) nested within finisher
where grp ~N(0, Iogrpz),
pen(grp).m = random effect of pen m (1-25) nested within farrowing group where
pen ~N(0, Iopcnz),
gn = random intercept for animal 11,
an = random linear regression coefficient on age for animal n,
Z; = week on test as a covariate, and
ejjkmm = random error.

The distributional assumptions on g = { g n} and a = { g n } were such that:

2
0 A0 A0"
[g] ~ N [ ] g 32“ , where a; is the intercept genetic variance for the
a 0 Aaga Ada

individuals, 022: is the linear age by animal genetic variance, and 0' 8a, is the genetic

covariance between the intercept and linear term for each animal.

40

Results and Discussion
Off-T est, Carcass Composition, and Meat Quality

Many of the off-test, carcass composition, and meat quality traits did not differ
significantly between pigs raised in the MOF building and those raised in the TS building
(Tables 11.3 and 11.4). These findings are consistent with Gentry et al. (2002), who
reported no differences for carcass measurements, 24 h pH, drip loss, sensory panel, and
shear force values for pigs raised in deep bedding versus those raised on slatted floors.
At 22 wk of age, BW was not significantly different between the two finishers, but once
pigs reached harvest age, BW was significantly higher in animals from the MOF than the
TS finisher (1 13.4 vs. 110.7 kg, respectively). This difference led to a trend for hot
carcass weight to be heavier in pigs from the MOF finisher compared to the TS finisher,
contrary to Gentry et al. (2002) who reported heavier cold carcass weights in pigs from
bedded facilities versus those on slatted floors. Since BW and hot carcass weight both
varied in the same proportion in this study, dressing percentages for pigs raised in these
two finishers was not different. While neither carcass temperature nor pH were different
between carcasses from pigs in these two finishers, the decline in pH from 45 min to 24 h
after harvest was greater among carcasses from pigs raised in the MOF versus the TS
finisher (0.89 vs. 0.83, respectively).

Off-test BF10 and LRF were significantly greater in pigs from the MOF building
than the TS building (Table 11.3). These differences were also discovered at slaughter
with carcass tenth rib backfat and last rib backfat greater in carcasses from pigs reared in
the MOF versus those reared in the TS building. Additionally, backfat at the last lumbar

vertebra was greater among carcasses from MOF reared pigs than those reared in the TS

41

building (Table 11.4). Gentry et al. (2002) reported similar results of more backfat on
pigs raised in deep bedded pens versus those raised on slatted floors.

All primal cut weights, meat quality evaluation traits, proximate analysis traits,
and sensory taste panel measurements were not significantly different between carcasses
from pigs raised in the MOF compared to those from the TS building (Table 11.4). 1
Additionally, drip loss and cook yield percentages were not significantly different.
Warner-Bratzler shear force values of longissimus chops from pigs raised in the MOF
were lower than values for chops from pigs raised in the TS building (3.14 vs. 3.36 kg,
respectively).

The observed differences in fat measurements may be a result of either differing
feeder types in the two buildings or changes in maintenance energy requirements. Wet-
dry feeders have been reported to cause pigs to have more subcutaneous fat than pigs fed
on dry feeders (Barnes et al., 1999). Since the MOF was equipped with wet-dry feeders,
this could explain part of the differences discovered in our study. An environmental
difference between the MOF and TS finishers was the amount of climate control inherent
to each building. The MOF building was naturally ventilated and had supplemental heat
in the winter. While the TS building had a curtain, wind blocks, and hovers in winter, it
did not have supplemental heat, and pigs were more exposed to the ambient temperature.
Considering these environmental differences, pigs in the TS may have had less fat
because they may have had an increased maintenance energy requirement in the TS
environment versus those raised in the MOF building. Since the energy requirement of
the pigs raised in the MOF may have been less, the excess energy may have been stored

as fat. These differences in subcutaneous fat may be partially responsible for the trend

42

(P = 0.116) of 24 h carcass temperature to be higher for pigs raised in the MOF versus
the TS building. Since higher carcass temperatures can allow faster pH declines, the
carcass temperature difference may have led to the greater pH decline from 45 min to 24
h after harvest for pigs finished in the MOF building than for those finished in the TS
building.

Serial Data Results

Although differences between finishers may not have been significant for some of
the off-test data, further evaluation of the serial data revealed some differences in grth
patterns between the MOF raised pigs and the TS raised pigs. The eight serial traits
analyzed were BW, BF10, LMA, LRF, F FTOLN, TOFAT, EBPRO, and EBLIPID.
These traits were plotted against week of age with standard error bars included in Figures
11.1-8. Figure 11.1 shows how BW changes differed between MOF and TS raised pigs.
While 10 wk and 22 wk BW did not differ, the shape of the BW curves differed slightly
with the pigs in the MOF building being heavier through the three midpoints of the
finisher phase. Pigs in the TS building grew faster in the latter stages of the 10 to 22 wk
of age period and BW was not different at 22 wk of age.

Figure 11.2 depicts the trend for increasing differences in BF10 as pigs became
older. Pigs reared in the MOF building had more BF10 than those reared in the TS and
continued this trend through harvest as indicated by the differences reported in the
carcass data. In Figure 11.3, LMA took a similar pattern to that reported for BW in this
study with no differences at 10 wk or 22 wk of age. Again, small differences occurred in
the shape of the LMA grth curve between pigs raised in the MOF compared to those

raised in the TS building; however, none were significant. Figure 11.4 for LRF indicated

43

a similar pattern to BF10 over time with pigs from the MOF building having an
increasing trend for more backfat than those from the TS building. These differences in
backfat were significant at 22 wk of age, and the harvest data suggested that the
differences between pigs raised in the MOF building versus the TS building continued to
increase through harvest.

These noted differences in backfat measurements combined with BW and LMA
measurements led to no differences between MOF and TS raised pigs for FFTOLN
(Figure 11.5). Conversely, differences in TOF AT continued to increase as the finishing
period from 10 to 22 wk of age progressed (Figure 11.6). The differences noted for BF10
and LRF contributed to the differences in TOFAT where the MOF reared pigs had
increasingly more TOF AT than the TS reared pigs. Regressions of EBPRO followed a
similar pattern to LMA with no differences between pigs raised in the two buildings
(Figure 11.7). Measures of EBLIPID followed BF10, LRF, and TOFAT trends for pigs
raised in the MOF versus those raised in the TS building, but were not statistically
different between pigs raised in the two buildings (Figure 11.8).

Serial Heritability Results

Estimates of heritability over time were calculated for all eight serial traits. The
heritability plotted over the time period in the finisher facilities from 10 to 22 wk of age
for the 4 measured traits of BW, BF10, LMA, and LRF increased over time (Figure 11.9).
Heritability for BW and LMA increased gradually throughout the period, and heritability
for BW reached 0.11 at 22 wk of age while heritability for LMA reached 0.25 at 22 wk of
age. Kuhlers et al. (2003) reported a similar heritability of 0.13 for body weight of Duroc

pigs at 168 d. Heritability for loin eye area, adjusted to 113.5 kg BW, of 0.32 was

44

reported by Chen et al. (2002). Estimates for BF 1 0 and LRF heritability followed similar
trends to one another with a rapid increase from 10 to 16 wk of age and then a plateau for
the rest of the finishing period to 22 wk of age. Chen et al. (2002) and Kuhlers et al.
(2003) both reported higher tenth rib backfat heritability estimates for Duroc pigs when
adjusted to either 113.5 kg (0.48) or adjusted to 168 d (0.58), respectively.

Heritability estimates for the composition traits of F FTOLN, TOFAT, EBPRO,
and EBLIPID were also calculated from 10 to 22 wk of age (Figure 11.10). Estimates of
heritability of the two leanness traits of F F TOLN and EBPRO were generally low, but
did increase from 10 to 22 wk of age. Heritability for TOF AT increased from 10 to 16
wk of age, but then declined to a similar level as that observed at 10 wk of age. The
EBLIPID heritability generally increased from 10 to 22 wk of age, and was 0.23 at 22 wk
of age. Huisman et al. (2002) also fit random regression models to serial data of BW in
pigs. Although residual error was kept constant throughout their study, heritability for
the models fluctuated around 0.17. Further research may be necessary on models that
account for genetic variance, permanent environment variance, and residual error
variance changes across time as they may have lower heritability estimates than some
other studies since more of the sums of squares are accounted for in variances other than
the genetic variance.

Implications

Analysis of data from the Michigan State University Duroc by Pietrain F2
resource population revealed some insight into the influence of finisher facilities on
growth, carcass merit, and meat quality. While many traits, including 22 wk body

weight, composition traits, and most meat quality traits, did not significantly differ

45

between finisher types in animals of similar genetic merit, some traits, including backfat
measurements, were influenced by finisher type. Collection of serial data allowed for
further characterization of animal growth data and allowed for estimation of serial
heritability from 10 to 22 wk of age. Animals of similar genetic merit can show
differences in grth patterns as influenced by differing finisher facilities. These
differences can lead to changes in feeding and management decisions, including diet

choices and marketing possibilities, based upon finisher facilities utilized.

46

Table ".1. Covariates for carcass and meat quality trait analyses.

 

Covariates“

 

Carcass Harvest
Trait Weight Age

 

Carcass measures
Off-farm BW, kg - X
Hot carcass weight, kg
Dressing percentage, %
45 min carcass temperature, °C
24 h carcass temperature, °C
45 min pH
24 h pH
45 min-24 h pH decline
Carcass length, cm
Number of ribs
First rib backfat, mm
Last rib backfat, mm
Last lumbar vertebra backfat, mm
Tenth rib backfat, mm
Longissimus muscle area, cm2
Primal cut weights
Ham weight, kg
Loin weight, kg
Boston shoulder weight, kg
Picnic shoulder weight, kg
Belly weight, kg
Spareribs weight, kg
Meat quality evaluation
Color, 1-6 -
Marbling, 1-10 - X
Firmness, 1-5 - X
L* - -
a* - -
b* - -
Proximate analysis
Moisture,°/o X
Fat, % X -
Protein, % X
Laboratory analyses
Drip Ioss,% - X
Cook yield, % - -
Warner-Bratzler shear force, kg - -
Sensory taste panel
Juiciness, 1-8 - -
Tenderness, 1-8 - -
Overall tenderness, 1-8 - -
Connective tissue, 1-8 - -
Off-flavor, 1-8 - -

><

><><><><><><|

><><><><><>< ><><><><><I

 

j‘Xindicates used in model, - indicates not used in model

47

Table ".2. Order of polynomial on week of age terms and other
significant terms for random regression analyses of serial growth data.

 

 

 

Significant term"
Week Sex* Finisher"

Trait 01‘ age weekz week2
Body weight, kg 4 - X
Tenth rib backfat. mm 2 X -
Longissimus muscle area, cm2 3 - -
Last rib backfat, mm 4 X -
Fat-free total lean, kg 4 - X
Total body fat tissue, kg 4 X -
Empty body protein, kg 2 - X
Empty body lipid, kg 4 - -

 

3X indicates used in model, - indicates not used in model

48

Table 11.3. Number of observations, least squares means, standard error of the mean, and

P-value of the difference between finisher means for 22 wk of age off-test traits.

 

 

 

MOFa TSa
Trait n Mean n Mean SEM P—value
Body weight, kg 513 99.6 460 100.4 3.23 0.687
Tenth rib backfat, mm 513 21.6 460 18.9 1.25 0.050
Longissimus muscle area, cm2 513 37.5 460 36.6 1.36 0.444
Last rib backfat, mm 513 15.1 460 13.6 0.87 0.050
Fat-free total lean, kg 513 38.1 460 38.6 1.22 0.571
Total body fat tissue, kg 513 25.5 460 24.7 1.92 0.519
Empty body protein, kg 513 14.9 460 15.1 0.48 0.414
Empty body lipid, kg 513 22.5 460 21.6 1.00 0.244
ADG (10-22 wk), g/d 513 874.5 460 876.0 31.1 0.961

 

aMOF = Modified Open Front finisher, T8 = Test Station finisher

49

Table ".4. Number of observations, least squares means, standard error of the mean, and
P-value of the difference between finisher means for carcass and meat quality traits.

 

 

 

 

MOF‘| TSa
Trait n Mean n Mean SEM P-value
Carcass measures
Off-farm BW, kg 50 113.4 45 110.7 1.77 0.015
Hot carcass weight, kg 50 82.5 45 80.7 1.31 0.074
Dressing percentage, % 50 72.8 45 72.9 0.46 0.883
45 min carcass temperature, °C 50 39.2 45 39.3 0.50 0.892
24 h carcass temperature, °C 50 3.02 45 2.46 0.26 0.116
45 min pH 49 6.41 45 6.36 0.03 0.125
24 h pH 48 5.52 45 5.55 0.02 0.358
45 min-24 h pH decline 47 0.89 45 0.83 0.01 0.019
Carcass length, cm 50 78.8 45 78.7 0.49 0.880
Number of ribs 29 15.1 33 14.9 0.17 0.416
First rib backfat, mm 44 41.6 42 40.5 1.13 0.385
Last rib backfat, mm 50 29.8 45 27.3 1.00 0.006
Last lumbar vertebra backfat, mm 50 23.0 45 21.3 1.08 0.021
Tenth rib backfat, mm 50 25.6 44 23.1 0.99 0.001
Longissimus muscle area, cm2 50 40.4 44 41.0 0.95 0.273
Primal cut weights
Ham weight, kg 50 9.48 45 9.60 0.10 0.087
Loin weight, kg 50 8.17 45 8.33 0.11 0.218
Boston shoulder weight, kg 50 3.78 45 3.81 0.14 0.857
Picnic shoulder weight, kg 50 3.92 45 3.84 0.14 0.679
Belly weight, kg 50 4.96 45 5.07 0.08 0.215
Spareribs weight, kg 50 1.47 44 1.48 0.04 0.717
Meat quality evaluation
Color, 16 50 3.06 45 3.14 0.17 0.392
Marbling, 1-10 50 2.84 43 2.96 0.12 0.193
Firmness, 1-5 50 2.76 45 2.88 0.14 0.178
U” 45 54.02 45 53.77 0.53 0.290
a* 45 17.09 45 17.11 0.52 0.967
b“ 45 9.22 45 9.10 0.36 0.781
Proximate analysis
Moisture,% 49 73.92 44 73.73 0.33 0.453
Fat, % 49 3.24 44 3.20 0.32 0.837
Protein, % 49 23.57 44 23.60 0.21 0.832
Laboratory analyses
Drip loss,% 50 1.62 45 1.73 0.25 0.444
Cook yield, % 49 77.65 44 77.72 0.63 0.840
Warner-Bratzler shear force, kg 49 3.14 44 3.36 0.13 0.044
Sensory taste panel
Juiciness, 1-8 50 5.18 44 5.34 0.23 0.074
Tenderness, 1-8 50 5.48 44 5.55 0.22 0.441
Overall tenderness, 1-8 50 5.56 44 5.63 0.20 0.362
Connective tissue, 1-8 50 6.36 44 6.38 0.16 0.783
Off-flavor, 1-8 50 1.14 44 1.12 0.04 0.255

 

°MOF = Modified Open Front finisher, TS = Test Station finisher

50

    

—°—MOF —¢ —TS

80 .
5° .
B 60 _
m -
40 ~
20 L- L.-. -A . L . 2. L W, m. .- _- W . . 7 L 7 7
8 10 12 14 16 18 20 22 24
Week of age

Figure 11.1. Body weight means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

51

22‘

     

E 13 ‘ —O—MOF —n—Ts
E /
e“
i: 14 ~
an
10 -
5- A 4-44 ..7.-* f 4. i. .4 ~45. *4 4.4
8 10 12 14 16 18 20 22 24
Weekofage

Figure 11.2. Tenth rib backfat means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

52

    

-'°—MOF '-¢ -TS

lO . ,y. ,4.— . u. -.. .7 -7 _‘_ .v .,,A_____ . ,_ v A. _7 . 7~ .7 . #_J 4. .

8 10 12 14 16 18 20 22 24
Week of age

Figure 11.3. Longissimus muscle area means with standard errors from 10 to 22 wk of
age for Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

53

16

14 ..

     

,2 . —°—MOF —n -TS

10

LRF, mm

8 10 12 14 16 18 2O 22 24
Week of age

Figure 11.4. Last rib backfat means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

54

A
O

U)
M

  
 

b.)
O

—°—MOF —6 -TS

FFTOLN, kg
8

8 10 12 14 16 18 20 22 24
Week of age

Figure 11.5. Fat-free total lean means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

55

b)
O

N
LII

   

N
O

  

—D—MOF d -TS

TOFAT, kg
3.

p—A
O

8 10 12 14 16 18 20 22 24
Week of age

Figure 11.6. Total body fat tissue means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

56

16-

14'

   

—'°—MOF -*-TS

  

10-

EBPRO, kg

8 l 0 l 2 14 16 1 8 20 22 24
Week of age

Figure 11.7. Empty body protein means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

57

N
U”.

  
 

20
31.9 —O-MOF --6 -TS
. 15 .
e __
a .
3 10 i

5 -.

O L H L , L L . L L. L 7 k i 7 ,7. n a ,# .L_

8 10 l2 l4 l6 18 20 22 24
Weekofage

Figure 11.8. Empty body lipid means with standard errors from 10 to 22 wk of age for
Modified Open Front building (MOF) vs. Test Station building (TS) raised pigs.

58

0.35 -,
-0-BW -D-BF10 -fi-LMA *LRF

 

 

 

0.3
0.25 é .
0.2 I 1
,, A
"1: 0.15 4 /fr
0.1 S 1"
0.05 if j,
0 .
-0.05 .4 4 4 4.. . - 4» 4 .4- .4 «.7 , - 4
8 10 12 14 16 18 20 22 24
Weekof age

Figure 11.9. Heritability for measured traits from 10 to 22 wk of age. BW = body
weight, BF 10 = tenth rib backfat, LMA = longissimus muscle area, LRF = last rib backfat

59

0.35 2
0 3 3 -0-FFTOLN -Cl-TOFAT +EBPRO *EBLIPID
0.25 Z
0.2 ;
"a 0.15
0.1 4

. i I
0.05 J
0 3 r 2

-0.055— + _- -z_-.n,_ ,_ 3 2 _._._ ._ a r- _,
8 10 12 14 16 18 20 22 24

Week of age

 

 

 

 

Figure 11.10. Heritability for calculated traits from 10 to 22 wk of age. F FTOLN = fat-
free total lean, TOFAT = total fat tissue, EBPRO = empty body protein, EBLIPID =
empty body lipid

60

CHAPTER III

QTL MAPPING IN AN F2 DUROC x PIETRAIN RESOURCE POPULATION:
1. GROWTH TRAITS2

Abstract

Pigs from the F2 generation of a Duroc x Pietrain resource population were
evaluated to discover quantitative trait loci (QTL) affecting growth and composition
traits. Body weight and ultrasound estimates of tenth rib backfat, last rib backfat, and
longissimus muscle area were serially measured throughout development. Estimates of
fat-free total lean, total body fat, empty body protein, empty body lipid, and average daily
gain from 10 to 22 wk of age were calculated, and random regression analyses were
performed to evaluate phenotypes representing intercept and linear rates of increase in
these eight serial traits. A total of 510 F2 animals were genotyped for 124 microsatellite
markers evenly spaced across the entire genome. Data were analyzed with line cross
least squares regression interval mapping methods using sex and litter as fixed effects.
Significance thresholds of the F-statistic for additive, dominance, and imprinted QTL
were determined at chromosome- and genome-wise levels by permutation tests. A total
of 55 QTL for 22 of the 29 measured traits were found to be significant at the 5%
chromosome-wise level. Of these 55 QTL, 16 were significant at the 1% chromosome-
wise, 11 at the 5% genome-wise, and 10 at the 1% genome-wise significance thresholds.
A total of 33 QTL for 15 of the 16 animal random regression terms were found to be
significant at the 5% chromosome-wise level. Of these 33 QTL, 11 were significant at

the 1% chromosome-wise, 2 at the 5% genome-wise, and 2 at the 1% genome-wise

 

2 This research was financially supported by the Michigan State University Department
of Animal Science, the Michigan Agricultural Experiment Station, a Michigan Animal
Initiative Coalition Grant, and USDA-CSREES NR1 Award 2004-35604-14580.

61

significance thresholds. Putative QTL were discovered for tenth rib and last rib backfat
on SSC 6, body composition traits on SSC 9, backfat and lipid composition traits on SSC
11, tenth rib backfat and total body fat tissue on SSC 12, and linear regression of body
weight, longissimus muscle area, and tenth rib backfat on SSC 18. These results will
facilitate fine mapping efforts to identify genes controlling growth and body composition
of pigs that can be incorporated into marker-assisted selection programs to accelerate
genetic improvement in pig populations.
Introduction

Enhancement of production efficiency and improvement of product quality are
major concerns for producers of food animals. Selection of improved breeding animals is
essential for achieving this goal, and more information on prospective parents leads to
better selection decisions. Advances in genetic technologies have allowed for more
information to be collected on prospective parents which has included identification of
quantitative trait loci (QTL). The search for regions of the genome that control these
traits has led to the creation of resource populations and the discovery of putative QTL.

The Duroc and Pietrain breeds are utilized worldwide as sire breeds, and these
breeds differ in growth phenotypes. In general, Duroc pigs and their offspring have been
found to grow faster, but also have more backfat than other breeds (Kennedy et al., 1996;
Blanchard et al., 1999; Edwards et al., 2006). Quiniou and Noblet (1995) used Pietrain
boars in their study of equations to predict composition because of their propensity
towards leanness, and Edwards et al. (2006) reported slower, but leaner, growth in
Pietrain-sired pigs versus Duroc-sired pigs. While each of these breeds has been utilized

individually in resource populations with some rustic and some commercial breeds

62

(Duroc: Grindflek et al., 2001; Sato et al., 2003; Stearns et al., 2005) (Pietrain: Nezer et
al., 2002), no study has reported a resource population utilizing both Duroc and Pietrain
breeds. The objective of this study was to conduct a full genome scan using
microsatellite markers to search for QTL affecting grth traits in an F2 Duroc x Pietrain
resource population.
Materials and Methods

Population Development

A three-generation resource population was developed at Michigan State
University to study traits of growth, body composition, and meat quality. Semen from
four F0 Duroc sires from a closed unselected control population (Kuhlers et al., 2003) and
sixteen F0 Pietrain dams from a closed herd propagated the F 1 generation. All
grandparents were determined homozygous normal for the RYR-l gene by a DNA test
(Fujii et al., 1991). All animals were produced through artificial insemination at the
Michigan State University Swine Teaching and Research Farm. From F. progeny, 51
females and six males (sons of three F0 sires) were retained to produce the 1259 F2 pigs
born alive in 142 litters across 11 farrowing groups. Females were retained across
multiple parities to produce F2 progeny.
Animal Management

All pigs were managed similarly in farrowing and nursery stages with dams
placed into farrowing crates one week prior to farrowing. Baby pigs were processed
(individually identified by ear tag, given 0.5 ml penicillin and 1 ml iron dextran
subcutaneously, and tails clipped) at approximately one (1 of age. At seven (1 of age,

males not kept for breeding purposes were castrated. Pigs were weaned at 16-25 (mean

63

of 19.8) d of age and sorted into nursery pens by sex and weight. All diets fed were
Michigan State University standard swine farm diets that met or exceeded NRC (1998)
requirements for all nutrients at each production stage. At 10 wk of age, F2 pigs were
placed into one of two finishing facilities at the Michigan State University Swine
Teaching and Research Farm. Farrowing groups 1, 3, 5, 7, 9, and 11 (n = 521 pigs) were
placed into a Modified Open Front (MOF) building with two-thirds solid, one-third
slatted floors and wet-dry feeders. Four larger pens (2.03 m by 6.91 m) with two-space
feeders were targeted to contain 16 pigs per group. Four smaller pens (1.42 m by 6.91 m)
with one-space feeders were targeted to contain 12 pigs per group. Farrowing groups 2,
4, 6, 8, and 10 (n = 465 pigs) were placed into a test station (TS) facility with solid floors
bedded with straw or wood shavings with one-space dry feeders and cup drinkers. All 25
pens utilized (1.42 m by 4.93 m) were targeted to contain four pigs per group. Pigs in
either facility had ad libitum access to feed and water.
Phenotype and Genotype Collection

Live animal traits collected on F2 animals included BW at birth, weaning, 6, 10,
13, 16, 19, and 22 wk of age. Additionally, B-mode ultrasound (Pie Medical ZOOSLC,
Classic Medical Supply, Inc., Tequesta, FL) estimates of tenth rib backfat (BF10), last rib
backfat (LRF), and longissimus muscle area (LMA) were recorded at 10, 13, 16, 19, and
22 wk of age. The ADG from 10 to 22 wk of age and the number of days to reach 105 kg
were calculated from these BW measures. At each of these time points measures of fat-
free total lean (F FTOLN), total body fat tissue (TOFAT), empty body protein (EBPRO),
and empty body lipid (EBLIPID) were calculated using equations similar to those used

by Wagner et al. (1999).

64

Whole blood was collected from all F o, F 1, and F2 animals for DNA isolation.
White blood cells were separated and frozen for subsequent DNA extraction. A total of
206 dinucleotide microsatellite genetic markers were considered for genotyping, and 128
informative markers that would amplify were chosen. Markers were selected from the
published pig genome linkage map (http://www.marc.usda.gov/genome/swine/
swine.html, USDA, 2005b), and the initial set of markers was tested for informativeness
(markers segregating between F0 breeds and estimated to have three or more alleles)
using fluorescent primers distributed by the US Pig Genome Coordinator (supported by
the USDA-CREES through the National Research Support Project 8). These 128
markers were genotyped for 510 F2 animals, their parents, and grandparents at a
commercial laboratory (GeneSeek Inc., Lincoln, NE). These 510 animals were sampled
across all farrowing groups from 61 entire litters. Pigs represented all P; sires with at
least 100 grand progeny from each F0 sire. Fifteen of the 16 F0 dams had a son or
daughter as a parent that produced multiple litters of the selected F2 pigs with the
remaining F0 dam represented by a single F. daughter with one litter in this group.
Mendelian inheritance of alleles could not be verified within F2 animals for SW1349 on
SSC 9, SW2067 on SSC 10, SW1632 on SSC 11, and $0229 on SSC 12, and these
markers were removed from the data set. The remaining 124 markers were utilized for a
whole genome scan of approximately 20 cM spacing across the 18 autosomes and X
chromosome. Among these 124 markers several individual animal genotypes could not
be verified as they did not match expected Mendelian inheritance patterns from pedigree
analysis and were removed. Markers used in this genome scan are listed in Table 111.1

along with the number of alleles segregating for each marker, the relative positions of

65

each marker, and number of unverified genotypes for each marker. Genetic linkage maps
were constructed for each of the 18 autosomes and the X chromosome using Crimap
version 2.4 software (Green et al., 1990).
Random Regression

Serial BW and ultrasound estimates from 10 to 22 wk of age were used to
generate random regression equations to model pig BW, BF10, LMA, LRF, FFTOLN,
TOFAT, EBPRO, and EBLIPID on age at measurement for individual animals. Age at
measurement was modeled as week on-test, calculated as age in weeks minus nine, (i.e.
1, 4, 7, 10, and 13 as distinct covariate values used in the analysis). A random intercept
for each animal and a linear regression on age for each animal were included in each
model. Main effects and interactions that contributed significantly to the increase of the
log likelihood were kept in the model. Table 111.2 lists the polynomial order of week of
age and interactions used in models for these eight traits as determined by log likelihood

tests of significance. The following model was used:

Yijklmn = u + Zweeki‘b + SCXj + 2(sex*week¢)j + fink + Z(fin*week¢)k + grp(fin)k1 +
Pen(grp)1m + gn + Oln*Zi + eijklmn
where
Yijklmn = record on the nth pig within jth sex, kth finisher,
regressed on 11'" polynomial week i,
u = overall mean of trait,
week? = fixed regression coefficient for polynomial terms 4) (1-4) of week i,
SCXj = fixed effect of sex of animal j (Barrow or Gilt),
fink = fixed effect of finisher k (MOF or TS),
grp(fin)k1 = random effect of farrowing group 1 (1-11) nested within finisher
where grp ~N(0, Iogrpz),
pen(grp)1m = random effect of pen m (1-25) nested within farrowing group where
pen ~N(0, 16,...2),
gn = random intercept for animal n,
(1,, = random linear regression coefficient on age for animal n,
Z, = week on test as a covariate, and
eijklmn = random error.

lth group, and mth pen

66

The distributional assumptions on g = { g n} and a = { g n } were such that:

2
0 A0 A0
[g] ~ N [ :l, g g2“ , where a"; is the intercept genetic variance for the
a 0 Aaga Ada

individuals, 002, is the linear age by animal genetic variance, and 0' ga is the genetic

covariance between the intercept and linear term for each animal. To account for residual
variances across serial measurement and the relationship between time points, the

e = {e yum} was specified as normally distributed with a general (co)variance structure

calculated from the data and specified within and across weeks with five variance and 10
covariance terms.
QTL Analysis

The genetic linkage maps constructed for each chromosome were used in an F2
least squares interval mapping framework for QTL analysis (Haley et al., 1994). Briefly,
this method proceeded in two steps. In the first step, marker positions and genotypes
were used to calculate the probability an individual inherited 0, 1, or 2 alleles from each
of the two founder lines. Secondly, phenotypic data were regressed on coefficients
derived from these probabilities. The F2 analysis option of the QTL Express sofiware
(Seaton et al., 2002) was used to search for single QTL with additive, dominance, or
imprinting effects on the 18 autosomes and additive effects on the X chromosome. For
traits listed in Table 111.3 the model used included the fixed effects of sex of animal and
litter. For birth and 3 wk BW the model also included parity of the female as a fixed

effect and number born alive as a covariate. The model for 10 to 22 wk ADG also

67

included a covariate of 10 wk BW to account for differences in BW when pigs were
placed into the finishers.

Tests of the full model including additive, dominance, and imprinting effects
versus the reduced model without these effects were carried out to determine F-ratios at
intervals of 1 cM across the entire genome for the traits listed in Table 111.3. Analyses of
the animal random regression terms were also carried out with QTL Express, but only for
additive effects. During the course of estimation of the animal intercept and linear terms,
dominance effects were confounded with common environment (as stated in Hill, 1999).
Since these traits have already been adjusted for fixed and random effects in random
regression models, no further effects were included in their QTL analyses.

Additive effects were defined as half the difference between Pietrain and Duroc
genotypes at the QTL. Dominance effects were defined as the difference between the
heterozygote genotype and the average of the homozygote genotypes. Imprinting effects
were defined as the difference between the heterozygous genotypes when the Duroc
allele was inherited from parents of opposite sex. Haley et a1. (1994) and Knott et al.
(1998) describe further details on the parameterization of these effects. Significance
thresholds of 5% and 1% at the chromosome-wise and 5% and 1% at the genome-wise
levels were determined through the use of permutation tests (Churchill and Doerge, 1994)
in QTL Express. These thresholds were determined for genome-wise levels and for each
chromosome based upon appropriate models for each chromosome and 30,000
permutations. For each QTL determined to be significant at the 5% genome-wise level,
confidence intervals of the QTL position were determined using a bootstrap method with

5,000 permutations in QTL Express (Visscher et al., 1996).

68

Results and Discussion
QTL Analysis

Trait means and standard deviations are listed in Table 111.3, and are similar to
those measured in other resource populations for similar traits (e. g. Malek et al., 2001b;
Stearns et al., 2005). Table 111.1 contains the list of the 124 markers used in this analysis
along with their linkage map positions on each chromosome, the number of alleles
segregating in this population, and the number of missing genotypes for F2 animals. The
average distance between markers was 24 cM, although the distance between some
markers was larger due to the difficulty in identifying informative markers in those
regions. Overall, our map generally agreed with or was slightly longer than the USDA
map (Rohrer et al., 1996; http://www.marc.usda.gov/genome/swine/swine.htrnl, USDA,
2005b) except for two marker pairs that switched order in our analysis. On SSC 9, SW21
and SW983 inverted order and were 13.2 cM apart on our map, whereas they were 11.2
cM apart on the USDA map. On SSC X, SW1608 and SW2059 inverted order and were
23.0 cM apart on our map and 17.4 cM apart on the USDA map. The longer length of
our map as compared to the USDA map was not unexpected since the marker density of
our map was much lower.

The permutation test results to determine significance thresholds for the F
statistics for models testing additive, dominance, and imprinting effects ranged from 3.44
to 4.35 across chromosomes for the 5% chromosome-wise thresholds and from 4.71 to
5.64 for the 1% chromosome-wise thresholds. Genome-wise threshold levels were 6.34
and 7.63 for 5% and 1% thresholds, respectively. Significance threshold levels for the

additive effect only models ranged from 5.39 to 7.64 and 8.68 to 10.99 for 5% and 1%

69

chromosome-wise levels, respectively, and they were 12.86 and 16.05 for 5% and 1%
genome-wise levels, respectively.

Estimates of positions and F -ratios of QTL significant at the 5% chromosome-
wise level for traits listed in Table 111.3 are listed in Table 111.4. These results were
derived from single QTL on each chromosome models. Information in the table is sorted
by chromosome and position within each chromosome. Additionally, the additive,
dominance, and imprinting effect at each QTL along with their standard errors are listed
in Table 111.4. A total of 55 QTL for traits in Table 111.3 were found to be significant at
5% chromosome-wise levels. Of these 55 QTL, 16 were significant at the 1%
chromosome-wise, 11 at the 5% genome-wise, and 10 at the 1% genome-wise
significance thresholds. No significant QTL were detected in this population for 3, 6, 10,
13, or 16 wk BW, 13 wk LMA, or 22 wk FFTOLN.

For each of the random regression analyses conducted, the animal-specific
intercept and linear random regression terms were tested for QTL. Estimates of QTL
positions and F-ratios for animal random regression terms significant at the 5%
chromosome-wise level are listed in Table 111.5. These results are for the most significant
position for each QTL as only single QTL on each chromosome models were analyzed.
Additionally, the additive effect at each QTL along with the standard error is listed in
Table 111.5. A total of 33 QTL for 15 of the 16 animal random regression terms were
found to be significant at 5% chromosome-wise levels. Of these 33 QTL, 11 were
significant at the 1% chromosome-wise, 2 at the 5% genome-wise, and 2 at the 1%
genome-wise significance thresholds. Only linear regression on FFTOLN did not have a

significant QTL that was detected in this population.

70

All chromosomes, except 2, 4, 12, 15, and 17, contained at least one QTL for
these traits related to grth and body composition. Although these two breeds have
both been selected for terminal sire breeding programs for several years, many important
QTL causing phenotypic differences between the breeds still exist and are possible
candidates to explore more thoroughly. Some QTL had breed effects in the same
direction as for traits in Duroc- and Pietrain-sired animals reported in Edwards et al.
(2006), but a few cryptic alleles exist that act in an opposite direction to the general trend
for the overall breed effects.

Body Weight

Four QTL influenced BW grth at different ages at the 5% chromosome-wise
level (Table 111.4). A QTL for birth weight occurred on SSC 5 with an imprinting effect
that indicated paternal expression. This QTL has not been reported before in other pig
resource populations (Hu et al., 2005). No QTL for growth between 3 and 16 wk of age
were detected in this analysis. However, a QTL on SSC 16 for BW was significant for
grth at 19 and 22 wk of age and influenced 10 to 22 wk ADG. This was a cryptic
allele, as the Duroc alleles caused lower BW and 10 to 22 wk ADG, which contrasted
with results from Edwards et al. (2006), who reported faster growth from Duroc-sired
versus Pietrain-sired progeny. The 1 cM F-ratio tests for these QTL are illustrated for
SSC 16 in Figure 111.] with the F-ratio plotted versus relative marker position. Another
QTL affected BW growth on SSC 7 and showed an imprinting inheritance pattern. Other
studies have reported a QTL for average daily gain on SSC 7, but for differing time
periods of development than in this study (Knott et al., 1998; Nezer et al., 2002) or in

different positions on the chromosome (Bidanel et al., 2001). Two QTL were found that

71

influenced the animal random regression intercept of BW (Table 111.5). One QTL was on
SSC 5 with the Pietrain allele increasing the intercept and another QTL was on SSC 8
with the Duroc allele increasing the intercept. Two additional QTL influencing the linear
rate of BW gain from 10 to 22 wk of age were found. One QTL on SSC 16 had an
additive effect with Pietrain alleles that increased the rate of BW gain. The other, on SSC
18, suggested that Duroc alleles increased the rate of BW gain. Neither of these linear
rate QTL, or traits with similar definitions, was reported in a summary of other pig QTL
studies (Hu et al., 2005). In addition, a QTL for days to 105 kg was discovered on SSC
9. Although ADG from 10 to 22 wk of age and days to 105 kg are similar traits,
differences in grth from birth to 10 wk of age that are included in the days to 105 kg
calculation may have caused QTL for these traits to be located on different chromosomes.
Backfat

Several regions of the genome contributed to fat tissue phenotypes at many ages
of development. When considering the associated traits of BF 10 and LRF, a region on
SSC 6 influenced both of these traits at all measured time points of 10, 13, 16, 19, and 22
wk of age with all of them significant at the 1% genome-wise level (Table 111.4). The
estimate of the peak position of this QTL ranged from 134 to 143 cM from the most
proximal marker (80099), so this may be the same pleiotropic QTL for all of these related
traits. Figure 111.2 demonstrates the F-ratio curves plotted versus relative marker
positions on SSC 6 and illustrates the similar shapes of F-ratios for 22 wk BF10 and 22
wk LRF. The estimate of this QTL in this analysis indicated that Duroc alleles
contributed to larger measures of BF 10 and LRF. A QTL in the same region affected the

animal linear random regression terms of BF 10 and LRF (Table 111.5). Again, Duroc

72

alleles contributed to a faster rate of backfat deposition at BF 10 and LRF. While most
backfat QTL on SSC 6 found in other populations occur more proximal than what was
reported here, Malek et al. (2001a) and Ovilo et al. (2002b) also reported putative QTL
for backfat in the same region as those reported in this study. Two other regions were
significant for multiple backfat phenotypes. One of these was on SSC 11, which
influenced BF10 at 19 and 22 wk of age and LRF at 19 wk of age (Figure 111.3). The
only other study which reported a backfat QTL on SSC 11 was Milan et al. (2002), but
their QTL was more proximal to the one reported here. Another region, on SSC 16, that
influenced BF10 at 19 wk of age and LRF at 22 wk of age is at 64 cM with an additive
effect characterized by Duroc alleles that caused less BF 10 and LRF (Figure 111.1). Only
one QTL, for small intestine length (Hu et al., 2005), has been identified on SSC 16, but
several of the traits in this study were influenced by QTL on SSC 16. A QTL on SSC 18
at 54 cM was significantly related to BF10 at 10 wk of age and LRF at 16 wk of age.
Additional locations of QTL affecting BF10 and LRF at different time points from 10 to
22 wk of age occur on SSC 3, 8, and 9 (Table 111.4). The random regression linear
animal terms for BF10 and LRF were also affected by a QTL on SSC 16. These effects
on SSC 16 were estimated to increase backfat when Pietrain alleles were present, which
would be considered a cryptic allele in comparison to results of Edwards et al. (2006).
Other QTL affecting random regression animal intercept and linear terms were found on
SSC10,11,18, and X.
Longissimus Muscle Area

Numerous QTL controlling LMA in this population were detected with most of

them significant only at the 5% chromosome-wise level. One chromosome with three

73

regions significant for LMA QTL was SSC 6 (Tables 111.4 and 111.5). One region that
was significant for 10 and 19 wk of age LMA had a dominance effect where Duroc
alleles contributed to a larger LMA. Another region was significant for 16 wk of age
LMA with a dominance effect where Pietrain alleles contributed to a larger LMA.
Lastly, a third region was significant for LMA random regression animal intercept where
Duroc alleles contributed to a larger LMA. A QTL influencing lean percentage was
reported in the same region in two other populations containing Duroc animals (Grindflek
et al., 2001; Szyda et al., 2003; Stearns et al., 2005). Additionally, SSC 1 had a QTL
region that affected 13 wk of age LMA where Pietrain alleles had additive effects that
contributed to larger LMA. The SSC 4 had a QTL region that affected 22 wk of age
LMA additively where Pietrain alleles contributed to a larger LMA. A few populations
have identified a region that influences LMA just distal to the findings reported here
(Pe'rez-Enciso et al., 2000; Malek et al., 2001a; Varona et al., 2002), while one used
AF LP markers and identified a QTL in the same region as reported here for LMA
(Wimmers et al., 2002). A region at the proximal end of SSC 18 affected the linear
random regression of LMA and was significant at the 1% genome-wise level (Figure
111.4). For this QTL, Duroc alleles caused a faster rate of increase in LMA.
Composition Traits

Estimates of body composition (F F TOLN, TOFAT, EBPRO, and EBLIPID) at 22
wk of age as well as random regression analyses from serial data of 10 to 22 wk of age
were tested for QTL. The use of random regression terms as traits for QTL analysis has
not been previously reported in other swine resource populations. Two QTL with

dominance effects for FFTOLN at 22 wk of age were discovered, with one on SSC 6 and

74

another on SSC 9 (Table 111.4). The Duroc alleles increased the amount of F F TOLN at
22 wk of age. Two QTL also affected the random regression intercept of F FTOLN with
one QTL on SSC 5 and another on SSC X (Table 111.5). In this case the Pietrain alleles
increased the F FTOLN intercept. Results for QTL for TOF AT indicated three QTL with
one each on SSC 6, 9, and 16 affecting TOFAT at 22 wk of age (Table 111.4). The
random regression animal intercept for TOFAT was influenced by QTL on SSC 5, 8, l4,
and 18 while random regression animal linear TOFAT was influenced by QTL on SSC 8
and 16 (Table 111.5). These QTL for TOFAT were all significant at the 5% chromosome-
wise level. Putative QTL with dominance effects were found on SSC 6 and 9 where
Duroc alleles contributed to an increase in EBPRO. Both of the protein composition
traits of FFTOLN and EBPRO had similar patterns for F-ratio curves on SSC 6 as
indicated in Figure 111.2. Random regression animal intercept for EBPRO was influenced
by QTL on SSC 5 and X with a decrease in EBPRO intercept by Duroc alleles. For
random regression linear EBPRO, QTL occurred on SSC 9 where Duroc alleles increased
the rate of EBPRO deposition. Another QTL was discovered on SSC X where Pietrain
alleles increased the rate of EBPRO deposition. This QTL also affected FF TOLN
intercept in the same region (Figure 111.5). Effects of QTL on SSC 6 and 11 were evident
for EBLIPID at 22 wk of age, as well as in the same region of SSC 9 that contributed to
differences in EBPRO (Table 111.4). A similar region on SSC X to the one that
influenced random regression EBPRO intercept also contributed to differences in random
regression EBLIPID intercept (Table 111.5). Additionally a region on SSC 6 also
contributed to an increase in random regression EBLIPID intercept when Duroc alleles

were present. The F-ratios for TOF AT and EBLIPID on SSC 6 were generally

75

concurrent (Figure 111.2), which indicated that the same chromosomal regions influenced
both traits. The random regression animal linear EBLIPID term associated with QTL
regions on SSC 18 and X was significant at the 5% chromosome-wise level for QTL.
Traits related to body composition of FFTOLN, TOFAT, EBPRO, and EBLIPID were
influenced by a QTL in between the two most proximal markers on SSC 9 (SW21 and
SW983) used in this study (Figure 111.6). These traits are highly related, and appeared to
be influenced by similar regions of the genome. No other studies have reported QTL for
these or similar traits on SSC 9 (Hu et al., 2005).
Confidence Intervals

For QTL that were significant at the 5% genome-wise level, 95% confidence
intervals were estimated using bootstrapping with resampling in QTL Express (Seaton et
al., 2002). These confidence intervals are listed in Table 111.6. Confidence intervals were
not calculated for QTL significant at the 5 or 1% chromosome-wise level since
preliminary analyses indicated that many of these intervals tended to encompass the
entire chromosome on which they reside.

Implications

Discovery of QTL in an F2 Duroc-Pietrain resource population revealed many
regions of the genome that potentially influence a portion of the phenotypic differences in
growth and composition traits. Not only were QTL found for measured traits, but also
for the animal terms from random regression analysis. Employing both measured and
composite traits allowed unique characterization of growth and composition phenotypes
from this population. The utilization of the Duroc and Pietrain breeds in this population

will allow for identified. QTL to be incorporated quickly into breeding schemes as these

76

two populations are already a major part of commercial pig production. While further
analyses and refining of QTL positions are in progress, results of this study are useful in
discovering potential QTL regions and leading to further understanding the role certain
regions of the genome have in determining phenotype differences among potential

parents for selection of breeding animals.

77

Table lll.1. Markers used in the QTL analysis, map positions determined for the F2 Duroc x
Pietrain resource population, number of alleles segregating for each marker, and number of
missing genotypes for each marker. Distances (in Kosambi cM) are relative to position of
first marker on each chromosome in this population.

 

 

Marker Chr Position Alleles DGa
SW1514 1 0 5 10
SW1515 1 21 .1 5 13
$0008 1 49.7 5 20
$0331 1 77.5 9 7
SW974 1 108.9 8 15
$0056 1 179.7 5 15
SW1301 1 235.1 5 8
SWR2516 2 0 3 4
SW240 2 41 .0 4 13
$0170 2 53.7 2 5
SW1026 2 64.8 8 10
$0370 2 93.4 6 13
SW1844 2 103.2 5 1
$0378 2 1 10.1 4 9
$0036 2 142.0 6 7
SW274 3 0 6 1
SW2021 3 22.7 1 1 8
$0206 3 68.7 4 3
ACTG2 3 86.4 7 9
SW2047 3 101 .3 5 17
SW2408 3 124.3 4 2
$0002 3 132.4 5 5
SW1327 3 141 .2 9 5
SW2532 3 159.6 6 8
SW2404 4 0 7 8
$0301 4 29.2 5 26
SW871 4 54.1 6 25
SW2454 4 61 .8 4 6
$0107 4 73.8 6 5
$0214 4 88.0 5 10
$0097 4 131 .3 3 1 1
SW413 5 0 3 36
ACR 5 1 1.5 5 2
SWR453 5 53.4 3 7
SW2 5 81 .5 6 9
$0005 5 108. 9 10 12
$0018 5 127.4 6 9
SW995 5 147.6 5 5
SW378 5 159.7 3 3

 

Marker Chr Position Alleles DG‘
$0099 6 0 5 42
SW2406 6 22.4 5 6
SW2525 6 50.7 4 10
$0087 6 81.4 4 10
$0220 6 98.2 2 13
SW122 6 103.9 5 9
SW1881 6 135.8 4 26
SW322 6 164.8 7 3
SW2419 6 181 .1 7 12
$0025 7 0 5 6
$0064 7 28.1 7 17
SW1369 7 48.0 7 14
SW859 7 92.5 4 41
$01 15 7 135.8 7 7
SWR773 7 151.6 3 10
$0101 7 164.0 4 9
SW764 7 186.7 3 5
SW2410 8 0 4 0
SW905 8 22.9 3 3
SWR1 101 8 55.0 7 44
$0017 8 95.1 5 10
SW2160 8 1 10. 7 3 9
SW1085 8 124.4 3 3
$0178 8 165.5 4 21
SW21 9 0 4 9
SW983 9 13.2 5 8
SW91 1 9 43.1 6 8
SW2401 9 63.7 3 4
SW539 9 72.4 2 7
SW989 9 97.0 6 9
SW21 16 9 127.4 4 14
SWR136 10 0 7 26
SW249 10 20.2 4 9
SWC19 10 44.3 5 17
SW1041 10 56.8 2 10
SW920 10 79.4 2 0
$0391 1 1 0 4 9
$0071 1 1 53.0 4 6
$0230 1 1 65.2 5 11
SW66 1 1 1 19.0 7 20

 

806 = number of unverified genotypes deleted for each marker (out of 510 F2 pigs)

78

Table Illflcont’d).
Marker Chr Position Alleles DG‘ Marker Chr Position Alleles DGa

 

 

SW2490 12 0 6 10 $0111 16 0 7 27
SW957 12 31.2 6 12 SW419 16 30.9 5 13
SW874 12 47.2 4 17 SW25‘17 16 64.8 3 14
80090 12 61.0 5 11 $0061 16 99.5 4 18
SW2180 12 94.1 6 17 SWR1004 17 0 5 25
$0219 13 0 3 8 SW2441 17 22.7 7 8
SWR1941 13 13.7 7 12 SW1031 17 42.9 2 7
SW344 13 40.9 7 7 SW2427 17 94.9 6 17
SWR1008 13 54.2 7 28 SW1023 18 0 4 3
$0068 13 65.1 5 13 SW1984 18 43.1 3 0
SW398 13 85.8 5 12 $0062 18 54.9 4 7
SW2440 1 3 103.2 5 5 SW949 X 0 6 13
$0215 13 122.6 3 3 SW980 X 14.8 5 9
SW857 14 0 5 6 SW2534 X 37.3 4 10
SW510 14 26.6 3 6 SW2126 X 47.7 5 4
SW21O 14 45.9 3 6 SW2470 X 56.1 4 9
SW886 14 64.7 7 13 SW1426 X 89.2 3 3
SW55 14 85.2 4 10 SW2059 X 146.6 5 18
SW1557 14 95.6 3 25 SW1608 X 169.6 4 11
SW027 14 117.6 3 2

SW1204 15 0 3 45

$0148 15 24.7 5 11

$0088 15 54.8 2 9

SW1683 15 64.5 4 1

SW1983 15 82.2 6 11

SW1119 15 96.0 5 9

 

‘06 = number of unverified genotypes deleted for each marker (out of 510 F2 pigs)

79

Table Ill.2. Order of polynomial on week of age terms and other
significant terms for random regression analyses of serial mwth data.

 

 

 

Significant terma
Week Sex* Finisher"

Trait Of age weekz week2
Body weight, kg 4 - X
Tenth rib backfat, mm 2 X -
Longissimus muscle area, cm2 3 - -
Last rib backfat, mm 4 X -
Fat-free total lean, kg 4 - X
Total body fat tissue, kg 4 X -
Empty body protein, kg 2 - X
Empty body lipid, kg 4 - -

 

3X indicates used in model, - indicates not used in model

80

Table lll.3. Number of records, means, and standard
deviations for growth traits measured.

 

Trait N Mean SD
Birth weight, kg 510 1.50 0.32
3 wk weight, kg 510 5.61 1.46
6 wk weight, kg 510 11.94 2.78
10 wk weight, kg 510 26.38 4.91
10 wk tenth rib backfat, mm 510 7.94 1.80
10 wk longissimus muscle area, cm2 510 11.63 2.60
10 wk last rib backfat, mm 510 6.09 1.09
13 wk weight, kg 510 41.56 6.55
13 wk tenth rib backfat, mm 510 9.71 2.66
13 wk longissimus muscle area, cm2 510 17.03 3.33
13 wk last rib backfat, mm 510 7.13 1.35
16 wk weight, kg 510 62.01 8.21
16 wk tenth rib backfat, mm 510 12.34 3.44
16 wk longissimus muscle area, cm2 510 25.03 3.92
16 wk last rib backfat, mm 510 9.53 2.29
19 wk weight, kg 510 80.85 10.15
19 wk tenth rib backfat, mm 510 16.03 5.12
19 wk longissimus muscle area, cm2 510 31.72 4.51
19 wk last rib backfat, mm 510 11.89 3.38
22 wk weight, kg 510 100.00 10.93
22 wk tenth rib backfat, mm 510 19.97 6.30
22 wk longissimus muscle area, cm2 510 37.47 5.04
22 wk last rib backfat, mm 510 14.43 4.13
10-22 wk ADG, g/d 510 880.21 107.99
Days to 105 kg 510 157.37 14.54
22 wk total body fat tissue, kg 510 25.01 7.00
22 wk fat-free total lean tissue, kg 510 38.45 4.41
22 wk empty body protein, kg 510 15.01 1.67
22 wk empty body lipid, kg 510 21.97 4.27

 

81

Table Ill.4. Position and significance levels of single point QTL significant at 5% chromosome-
wise level with additive, dominance, and ian'ntigg effects and standard errors of the QTL.

 

 

 

Pos. Additive Dominance Imprinm

Chr Trait (cM) Effect SE Effect SE Effect SE F-ratio'
1 13 wk longissimus muscle area, cm2 87 -0.66 0.22 0.56 0.36 0.33 0.23 4.55
3 10wktenth rib backfat, mm 45 -0.03 0.15 -0.06 0.32 0.58 0.15 5.34

4 22 wk longissimus muscle area, cm2 45 -1.30 0.36 -1.02 0.59 -0.26 0.34 5.20

5 Birth weight, kg 58 -0.05 0.02 -0.04 0.04 0.07 0.03 4.44

6 10 wk longissimus muscle area, cm2 104 0.11 0.15 0.75 0.21 -0.06 0.14 4.48
6 19 wk longissimus muscle area, cm2 104 -o.05 0.29 1.56 0.42 -o.41 0.29 4.98
6 22 wk fat-free total lean, kg 104 -0.74 0.29 1.23 0.42 -0.54 0.29 5.97"
6 22 wk empty body protein, kg 104 -0.31 0.11 0.35 0.16 -0.18 0.11 5.11

6 10 wk last rib backfat, mm 134 0.40 0.06 -0.37 0.10 -0.07 0.0717.59***
6 10 wk tenth rib backfat, mm 137 0.67 0.10 -0.76 0.14 —0.04 0.112577“
6 19 wk tenth rib backfat, mm 138 1.81 0.25 -1.80 0.38 -0.25 0.28 24.97”“
6 19 wk last rib backfat, mm 138 1.36 0.18 -1.21 0.27 -0.25 0.20 26.15"”
6 13 wk last rib backfat, mm 139 0.55 0.08 -0.50 0.13 -0.04 0.09 20.07“"
6 22 wk last rib backfat, mm 139 1.44 0.23 -1.11 0.36 -0.10 02516.45“
6 22 wk total body fat tissue, kg 139 1.32 0.42 -0.87 0.66 -1.06 0.47 5.48

6 13 wk tenth rib backfat, mm 140 0.97 0.15 -0.93 0.23 -0.18 0.16 20.36""
6 16wklast rib backfat, mm 140 0.85 0.14 -0.92 0.22 -0.16 0.1519.31***
6 16 wk tenth rib backfat, mm 141 1.37 0.19 -1.09 0.30 0.09 0.2122.09***
6 22 wk tenth rib backfat, mm 142 2.09 0.35 -1.98 0.58 -0.03 0.391517”
6 22 wk empty body lipid, kg 142 0.91 0.26 -1.04 0.43 -0.43 0.29 6.48”
6 16 wk longissimus muscle area, cm2 150 0.30 0.27 -1.45 0.48 -0.57 0.29 4.87
7 10—22 wk ADG, g/d 148 15.76 6.53 -3.07 11.3 20.08 6.73 4.75

8 13 wk tenth rib backfat, mm 115 0.51 0.15 -0.10 0.25 -0.33 0.16 5.20

9 22wkfat-free total lean, kg 4 -0.16 0.30 1.43 0.46 0.63 0.33 4.26

9 Days t0105 kg 5 0.42 0.96 -5.02 1.47 -2.22 1.06 5.08

9 22 wk empty body lipid, kg 6 0.05 0.26 1.56 0.40 0.80 0.29 7.21“
9 22 wk empty body protein, kg 6 -0.10 0.11 0.58 0.17 0.16 0.13 4.30

9 22 wk total body fat tissue, kg 8 -0.20 0.44 2.40 0.66 0.84 0.48 5.26
9 22 wk tenth rib backfat, mm 86 -1.02 0.43 —2.46 0.82 0.44 0.44 4.67
11 10 wk last rib backfat, mm 0 0.04 0.10 0.51 0.22 -0.24 0.10 3.97
11 19 wk weight, kg 65 -1.27 0.62 -2.08 0.87 1.06 0.63 4.36
11 19 wk last rib backfat, mm 65 -0.21 0.19 -1.10 0.27 0.16 0.20 6.34”
11 19 wk tenth rib backfat, mm 68 -0.64 0.28 -1.30 0.42 -0.01 0.29 5.25‘
11 22 wk tenth rib backfat, mm 75 -0.34 0.42 -2.33 0.73 0.29 0.43 3.80
11 22 wk empty body lipid, kg 75 -0.46 0.30 -1.87 0.52 0.21 0.31 530*
16 22 wk last rib backfat, mm 64 -1.10 0.30 -0.17 0.50 0.47 0.30 535*
16 19 wk tenth rib backfat, mm 65 -1.42 0.33 -0.26 0.55 0.04 0.34 6.40"
16 22 wk total body fat tissue, kg 65 -2.06 0.52 -1.24 0.86 0.89 0.53 7.08“
16 22 wk weight, kg 67 -2.71 0.88 -1.59 1.51 0.98 0.914.02

16 19 wk weight, kg 69 -2.96 0.80 -0.18 1.42 1.17 0.83 4.97
16 10-22 wk ADG, g/d 97 -14.67 5.98 -15.61 8.83 16.50 6.04 545*
18 10wktenth rib backfat, mm 54 0.30 0.11 0.30 0.16 0.18 0.12 4.14
18 16 wk last rib backfat, mm 54 0.40 0.14 0.42 0.21 0.06 0.16 3.75

 

“ Significant at * = 1% chromosome-wise, ** = 5% genome-wise, *** = 1% genome-wise levels

82

Table lll.5. Position and significance levels of random regression QTL significant at 5%
chromosome-wise level with additive effects and standard errors of QTL at those positions.

 

 

Location Additive
Chr Trait (cM) Effect SE F-ratio‘
5 Total fat tissue intercept, kg 57 -0.0078 0.0027 8.48
5 Body weight intercept, kg 120 -0.0217 0.0070 9.52
5 Fat-free total lean tissue intercept, kg 123 -0.0075 0.0022 11.56"
5 Empty body protein intercept, kg 126 -0.0024 0.0008 8.12
6 Last rib backfat intercept, mm 36 0.0059 0.0017 12.27‘
6 Empty body lipid intercept, kg 104 0.0114 0.0036 9.91
6 Last rib backfat linear, mka 140 0.0052 0.0018 8.07
6 Tenth rib backfat linear, mm/wk 145 0.0211 0.0061 12.08“
6 Longissimus muscle area intercept, cm2 181 0.0138 0.0048 8.35
8 Total fat tissue linear, kg/wk 55 0.0093 0.0033 8.11
8 Body weight intercept, kg 114 0.0278 0.0067 17.10““
8 Total fat tissue intercept, kg 116 0.0068 0.0024 7.95
8 Last rib backfat intercept, mm 117 0.0044 0.0015 8.41
9 Empty body protein linear, kg/wk 63 0.0036 0.0012 9.53
10 Tenth rib backfat intercept, mm 26 0.0125 0.0042 8.75
11 Last rib backfat linear, mm/wk 119 0.0077 0.0024 10.14'
14 Last rib backfat intercept, mm 109 -0.0053 0.0018 8.99
14 Total fat tissue intercept, kg 117 -0.0078 0.0028 7.94
16 Tenth rib backfat linear, mm/wk 0 -0.0171 0.0057 9.14
16 Total fat tissue linear, kg/wk 0 —0.0090 0.0032 8.12
16 Body weight linear, kg 0 -0.0298 0.0094 10.03“
16 Last rib backfat linear, mm/wk 3 -0.0055 0.0019 8.04
18 Longissimus muscle area linear, cm2/wk 0 0.0517 0.0128 16.30“"
18 Empty body lipid linear, kg/wk 0 0.0166 0.0069 5.87
18 Body weight linear, kg 0 0.0455 0.0138 1080*
18 Tenth rib backfat linear, mka 7 0.0295 0.0087 11.45‘
18 Total fat tissue intercept, kg 49 0.0064 0.0024 7.17
X Fat-free total lean tissue intercept, kg 35 -0.0069 0.0021 10.71'
X Empty body protein linear, kng 37 -0.0034 0.0010 10.86"
X Empty body protein intercept, kg 160 -0.0026 0.0009 7.39
X Last rib backfat linear, mm/wk 169 -0.0061 0.0019 9.81
X Empty body lipid intercept, kg 169 -0.0108 0.0040 7.48
X Empty body lipid linear, kg/wk 169 -0.0130 0.0050 6.84

 

’ Significant at * = 1% chromosome-wise, ** = 5% genome-wise, '** = 1% genome-wise levels

83

Table Ill.6. Position and 95% confidence interval lower and upper limits of
HLOWth QTL significant at the 5% genome-wise level.

Position Lower limit Upper limit

 

Chr Trait (cm (CM) (cM)
6 10 wk last rib backfat, mm 134 122 144
6 10 wk tenth rib backfat, mm 137 129 143
6 19 wk tenth rib backfat, mm 138 127 144
6 19 wk last rib backfat, mm 138 127 144
6 13 wk last rib backfat, mm 139 125 145
6 22 wk last rib backfat, mm 139 124 146
6 13 wk tenth rib backfat, mm 140 127 146
6 16 wk last rib backfat, mm 140 126 147
6 16 wk tenth rib backfat, mm 141 128 147
6 22 wk tenth rib backfat, mm 142 124 149
6 22 wk empty body lipid, kg 142 23 181
8 Body weight intercept, kg 114 2 139
9 22 wk empty body lipid, kg 6 0 104

16 19 wk tenth rib backfat, mm 65 0 79.5

16 22 wk total fat tissue, kg 65 0 80

18 Longissimus muscle area linear, cmzlwk 0 0 10

 

84

o 19 wk weight, kg 0 19 wk tenth rib backfat, mm A 22 wk weight, kg
x 22 wk last rib fat, mm x 22 wk total body fat tissue, kg

8 : macaw-rise _

5% genome-wise

 
  
   
    

T 1% chromosome-wise

5% chromosome-wise

F-ratio
&

11111

‘A
.— ,...o'eo¢‘;‘ a \
Q A‘ u 6
1 ....00 ‘~“"\ A
k A.‘

0 1 y _ -m + if .._. _ -. , ,7 7, , , . . . d _._._
5— 20 40 60 80 100
Position, cM

 

Figure 111.1. F-ratio plots versus relative positions on SSC 16. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

85

16 i
14 9
12 f

10;

F-nlfio
co

_ 5% gengrrre-wise

D 22 wk tenth rib backfat, mm x 22 wk last rib backfat, mm

x 22 wk total body fat tissue, kg 0 22 wk fat-free total lean, kg
+ 22 wk empty body lipid, kg - 22 wk empty body protein, kg

   
  

1% gegmejvise

1% caromghre-wise,,

5% chmmsomryvjse

Position, cM

Figure 111.2. F-ratio plots versus relative positions on SSC 6. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

86

o 19 wk weight, kg
x 22 wk tenth rib backfat, ran at 22 wk empty body lipid, kg

8 1%figengme-lvisefi , 7

i 5% genome-wise
(5 : ‘ A
1% chronpsomejvise ~ , . ..

; 5% chromosome-wise

    

F -ratio
A

\
\
\‘\

0 ”*— * ‘ fl * ' 7 r
E 20 40 60 80
Position,cM

 

D 19wk tenth rib backfat, run A 19wk last rib backfat, mm

.......

AA
“{‘Aaaaa

100 0

Figure [11.3. F -ratio plots versus relative positions on SSC 11. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

87

18 j 0 body weight linear El tenth rib backfat linear

A longissimus muscle area linear uierrpty body lipid linear _ 1% genome-wise

16 913A];
. AA
A
14 - AA .
, A A 5% genome-Wise
12 : QDDDDDDAAB
' DD DD
9800 nBUD
. o A on
O 10 00 DD .
'5 f 00 AA Duo 1% chromosome-Wise
a: = 0004.13 Jone
u. 8 o A no
00 AA 000
00 A 0000 0000000
0 A Do on .
6 o A Do 0 0/ h ,
xxxxxXXX °o 7AA A , 0:1 q, 5 oc romosome wrse
' xxx 000 KA
4 ; XxxXXx 0° AAAA
. Xxxxxxx 00°00 AAAAA
2 1 XXXXXX§§°°00 AAAAAA oo
. Xxxxflflfigggggfififififlfigxx
0‘ . A, . , _. . . _ ”2- ,2", 1 2.. ;+ - _‘A-__. _ .4
O 10 20 30 40 50 60
Position, cM

Figure [11.4. F-ratio plots versus relative positions on SSC 18. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

88

18 L, A fat-fi'ee total lean int 0 empty body protein linear
- 1% genome-wise

wise

l4 7 7 75%vgeno‘me-wise _

.__ 1% chm-gsom-wise

 

.2
H
E
L; .
.x'm , ‘ 5% chromosome-Wise
"'I ,7 2* luau" ” w ‘1 _ u; q
I rl'lr'lr.' Illio,”"’lu ll”, / “"0“
Irl|r ‘IA‘
”fur A
I].
ninth..." "I
‘ ' III ”I" I
2 t " r”Fir
I
O f _ M .2 . A _ _ , A- “A, ._,,A , A . k ”A, .2 .12- -_. W _
l l l l l l l I
0 20 40 6O 80 100 120 140 160

Position, cM

Figure 111.5. F-ratio plots versus relative positions on SSC X. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

89

o 22 wk total body fat tissue, kg 0 22 wk fat-fi'ee total lean, kg
A 22 wk empty body protein, kg x 22 wk empty body lipid, kg

 

8 I 7 i , , , i 7 ,7 17%kgenorne-wise 7,
7 f M;
2%: ’§*” 7, 5% genoArne-wise__
6
3.7 , ,i 7 i H 7 718/3 chromosome-wise
O 5 : We
a ‘ ., R 5% chromosome-wise
g 4 , "*§g7<*** if if- HAWWH , ._2_
h.

 

 

 

 

 

Position, cM

Figure [11.6. F-ratio plots versus relative positions on SSC 9. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

90

CHAPTER IV

QTL MAPPING IN AN F2 DUROC x PIETRAIN RESOURCE POPULATION:
II. CARCASS AND MEAT QUALITY TRAITS3

Abstract

Pigs from the F2 generation of a Duroc x Pietrain resource population were
evaluated to discover quantitative trait loci (QTL) affecting carcass composition and meat
quality traits. Carcass composition phenotypes included primal cut weights, skeletal
characteristics, backfat thickness, muscle pH, and temperature. Meat quality data
collected on boneless longissimus muscle chops included objective and subjective color
information, marbling and firmness scores, and drip loss. Additionally, chops were
analyzed for moisture, protein, and fat composition as well as cook yield and Warner-
Bratzler shear force measurements. Palatability of chops was determined by a trained
sensory taste panel. A total of 510 F2 animals were genotyped for 124 microsatellite
markers evenly spaced across the entire genome. Data were analyzed with line cross
least squares regression interval mapping methods using sex and litter as fixed effects and
carcass weight or harvest age as covariates. Significance thresholds of the F -statistic for
additive, dominance, and imprinted QTL were determined on chromosome- and genome-
wise levels by permutation tests. A total of 94 QTL for 35 of the 38 traits analyzed were
found to be significant at the 5% chromosome-wise level. Of these 94 QTL, 43 were
significant at the 1% chromosome-wise, 27 at the 5% genome-wise, and 14 at the 1%
genome-wise significance thresholds. Putative QTL were discovered for 45 min pH and

pH decline on SSC 3, marbling score and carcass backfat on SSC 6, carcass length and

 

3 This research was financially supported by the Michigan State University Department
of Animal Science, the Michigan Agricultural Experiment Station, a Michigan Animal
Initiative Coalition Grant, and USDA-CSREES NRI Award 2004-35604-14580.

91

number of ribs on SSC 7, marbling score on SSC 12, and color measurements and
tenderness score on SSC 15. These results will facilitate fine mapping efforts to identify
genes controlling carcass composition and meat quality traits that can be incorporated
into marker-assisted selection programs to accelerate genetic improvement in pig
populations.
Introduction

Enhancement of production efficiency and improvement of product quality are
major concerns for producers of food animals. Swine have been selected for increased
lean growth, but the antagonistic relationship with meat quality, shown by significant
genetic correlations of carcass leanness to ultimate pH (-O.l3), reflectance (0.16), and
drip loss (0.05) (Sellier, 1998), has caused a decrease in meat quality. Additionally,
Wood (1985) reported increased occurrence of less juicy pork products with leaner pigs.
Locating specific favorable quantitative trait loci (QTL) for meat quality and using this
information in genetic improvement programs will help overcome this natural
relationship and allow improvement in both efficient production and product quality.

The Duroc and Pietrain breeds are utilized worldwide as sire breeds, and these
breeds differ in carcass and meat quality phenotypes. Quiniou and Noblet (1995) used
Pietrain boars in their study because of the breed’s propensity towards leanness.
Affentranger et al. (1996) compared Duroc and Pietrain pigs and reported more backfat
for Duroc animals as compared to Pietrain animals. In Edwards et al. (2003), Duroc-
sired pigs had longer carcass lengths than Pietrain-sired pigs, while Pietrain-sired pigs
had less backfat at the tenth rib and larger longissimus muscle area at harvest than Duroc-

sired pigs. Meat quality was more favorable for Duroc- versus Pietrain-sired pigs in

92

Affentranger et al. (1996) and Edwards et al. (2003). In general, Duroc and Duroc-sired
pigs have favorable meat quality (Langlois and Minvielle, 1989b; Jeremiah et al., 1999),
whereas Pietrain and Pietrain-sired animals are leaner with average meat quality
(Edwards at al., 2003). The objective of this study was to conduct a full genome scan
using microsatellite markers to search for QTL affecting carcass and meat quality traits in
an F2 Duroc x Pietrain resource population.
Materials and Methods

Population Development

A three-generation resource population was developed from four F0 Duroc sires
and sixteen F0 Pietrain dams at Michigan State University to study traits of growth, body
composition, and meat quality. All grandparents were determined homozygous normal
for the RYR-l gene by a DNA test (Fujii et al., 1991). Further details of population
development and animal management are found in Edwards (2005a).
Phenotype Collection

At harvest, pigs were transported to one of two abattoirs. A total of 176 pigs were
harvested at the Michigan State University Meats Laboratory (East Lansing, MI), and the
remaining pigs were transported to a small federally inspected plant in western Michigan
(DeVries Meats, C00persville, MI). Both groups were fasted and allowed to rest
overnight with access to water. Ear tag and tattoo numbers were collected at harvest to
maintain identity of each carcass. Carcass traits collected included hot carcass weight
and longissimus dorsi pH and temperature at 45 min and 24 h postmortem. Dressing
percentage was calculated by dividing hot carcass weight by harvest live weight. After

overnight chilling, measurements were taken according to National Pork Producers

93

Council guidelines (N PPC, 2000) and included midline first rib backfat, last rib backfat,
last lumbar backfat, number of ribs, and carcass length. Weights of primal cuts of ham,
closely trimmed loin, picnic shoulder, Boston shoulder, belly, and spareribs were
recorded. During carcass fabrication, measurements of tenth rib backfat and longissimus
muscle area were also recorded. A section of loin from the tenth rib to the last rib was
further evaluated for meat quality traits at Michigan State University. All measurements
were taken from the left side of each carcass.

Boneless longissimus dorsi were removed from loin sections and external fat
removed. A small portion of longissimus dorsi was diced and frozen for proximate
analysis. Two 2.54 cm thick chops were cut from the anterior end of the longissimus
dorsi for fresh meat quality analysis. The two chops were allowed to bloom for a
minimum of 10 minutes and evaluated for subjective scores of color and marbling
(NPPC, 2000) and firmness (N PPC, 1991). The color score scale ranged from 1 (pale
pinkish gray) to 6 (dark purplish red). The marbling score scale was 1 to 10 (closely
approximating fat percentage). The firmness score scale was 1 (very sofi and watery) to
5 (very firm and dry). Additionally, objective color scores of CIE L* (lightness), a*
(redness), and b* (yellowness) were obtained using a Minolta CR-3 10 colorimeter
(Ramsey, NJ) with a D65 illuminant and a two-degree standard observer. Chops were
weighed, hung in sealed plastic bags for 24 h at 4°C, and then weighed again for drip loss
measurement. The remaining section of the longissimus dorsi was vacuum packaged,
aged at 4°C until 7 d postmortem, and frozen for further meat quality tests of cook yield,

shear force, and sensory taste panel analysis.

94

From frozen loin sections, two chops were cut for cook yield and Warner-Bratzler
shear force (WBS) analysis. For cook yield measurements, each chop was thawed,
weighed, cooked to 71°C internal temperature on a Taylor clamshell grill (Model QS24,
Taylor Co., Rockton, IL), cooled to room temperature, and weighed again. From these
chops, six cores (three cores from each chop) were taken parallel to the muscle fiber
direction using a drill press-mounted corer. Cores were sheared perpendicular to muscle
fibers using a Warner-Bratzler head on a TA-HDi texture analyzer (Texture Technologies
Corp., Scarsdale, NY). The cross-head speed was 3.30 mm/s. Samples for proximate
analysis were ground using dry ice and measured for moisture (oven drying), fat (soxhlet
ether extraction), and protein (nitrogen combustion, Model FP-ZOOO, Leco Inc., St.
Joseph, MI), following AOAC procedures (2000). A total of 958 animals had carcass and
meat quality traits measured. 1
Trained Sensory Panel Evaluation

A trained panel of seven healthy adults (ages 20-65) was utilized to determine
specific sensory attributes of each longissimus chop. The sensory panel was trained
according to Meilgaard et al. (1991) and AMSA (1995). All panelists had experience in
sensory evaluation and were previously trained to evaluate various meat products. Each
sample was evaluated for juiciness, muscle fiber and overall tenderness, connective
tissue, and off-flavor using an 8 point hedonic scale. Higher scores in each of the first
four categories were more favorable and indicated extremely juicy, extremely tender, or
no connective tissue for each of these attributes, respectively, while lower scores for off-

flavor were indicative of less off-flavor.

95

Frozen chops were thawed for 24 h at 2.6°C and then cooked on a Taylor
clamshell grill (Model QS24, Taylor Co., Rockton, IL). The upper plate was set to
104.4°C and the bottom plate was set to 102.8°C with a 2.16 cm gap between plates.
Temperature was monitored by inserting a copper constantan thermocouple (0.051 cm
diameter, 15.2 cm length, Omega Engineering Inc., Stamford, CT) into the geometric
center of the pork chop. Chops were cooked to a final internal temperature of 71°C.
Sample preparation included cutting 1.27 cm cubes from the center portion of each chop,
and two cubes were placed in 2 oz. soufflé cups and covered with a lid. Souffle cups
were placed in a Pyrex two quart bowl with a lid and the bowl was covered with warm
towels to keep the samples warm. The insulated bowl was placed in an insulated
container and transferred to the sensory evaluation room.

Testing took place in climate controlled, partitioned booths with cool
incandescent light. The order of sample preparation was randomized within each session
to minimize positional bias and a 3 digit random code was used to label the samples. The
samples were picked up with a toothpick, chewed with the molars, and evaluated.
Expectorant cups were provided to prevent taste fatigue and distilled, deionized water
was used to clean the palate between samples. The panelists were standardized each day
by evaluating a warm-up sample and discussing the results. A total of 18-24 samples
were evaluated on each day, and the day was divided into three sessions with a 15 min
break between each session.

Genotype Collection
Whole blood was collected from all F0, F1, and F2 animals for DNA isolation.

White blood cells were separated and frozen for subsequent DNA extraction.

96

Dinucleotide microsatellite genetic markers were utilized in this study to obtain
genotypes for a whole genome scan of approximately 20 cM spacing across the 18
autosomes and X chromosome. Markers were selected from the published pig genome
linkage map (http://www.marc.usda.gov/genome/swine/swine.htrnl, USDA, 2005b), and
the initial set of markers were tested for informativeness (markers segregating between F o
breeds and estimated to have three or more alleles) using fluorescent primers distributed
by the US Pig Genome Coordinator (supported by the USDA-CREES through the
National Research Support Project 8). The resulting 124 markers were genotyped for 510
F2 animals, their parents, and grandparents at a commercial laboratory (GeneSeek Inc.,
Lincoln, NE). These 510 animals were sampled across all farrowing groups from 61
entire litters. Pigs represented all F1 sires with at least 100 grand progeny from each F o
sire. Fifteen of the 16 F0 dams had a son or daughter as a parent that produced multiple
litters of the selected F2 pigs with the remaining F0 dam represented by a single F.
daughter with one litter in this group. Markers used in this genome scan were those
reported in Edwards (2005a).
QTL Analysis

Genetic linkage maps were constructed for all 18 autosomes and the X
chromosome using Crimap version 2.4 software (Green et al., 1990) and are reported in
Edwards (2005a). These maps were used in an F2 least squares interval mapping
framework for QTL analysis (Haley et al., 1994) similar to that in Edwards (2005a). The
F2 analysis option of the QTL Express software (Seaton et al., 2002) was used to search
for single QTL with additive, dominance, or imprinting effects on the 18 autosomes and

additive effects on the X chromosome for the carcass and meat quality traits with the

97

fixed effects and covariates as listed in Table IV. 1. The terms that were included in each
model were those that were found to be important to the model (P < 0.20) when each trait
was analyzed by ordinary least squares analysis of variance without QTL effects. Fixed
effects of sex of the animal and harvest date were included in every model. Some models
contained the covariates of carcass weight or harvest age, while others had neither
covariate (Table IV.1).

Tests of the full model including additive, dominance, and imprinting effects
versus the reduced model without these effects were carried out to determine F -ratios at 1
cM intervals across the entire genome. Significance thresholds of 5% and 1% at the
chromosome-wise and 5% and 1% at the genome-wise levels were determined through
the use of permutation tests (Churchill and Doerge, 1994) in QTL Express using 30,000
permutations. For each QTL determined to be significant at the 5% genome-wise level,
confidence intervals of the QTL position were determined using a bootstrap method with
5,000 permutations in QTL Express (V isscher et al., 1996).

Results and Discussion

QTL Analysis

Trait means and standard deviations for genotyped animals with measured
phenotypes are reported in Table IV.2. These means and standard deviations are similar
to those measured in other resource populations reporting similar traits (e. g. Malek et al.,
2001b; Stearns et al., 2005). The permutation test results of F -ratio calculations to
determine significance thresholds for the model testing additive, dominance, and
imprinting effects ranged from 3.44 to 4.35 across different chromosomes for the 5%

chromosome-wise levels and from 4.71 to 5.64 for the 1% chromosome-wise levels.

98

Genome-wise F -ratio threshold levels were 6.34 and 7.63 for 5% and 1% levels,
respectively. Significance levels of the F-statistic for the additive effect only model for
SSC X were 6.68 and 9.93 for 5% and 1% chromosome-wise levels, respectively,
whereas the significant F-ratios were 12.86 and 16.05 for 5% and 1% genome-wise
levels, respectively.

Estimates of position and F -ratio for carcass and meat quality QTL significant at
the 5% chromosome-wise level are listed in Table IV.3. The table is sorted by
chromosome and position within each chromosome. Additionally, the additive,
dominance, and imprinting effect of each QTL along with the standard errors are listed in
Table IV.3. A total of 94 QTL for 35 of the 38 traits were found to be significant at 5%
chromosome-wise levels. Of these 94 QTL, 43 were significant at the 1% chromosome-
wise, 27 at the 5% genome-wise, and 14 at the 1% genome-wise significance thresholds.
No significant QTL were detected in this population for subjective firmness, Boston
shoulder weight, or picnic shoulder weight.

All chromosomes, except 13, contained at least one QTL for these 38 traits.
Although these two breeds are used in similar capacities in many pork production chains,
many important QTL that impact phenotypic differences between the two breeds still
exist and are possible candidates to explore more thoroughly. Differences in carcass and
meat quality traits between animals sired by Duroc or Pietrain boars in a crossbred
progeny study (Edwards et al., 2003) have been previously reported. Some of the breed
allelic QTL effects in the current study had effects in the same direction as breed effects
reported in Edwards et al. (2003), but a few cryptic alleles were detected that acted in an

opposite direction to the general trend for the overall breed effects.

99

Carcass Measurements

The carcass measurement traits included many of the classically measured carcass
traits that impact prices paid for market pigs. Measures of off-farm BW and hot carcass
weight are related and QTL for these two traits appeared in similar positions. Two
significant QTL were identified, one on SSC 4, expressing an additive effect where
Pietrain alleles increased these weights, and one on SSC 10, which indicated an
imprinting effect where alleles from Pietrain dams increased these weights. These QTL
were significant at the 5% genome-wise and 1% chromosome-wise levels, respectively.
While other studies have reported QTL for live weight at harvest on SSC 4 (Cepica et al.,
2003; Marklund et al., 1998), the position of these QTL were more distal than the 12 cM
distance from the first marker observed in this study. By utilizing the population of
Cepica et al. (2003), another analysis by Dragos-Wendrich et al. (2003) found a QTL for
live weight at harvest on SSC 10, but, again, not in the same position as reported here.
Several studies have reported QTL for carcass weight on SSC 4, but all of these studies
reported positions more distal (85 to 121 cM from SW2404) than was discovered in this
population (Pérez—Enciso et al., 2000; Malek et al., 20013; Cepica et al., 2003).

Carcass temperature and pH at 45 min and 24 h postmortem are important early
indicators of meat quality. The decline between these two points is a rarely studied trait,
but is indicative of changes in meat properties that affect further quality parameters.
Figure IV.1 demonstrates the F -ratio curves plotted versus relative marker positions on
SSC 3 and illustrates the similar shapes of F -ratios for 45 min pH and pH decline from 45
min to 24 h postmortem. On SSC 3, a QTL that was significant at the 1% chromosome-

wise level was discovered in this population which was additive and increased 45 min pH

100

when Pietrain alleles were present. In addition this QTL also caused a smaller decrease
in pH from 45 min to 24 h with Pietrain alleles. The Pietrain alleles caused favorable
change for these two traits which should improve meat quality. These can be considered
cryptic alleles since previous studies reported lower pH values for Pietrain or Pietrain-
cross pigs compared to animals from other breeds (Affentranger et al., 1996; Garcia-
Macias et al., 1996; Edwards et al., 2003). This QTL affecting pH on SSC 3 was in a
similar location to a pH QTL reported by Beeckmann et al. (2003) in a population that
containediPietrain germplasm. Additional QTL significant at the 5% genome-wise level
were found for 24 h carcass temperature on SSC 5, which was estimated to have a
positive dominance effect and another on SSC 6, which was additive in its effect and
indicated animals with Duroc alleles increased 24 h carcass temperature. On SSC 9, QTL
for 45 min temperature and 24 h temperature were found, but in separate regions of the
chromosome. No other QTL studies have reported taking carcass temperature
measurements, but our work indicated that significant QTL did control carcass
temperature at 45 min and 24 h after harvest.

Measurements of skeletal characteristics of a carcass are important in determining
size and ability to grow to heavier weights while maintaining leanness. Carcass length
QTL were present on SSC 6 and 7 with both significant at the 1% genome-wise level.
The QTL on SSC 6 showed a positive dominance effect while the one on SSC 7 affected
both carcass length and dressing percentage in an additive manner (Figure IV.2). Duroc
alleles caused a lengthening of the carcass and decreased dressing percentage. A similar
dominance effect QTL was reported on SSC 6 by Malek et al. (2001a), while the QTL on

SSC 7 was in a similar position to one reported by Rohrer and Keele (1998b). Two

101

putative QTL were detected for the number of ribs counted in each carcass with one on
SSC 7 and one on SSC 18. The QTL for number of ribs on SSC 7 was found in the same
location as one detected in several populations by Mikawa et al. (2005). Of nine families
reported in their study, the one that contained Duroc germplasm had the highest F-ratio
corresponding to a significant QTL on SSC 7. Their study reported an additive effect for
this QTL, but they did not test imprinting effects. Our QTL was estimated to have an
imprinting effect with animals that received a Duroc allele from the sire having an
increase of 0.34 ribs.

Carcass backfat thickness and longissimus muscle area are predictive of body
composition, and have been used to determine pricing structure at packing plants.
Traditionally, carcass measurements of backfat have been taken along the midline at the
first rib, last rib, and last lumbar vertebra in pigs, while measurements of backfat and
longissimus muscle area at the tenth rib have been taken on a ribbed carcass. All of these
measures can be used to determine body composition estimates. A region of SSC 6
(Figure IV.3) contained a significant QTL that affected many of these related backfat
traits as well as carcass longissimus muscle area. This was an additive QTL in which
Duroc alleles increased backfat and reduced longissimus muscle area. These results were
consistent with overall breed results reported in Edwards et al. (2003) where Duroc
germplasm contributed to an increase in backfat compared to Pietrain-sired pigs. The
region of SSC 6 that affected tenth rib backfat was also found significant by Malek et al.
(20013). Ovilo et al. (2002b) also reported that the same region of SSC 6 affected
longissimus muscle area as reported here. An additional QTL on SSC 18 that was

dominant and increased last rib and tenth rib backfat when Duroc alleles were present

102

was discovered in this analysis. Finally, a QTL significant at the 1% chromosome-wise
level was found for first rib backfat on SSC 5, and no previous pig QTL studies have
identified a first rib backfat QTL in this region (Hu et al., 2005). Su et al. (2004) reported
a QTL for carcass longissimus muscle area on SSC 7 in a similar region to the QTL
significant at the 1% chromosome-wise level discovered in this study. Sato et al. (2003)
also discovered a QTL for longissimus muscle area on SSC 7 in a population with a
Duroc founder boar, but their QTL was located in a more proximal position than the QTL
reported here.
Primal Cut Weights

Since primal cut weights were adjusted for carcass weight in the model to
determine significance of QTL, these adjusted primal weights have a similar
interpretation to primal cut weights as a percentage of carcass weight. Although QTL
were not found in this study for Boston shoulder and picnic shoulder weights, these
primal cut weights were highly dependent on where they were separated from one
another and had higher standard deviations than other primal cuts compared to their mean
weights (Table IV.2), so larger environmental variation may have prevented discovery of
significant QTL. Other primal cuts had several putative QTL in this study. Again, SSC 6
contained a QTL that indicated more muscling in the ham and loin for animals with
Pietrain alleles, which paralleled results of Edwards et al. (2003) that reported Pietrain-
sired pigs had larger ham and loin primal weights than Duroc-sired pigs. Other pig QTL
studies that reported loin weight, summarized by Hu et al. (2005), had not found QTL on
SSC 6. Geldermann et al. (2003) reported a QTL for ham weight on SSC 6, but it was

slightly more proximal to S0099, our first marker, than the QTL reported here. An

103

additional loin weight QTL, significant at the 5% genome-wise level, was found on SSC
3 and was estimated to have a negative dominance effect. Two QTL for belly weight
were found, one on SSC 12 and one on SSC 14, at the 1% chromosome-wise and 5%
genome-wise levels, respectively. The QTL on SSC 12 indicated that Duroc alleles
additively increased belly weight whereas the QTL on SSC 14 was dominant and
negative in its effect. These two QTL have not been previously reported (Hu etal.,
2005). The analysis for sparerib weight indicated QTL located on SSC 1, 4, 6, 7, 8, and
18, although only the QTL on SSC 1 exceeded the 5% chromosome-wise significance
threshold.
Meat Quality

Consumer perception of fresh meat products is affected at least in part by color,
marbling, and firmness. Evaluation of these parameters on chops cut from longissimus
dorsi in this study revealed QTL affecting traits of color and marbling, but not firmness.
Putative QTL for subjective color were located on SSC 4, 6, 7, and 15. The QTL on SSC
15 was significant at the 1% chromosome-wise level for subjective color as well as for a*
and at the 1% genome-wise level for If (Figure WA). The additive effects indicate that
Duroc alleles lowered the color score and raised the L* value while they lowered the a*
value. The population studied by Malek et al. (2001b) also contained a QTL for L* on
SSC 15, but it was slightly more distal than the position reported here. A QTL in a
similar position to the one reported here for a* was also reported by de Koning et al.
(2001). A QTL for L* was also discovered in this study at the same position on SSC 7 as
the QTL reported for subjective color score. While not in the same position on SSC 7 as

the QTL reported here, Ovilo et al. (2002a) also reported a QTL for L* on SSC 7. A

104

QTL at the 5% genome-wise level for a* was located on SSC 6 and was additive in its
effect where Duroc alleles increased the a* value. The only QTL for b* found in this
study was on SSC 9 and was significant only at the 5% chromosome-wise level.

Another factor that impacts a consumer’s choice of meat and subsequent eating
experience is marbling. The subjective marbling score analysis revealed three significant
QTL on SSC 6, 10, and 12. The QTL on SSC 6 was additive and indicated that Duroc
alleles increased marbling and was significant at the 1% genome-wise level while the one
on SSC 12 was significant at the 5% genome-wise level. Marbling score and
intramuscular fat percentage showed similar F -ratio curves when plotted versus relative
marker position on SSC 6 (Figure IV.3), which indicated QTL within this region are
controlling these traits.

The three traits of percentage moisture, fat, and protein from proximate analysis
of longissimus dorsi are highly related as they are derived from percentages of the same
whole. As subjective marbling score approximates fat percentage in each chop, two
significant QTL for fat percentage were found at the same locations as the QTL for
marbling score on SSC 6 and 12, again significant at the 1% and 5% genome-wise levels,
respectively. Nearly the same position on SSC 6 for a QTL affecting intramuscular fat
was reported by Ovilo et al. (2002b). The QTL on SSC 6 also affected moisture
percentage, and was previously reported by Su et al. (2004). This chromosomal region
also affected protein percentage, while the QTL on SSC 12 affected moisture percentage.
An additional QTL, significant at the 1% genome-wise level, for protein percentage was

located on SSC 15. Selection for any of these QTL will directly impact the other two

105

traits since these three traits are not strictly independent, as evidenced by similar F-ratio
curves for fat and moisture percentages on SSC 12 (Figure IV.5).

Several traits measured are directly related to storage, preparation, and eating
quality of chops. One was drip loss, and a QTL was found on SSC 9 with an estimated
dominance effect that indicated that Duroc alleles caused a higher percentage of drip loss.
Cook yield, which measured moisture loss during cooking, had two QTL, with one on
SSC 5 significant at the 1% chromosome-wise level and an additional one on SSC 15.
None of the drip loss or cook yield QTL discovered in this study, including related traits,
were previously reported in a summary of pig QTL studies (Hu et al., 2005). Our
analyses indicated significant QTL for WBS on SSC 7 and 15, with both significant at the
1% chromosome-wise level. A similar trait, Instron (star probe) force, was reported to
have a QTL on SSC 15 (Malek et al., 2001b) in a similar position to the WBS QTL in our
study, but the QTL on SSC 7 had not been reported previously (Hu et al., 2005). These
same two QTL regions on SSC 7 and 15 reported here for WBS also affected tenderness
and overall tenderness as determined by the trained sensory taste panel and were
significant at the 5% genome-wise level for these two important eating quality traits. The
QTL on SSC 7 acted in an additive mode of inheritance where Pietrain alleles led to
lower, and less favorable, tenderness scores. On SSC 15, the mode of inheritance
followed an imprinting pattern where Pietrain alleles from the sire again led to lower
tenderness scores. Furthermore, tenderness, overall tenderness, and WBS were
controlled by the same chromosomal region on SSC 15 (Figure IV.6). Malek et al.
(2001b) also reported a similar position for a tenderness QTL on SSC 15. Additional

QTL for both tenderness and overall tenderness were located on SSC 9 and 10. There

106

was only one QTL for sensory taste panel juiciness, on SSC 2, possibly due to its high
propensity to be affected by meat preparation conditions. A QTL on SSC 2 for juiciness
was also reported by Stearns et al. (2005), but it was located more proximal than the QTL
discovered here. Additionally, a QTL for off-flavor was located on SSC 2, a
chromosome in which Malek et al. (2001b) reported two QTL for off-flavor. A QTL
significant at the 1% chromosome-wise level which impacted connective tissue scores
was discovered on SSC 10. This QTL had a dominance mode of inheritance. Refining
the location for these QTL for sensory taste panel attributes will provide potentially
important information for selection of prospective parents in swine populations since
these attributes ultimately impact a consumer’s dining experience.
Confidence Intervals

For QTL that were significant at the 5%genome-wise level, 95% confidence
intervals were estimated using bootstrapping with resampling in QTL Express (Seaton et
al., 2002). These confidence intervals are listed in Table 1V.4. Confidence intervals
were not calculated for QTL significant at the 5 or 1% chromosome-wise level since
preliminary analyses indicated that many of these intervals tended to encompass the
entire chromosome on which they reside.

Implications

Numerous QTL that control economically important traits of carcass composition
and meat quality were revealed in this F2 Duroc-Pietrain resource population. These
QTL are extremely important as they give us insight into traits that are not easily
measured in breeding animals but add value to end products. These QTL will become

even more important as the traits they influence achieve greater economic value. Many

107

QTL for meat quality, with both desirable and undesirable alleles, existed in the Duroc
and Pietrain breeds utilized in this study, and desirable QTL could be incorporated
systematically into breeding schemes as these two populations are already a major part of
commercial pig production. Chromosomal regions of interest discovered in these
populations are being subjected to further analyses and refinement of QTL positions may
lead to their incorporation into selection programs for prospective parents and

subsequently increase value in these pork products.

108

Table lV.1. Fixed effects and covariates for carcass and meat quality trait QTL analyses.

 

 

 

Fixed Effectsa Covariates“
Harvest Carcass Harvest
Trait Sex Date wiight Ag:
Carcass measures
Off-farm BW, kg X X - X
Hot carcass weight, kg X X - X
Dressing percentage, % X X - -
45 min carcass temperature, °C X X X -
24 h carcass temperature, °C X X X -
45 min pH X X X -
24 h pH X X X -
45 min-24 h pH decline X X X -
Carcass length, cm X X X -
Number of ribs X X - -
First rib backfat, mm X X X -
Last rib backfat, mm X X X -
Last lumbar vertebra backfat, mm X X X -
Tenth rib backfat, mm X X X -
Longissimus muscle area, cm2 X X X -
Primal cut weights
Ham weight, kg X X X -
Loin weight, kg X X X -
Boston shoulder weight, kg X X X -
Picnic shoulder weight, kg X X X -
Belly weight, kg X X X -
Spareribs weight, kg X X X -
Meat quality evaluation
Color, 1-6 X X - -
Marbling, 1—10 X X - X
Firmness, 1-5 X X - X
L* X X - -
a“ X X - -
b* X X - -
Proximate analysis
Moisture,% X X X -
Fat, % X X X -
Protein, % X X X -
Laboratory analyses
Drip loss,% X X - X
Cook yield, % X X - -
Warner-Bratzler shear force, kg X X - -
Sensory taste panel analyses
Juiciness, 1-8 X X - -
Tenderness, 1-8 X X - -
Overall tenderness, 1-8 X X - -
Connective tissue, 1-8 X X - -
Off-flavor, 1-8 X X - -

 

a X indicates used in model, - indicates not used in model

109

Table IV.2. Number of records, means, and standard deviations for

carcass and meat quality traits measured.

 

Trait N Mean SD
Carcass measures
Off—farm BW, kg 504 111.89 9.12
Hot carcass weight, kg 504 81.35 7.19
Dressing percentage, % 504 72.69 2.12
45 min carcass temperature, °C 503 39.20 2.26
24 h carcass temperature, °C 502 2.74 1.19
45 min pH 497 6.38 0.23
24 h pH 484 5.53 0.15
45 min-24 h pH decline 478 0.84 0.23
Carcass length, cm 503 78.92 2.59
Number of ribs 357 14.84 0.86
First rib backfat, mm 417 40.67 7.19
Last rib backfat, mm 503 28.65 6.29
Last lumbar vertebra backfat, mm 502 22.29 6.59
Tenth rib backfat, mm 499 24.15 7.25
Longissimus muscle area, cm2 500 40.94 4.60
Primal cut weights
Ham weight, kg 503 9.56 0.81
Loin weight, kg 503 8.17 0.86
Boston shoulder weight, kg 503 3.72 0.63
Picnic shoulder weight, kg 503 3.89 0.65
Belly weight, kg 503 4.94 0.69
Spareribs weight, kg 502 1.48 0.20
Meat quality evaluation
Color, 1-6 502 3.21 0.82
Marbling, 1-10 503 2.88 0.80
Firmness, 1-5 489 2.86 0.81
L* 487 53.63 2.21
a* 487 17.26 1.91
b* 487 9.10 1.61
Proximate analysis
Moisture,% 494 73.88 1 .50
Fat, % 494 3.27 1.33
Protein, % 493 23.39 1.09
Laboratory analyses
Drip loss,% 503 1.71 1.18
Cook yield, % 498 77.46 2.94
Warner-Bratzler shear force, kg 497 3.19 0.69
Sensory taste panel analyses
Juiciness, 1-8 501 5.28 0.58
Tenderness, 1-8 501 5.59 0.61
Overall tenderness, 1-8 501 5.65 0.55
Connective tissue, 1-8 501 6.39 0.38
Off-flavor, 1-8 501 1.14 0.21

 

110

Table IV.3. Position and significance level of carcass and meat quality QTL significant at the 5%
chromosome-wise level with additive, dominance, and imprinting effects and standard errors.
Pos. Additive Dominance lmprintng_

Chr Trait (cM) Effect SE Effect SE Effect SE F-ratio‘I
Longissimus muscle area, cm2 10 -123 0.31 0.44 0.56 0.13 0.30 5.47
Dressing percentage,% 71 -0.42 0.12 0.05 0.20 -0.38 0.13 6.03"
Spareribswt, kg 172 0.02 0.01 -0.11 0.02 -0.04 0.01 1124*"
Juiciness,1-8 54 -0.10 0.04 -0.16 0.07 -0.08 0.04 4.64
Off-flavor,1-8 70 -0.05 0.01 0.01 0.02 -0.02 0.02 4.70
45 min-24th decline 86 -0.05 0.02 -0.03 0.02 0.04 0.02 5.70"
45 min pH 91 -0.06 0.02 -0.02 0.02 0.03 0.02 6.49”
Loinwt, kg 114 0.06 0.04 -0.21 0.06 -0.08 0.04 6.60”
Longissimus muscle area,cm2 155 0.16 0.27 -1.37 0.41 -0.27 0.27 4.29
Longissimus muscle area, cm2 0 -0.33 0.26 4.09 0.35 0.35 0.24 4.71
Hot carcass weight, kg 12 -2.18 0.50 0.94 0.83 -0.47 0.47 7.14”
Off-farm BW, kg 13 -2.70 0.65 0.76 1.08 -0.71 0.61 6.56”
Color,1-6 45 0.02 0.06 0.35 0.10 0.04 0.06 4.21
Spareribswt, kg 53 0.02 0.01 -0.03 0.01 0.00 0.01 4.31
First rib backfat, mm 86 0.64 0.47 1.16 0.74 1.79 0.47 628*
24hcarcasstemperature,°C 117 0.02 0.03 0.20 0.05 0.09 0.03 7.13"
Cook yield,% 159 -0.28 0.20 0.37 0.32 0.82 0.22 5.52'
a“ 24 0.24 0.07 0.15 0.11 -0.14 0.08 6.49”
Color, 1-6 92 0.17 0.06 0.04 0.10 -0.15 0.06 4.77
Warner-Bratzler shearforce, kg 101 -0.14 0.05 -0.14 0.07 0.05 0.05 4.64
Marbling,1-10 116 0.43 0.06 -0.15 0.11 0.04 00618.41“
Fat,% 117 0.71 0.10 -0.37 0.17 0.14 0.10 2001*"
Ham wt, kg 121 -0.17 0.03 0.18 0.06 -0.04 0.031179m
Tenth rib backfat, mm 125 3.66 0.39 -2.09 0.70 0.42 0.42 32.18“"
Last lumbar backfat, mm 129 2.11 0.42 -2.31 0.70 0.03 0.4611.87***
Carcass length, cm 139 -0.57 0.13 1.02 0.20 0.16 0.1414.57***
Loin wt, kg 141 -0.18 0.03 0.17 0.06 0.04 0.0412.25***
Last rib backfat, mm 144 1.04 0.38 -2.19 0.65 -0.17 0.41 610*
Protein,% 150 -0.39 0.08 0.24 0.14 -0.17 0.091006”
Spareribswt, kg 154 —0.03 0.01 0.04 0.02 -0.01 0.01 4.84
Longissimus muscle area, cm2 156 -1.27 0.28 -0.69 0.48 0.06 0.30 7.40“
24hcarcass temperature, °C 176 0.14 0.03 -0.04 0.05 0.01 0.04 6.70“
Moisture,% 181 -0.19 0.09 -0.34 0.17 -0.34 0.10 6.21“

3 Significant at * = 1% chromosome-wise, ** = 5% genome-wise, *** = 1% genome-wise levels

 

 

mmmmmmmmmmmmmmmmmmmnnebewwwwNN—x—x—s

 

111

Table IV.3 (cont’d).
Pos. Additive Dominance lmprinting_

 

 

Chr Trait (cM) Effect SE Effect SE Effect SE F-ratio'
7 L* 1 0.23 0.14 0.24 0.21 -0.50 0.16 4.32
7 Color, 1-6 2 -0.05 0.05 -0.07 0.08 0.21 0.06 4.52
7 Protein,% 6 -0.24 0.07 -0.21 0.13 0.02 0.08 4.48
7 Carcass length, cm 67 0.81 0.17 0.40 0.34 0.19 0.18 9.01“"
7 Dressing percentage, % 70 -0.73 0.15 -0.65 0.31 -0.05 0.1610.13”*
7 Longissimus muscle area,cm2 108 -1.28 0.36 -1.61 0.72 0.17 0.37 628*
7 Numberof ribs 124 0.19 0.06 0.10 0.11 0.34 00615.13”
7 Spareribswt, kg 135 0.01 0.01 -0.01 0.01 0.03 0.01 4.93
7 Warner-Bratzler shear force, kg 158 0.19 0.05 0.05 0.08 0.07 0.05 5.74"
7 Overall tenderness, 1-8 159 -0.15 0.04 0.00 0.06 -0.01 0.04 4.72
7 Tenderness, 1-8 160 -0.16 0.04 -0.01 0.07 -0.01 0.04 4.75
7 45 min-24th decline 164 0.04 0.02 0.01 0.02 0.04 0.02 4.35
8 Spareribs wt, kg 62 0.03 0.01 0.02 0.02 -0.01 0.01 4.41
8 Longissimus muscle area, cm2 96 -0.82 0.24 -0.38 0.36 -0.15 0.26 4.68
8 Moisture,% 152 -0.35 0.12 0.45 0.21 -0.03 0.12 4.19
9 Drip loss,% 1 -0.04 0.07 0.36 0.10 -0.05 0.07 4.75
9 Off-farm BW, kg 3 -0.37 0.56 2.89 0.83 0.77 0.59 4.39
9 Hot carcassweight, kg 3 -0.40 0.44 2.38 0.65 0.76 0.46 5.25
9 Overall tenderness,1-8 5 0.00 0.04 0.04 0.06 0.14 0.04 4.42
9 Ham wt, kg 8 -0.11 0.03 0.01 0.05 -0.04 0.03 5.09
9 24hcarcass temperature, °C 13 0.09 0.03 -0.02 0.04 -0.06 0.03 4.08
9 Warner-Bratzler shear force, kg 17 -0.01 0.04 -0.09 0.07 -0.18 0.05 4.77
9 Tenderness, 1-8 28 -0.02 0.05 0.07 0.09 0.19 0.05 4.85
9 Fat,% 62 -0.14 0.10 -0.11 0.16 0.35 0.10 4.66
9 Moisture,% 64 0.21 0.10 -0.01 0.17 -0.34 0.11 4.14
9 45 min carcass temperature, °C 108 0.06 0.10 0.67 0.19 0.03 0.11 4.27
9 b“ 127 -0.13 0.05 0.14 0.07 0.01 0.07 4.25
10 Tenderness,1—8 0 0.06 0.04 0.19 0.06 0.01 0.04 4.08
10 Overall tenderness,1-8 0 0.05 0.04 0.19 0.05 -0.02 0.04 4.65
10 Connective tissue, 1-8 61 0.08 0.04 0.15 0.07 -0.05 0.03 5.36'
10 Off-farm BW, kg 74 -0.59 0.82 2.22 1.35 -2.72 0.77 5.62'
10 Hot carcassweight, kg 75 -0.37 0.63 0.93 1.03 -2.38 0.60 593*
10 Marbling,1-10 79 -0.12 0.08 0.30 0.11 -0.14 0.07 4.22
11 45 min-24th decline 107 0.07 0.02 -0.08 0.06 0.03 0.03 3.80
11 Fat,% 119 0.28 0.12 -0.43 0.25 0.36 0.13 4.78
11 Moisture,% 119 -0.40 0.13 0.30 0.27 -0.28 0.14 4.69
12 Bellywt, kg 53 0.11 0.03 0.01 0.04 0.01 0.03 6.04*
12 Moisture,% 54 -0.37 0.10 0.31 0.16 -0.11 0.10 7.26“
12 Fat,% 57 0.30 0.09 -0.32 0.15 0.11 0.09 6.99”
12 24th 61 -0.02 0.01 -0.04 0.01 0.01 0.01 4.23
12 Marbling,1-10 67 0.21 0.06 -0.24 0.10 0.02 0.06 7.04“

 

’ Significant at * = 1% chromosome-wise, ** = 5% genome-wise, *** = 1% genome-wise levels

112

Table lV.3 (cont’d).
Pos. Additive Dominance Imprinting_

 

 

Chr Trait (cM) Effect SE Effect SE Effect SE F-ratio'
14 a* 51 -0.28 0.07 0.05 0.12 0.01 0.07 5.05
14 Bellywt, kg 114 0.06 0.03 -0.24 0.06 -0.05 0.03 7.53”
15 Protein, % 57 0.38 0.08 -0.01 0.14 0.41 00814.89“
15 L* 60 0.70 0.17 -0.70 0.29 0.18 0.17 8.03"“
15 a“ 64 -0.25 0.07 0.15 0.11 -0.13 0.07 560*
15 Color, 16 64 -0.16 0.06 0.16 0.09 -0.13 0.06 5.03'
15 Tenderness, 1-8 65 -0.12 0.04 0.00 0.07 -0.17 0.04 7.57”
15 Overall tenderness, 1-8 65 -0.10 0.04 0.00 0.06 -0.15 0.04 7.25“
15 Wamer-Bralzler shearforce, kg 70 0.08 0.05 0.04 0.08 0.18 0.05 5.02“
15 Cook yield,% 75 0.55 0.21 -0.15 0.34 0.42 0.21 3.86
16 24th 81 0.02 0.01 -0.04 0.02 0.03 0.01 4.90
17 Protein, % 23 -0.03 0.07 0.38 0.11 0.04 0.07 3.98
17 45 min-24th decline 51 -0.06 0.02 -0.14 0.06 -0.01 0.03 3.94
18 Last rib backfat, mm 0 1.33 0.54 1.45 0.78 1.36 0.56 4.10
18 24th 0 -0.01 0.01 0.07 0.02 -0.01 0.01 4.60
18 Tenth rib backfat, mm 5 0.21 0.59 3.46 0.97 1.06 0.63 4.77“
18 Spareribsvvt, kg 19 -0.02 0.01 0.01 0.03 -0.05 0.01 3.83
18 Numberofribs 48 -0.11 0.05 0.15 0.08 0.06 0.06 3.45
X Moisture,% 14 -0.36 0.13 NA NA 8.02
X Fat,% 169 0.31 0.11 NA NA 7.68

 

3 Significant at * = 1% chromosome-wise, ** = 5% genome-wise, *** = 1% genome-wise levels

113

Table 1V.4. Position and 95% confidence interval lower and upper limits of
carcass merit QTL significant at 5% genome-wise level.

Position Lower limit Upper limit

 

Chr Trait (cM) (cM) (0M)
1 Spareribs wt, kg 172 146 184
3 45 min pH 91 43 149.5
3 Loin wt, kg 114 88 156
4 Hot carcass weight, kg 12 0 61
4 Off-farm BW, kg 13 0 61
5 24 h carcass temperature, °C 117 12 123
6 a* 24 4 117
6 Marbling, 1-10 116 94 144
6 Fat, % 117 96 144
6 Ham wt, kg 121 113 175
6 Tenth rib backfat, mm 125 118 148
6 Last lumbar vertebra backfat, mm 129 99 146
6 Carcass length, cm 139 99 144
6 Loin wt, kg 141 122 149
6 Last rib backfat, mm 144 0 173.5
6 Protein, % 150 30 164
6 Longissimus muscle area, cm2 156 80.5 166
6 24 h carcass temperature, °C 176 81 181
7 Carcass length, cm 67 41 152
7 Dressing percentage, % 70 56 80
7 Number of ribs 124 106 144

12 Fat, % 57 12 75

12 Marbling, 1-10 67 17 79

14 Belly wt, kg 114 36 117

15 Protein, % 57 46 71

15 L* 60 41 67

15 Overall tenderness, 1-8 65 34 93

15 Tenderness, 1-8 65 33 86.5

 

114

0 45 min-24 h pH decline 0 45 min pH

 

8 1% genome-wise
7 o .
5 /o genome-Wise
7 if v llIi"L" "l'rrr "7 """"
6 : 1% chromosome-Wise 14“ "
’0 I
_ 7 A ~ 3:»... ~ — “r. 4 f -
5 ' L-;.‘ .. .-
° 5% chromosome-wise Luv"? ’. -
.a 4 "W T LLLL .O‘. 7 A—i _ _ a .0 — 7 .' 7 i _ T
I. o. 0
E L‘— o... .. I.
l . ’( ( ’0
I o '0 I
I O O
2 ‘5 O. .0. .I
o , I
‘i .0 ...:f'_r . , .
L. O. 0’ Ir, , illll’fl'ulllllg-
I |. 9 s . L
1 rr 5 o r,- ~ I
Irr' LL 0 fr \‘..I
no r" LLL .0. infl-
' "co/o; riIu'lullli'lll'Ll 0..
O '_ . ':'(’((’(( ((\((\(\‘.\_ , . . . .- .. . . .. . - , l . ._ .. . . .- . _._ .2 L .l. A . .. _. _ -..
E l l A I l T l
0 20 40 60 80 100 120 140 160

Position, cM

Figure IV.l. F-ratio plots versus relative positions on SSC 3. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

115

l6

    
 
 
  
   

14 o Dressing percentage, %
0 Carcass length, cm
12 A Number of ribs

10;

3 L 1% genome-wise i

F-ratio

;5% genome-wise
6 ~~ 1% chromosome-wk

  

-20 F 20 I 40 A 60 80 A 100 120 540 1—16& 180
Position, cM

Figure IV.2. F-ratio plots versus relative positions on SSC 7. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

116

30 ; oMarbllng, 1-10

 
   
 
 

0 Fat, %
A Carcass tenth rib backfat, mm A
25 + Last rib backfat, mm AA
_ )1: Last lumbar backfat, rrm 1:
- A
e 20 AA
‘5 . A
Li. 15 ’
» 1% genome-wise . ,2, g 1‘5 .' , ' -r_

l 5% genome-Wise
5 j 1% chromosome-wise fl
. 5% chromosome-Wise

ru-

‘ . —--"-___
. 9. ”in fun“
. I ‘(\ l
F-‘s .. \‘P '7-
.. l.‘

011. "'1'...

 

 

0 III!) unr.......-- --f--'----'-”' ., «- I . ‘ .‘1 ‘ . _1_-z ‘ ,__. . ‘ ‘ L i
0 20 40 60 80 100 120 140 160 180
Position, cM

Figure IV.3. F-ratio plots versus relative positions on SSC 6. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

117

0 Color, 1-6 D L* A a*
8 . 1% genome-wise

’ 5% genome-wise

”[1,".
"u ‘5A“

     

' 1%chr0mosome-wise , “.27

4 - 5% chromosome-wise ..-' 49."

‘7 ‘_ C I O

F-ra tio

I
I"u""IIIIIIIIIIIIIIIII' ‘

 

Position, cM

Figure 1V.4. F-ratio plots versus relative positions on SSC 15. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

118

9 .

8 ~ 1% genome-wise

5% genome-wise

U Belly wt, kg A Moisture, % 0 Fat, % x Marbling, 1-10

 

 

 

 

 

6 .
e ” 1% chromosome-wise
I; —o‘—
g .
F;- 4 , 5% chromosome-wise
3 . f
2 --
1 .
0 r , *‘
-20 0

f a,

 

, _,L,l_ . ,+_ IL- .,,1 _*. 2.2-- . LL 2.--}
20 40 60 80

Position, cM

Figure IV.5. F-ratio plots versus relative positions on SSC 12. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

119

o Warner-Bratzler shear force, kg 0 Tenderness, 1-8 A Overall tenderness, 1-8
8 ~ 1% genome-wise

 

 

 

I'.' M fl *
I
5:75.32
0 " A ‘
- A ‘-
S/o genome wrse 9,, 3..
# - A #2 _ . . L -. ,- ..- _..__ A- I. A ‘2. # ,7
‘ A‘-
6 I; “I
a}? "I
0 ' P A A
l/o chromosome-Wise flag. ‘3.
__ _ _ __ _ __¢ __ .V A‘ ____ . ‘ ______ _
a .‘.:V.‘ ..4oe(oe.....=='
o- - .'|:-‘ ‘ o. .3:-
"' 5"/ h e- ' 5"“ 6’ ‘~""
a 4 _ 0 C I'OITlOSOl'n Wise I. A‘ A‘
I A o AAVV
h ’ A A‘ F‘.
' r if 7— Il=fl _— ‘ y ‘A- —— T_fiT ‘xvl _
‘1‘ ptt‘ . .O.‘A~..
'2‘ .9 0:4 I.
rri‘ .0 ‘A
A . %
rrrVA e
rrrrrrrrrsififlifix““ . o
2 fi'fiAA““‘3oqoo“'0'“00400(oeoqoo¢...(..uo°‘°

20 40 60 80 100
Position, cM

Figure 1V.6. F-ratio plots versus relative positions on SSC 15. Arrows on x-axis indicate
relative positions of markers on the linkage map. Significance thresholds are indicated
by horizontal lines for 5 and 1% genome- and chromosome-wise significance levels.

120

SUMMARY AND CONCLUSIONS

Creation and analysis of the Michigan State University Duroc x Pietrain F2 pig
resource population allowed for the discovery of quantitative trait loci (QTL) affecting
growth, body composition, and meat quality. In order to supplement these objectives, an
analysis regarding the influence of two different finishing facility types used to rear pigs
within the population was undertaken. These analyses were designed to determine what
phenotypes, if any, may have been influenced by the type of finisher pigs were housed in
during their growth phase. A total of 1259 F2 pigs were born in 11 farrowing groups with
958 pigs completing all growth and meat quality trait collection. These pigs were
managed similarly during farrowing through nursery stages and placed into one of two
different finishers at 10 wk of age. The Modified Open Front building (MOF) had eight
pens, in which four pens differed in size from the other four. Four larger pens (2.03 m by
6.91 m) with two-space feeders were targeted to contain 16 pigs. Four smaller pens (1 .42
m by 6.91 m) with one-space feeders were targeted to contain 12 pigs. Each pen had
two-thirds solid, one-third slatted floors and wet-dry feeders. The Test Station building
(TS) had pens with solid floors that were bedded with straw or wood shavings and had
single-space dry feeders and cup drinkers. All 25 pens utilized in the TS (1.42 m by 4.93
m) were targeted to contain four pigs.

Although 10 to 22 wk ADG of pigs was not different between the two finisher
types, shapes of the growth curve derived from random regression analyses revealed
some differences. Pigs finished in the MOF were heavier at harvest and had more
backfat at 22 wk of age and at harvest at the tenth and last rib than pigs raised in the TS.

Backfat differences as determined through random regression analyses were significantly

12]

different between pigs raised in the two buildings during the test period and continued to
increase through harvest. This increased backfat thickness may have been more
insulating during carcass chilling and led to a trend for higher 24 h carcass temperatures
in the pigs raised in the MOF versus those raised in the TS (3.02 vs. 2.46 °C, P = 0.116).
These higher temperatures may have led to a greater decline in pH from 45 min to 24 h
postmortem for those pigs raised in the MOF than those raised in the TS (0.89 vs. 0.83, P
= 0.019). Most measures of carcass composition and meat quality did not differ between
pigs raised in the two finishing facilities. Ham primal cut weight indicated a trend for
heavier hams in pigs raised in the TS versus the MOF (9.60 vs. 9.48 kg, P = 0.077), while
sensory panel juiciness also indicated a trend for a higher score (more favorable) in TS
raised pigs versus MOF raised pigs (5.34 vs. 5.18, P = 0.064). Warner-Bratzler shear
force was significantly higher in TS raised pigs versus MOF raised pigs (3.36 vs. 3.14 kg,
P = 0.036). The knowledge that finisher facilities influenced some traits was
incorporated into QTL analyses to account for these finisher differences.

The remaining two objectives related to the the detection of QTL. The first
objective was to discover QTL affecting traits that described growth and body
composition, and the second estimated QTL effects for carcass merit and meat quality
traits. Whole blood was collected from all animals for DNA isolation. White blood cells
were separated and frozen for subsequent DNA extraction. A total of 206 dinucleotide
microsatellite markers were considered for genotyping, and 128 informative markers
were selected. Inheritance patterns of four marker genotypes could not be resolved and
were not included in the analysis, therefore 124 markers were utilized encompassing the

entire genome with an average spacing of 24 cM between the markers. A total of 510 F2

122

pigs, their parents, and grandparents were genotyped, and these 510 F2 pigs were sampled
across all farrowing groups from 61 entire litters. Pigs represented all F1 sires with at
least 100 grand progeny from each F0 sire. Fifteen of the 16 F0 dams had a son or
daughter as a parent that produced multiple litters of the selected F2 pigs with the
remaining F0 dam represented by a single F1 daughter with one litter in this group.
Genetic linkage maps were constructed for each of the 18 autosomes and the X
chromosome, and least squares interval mapping was used for QTL detection. Additive,
dominance, and imprinting effects were tested with significance thresholds determined by
permutation tests.

Several measures of growth and body composition were obtained and tested for
significance of QTL associated with the traits measured in the population. Weight at
birth, weaning, and six wk of age was recorded. Additionally, weight and ultrasound
estimates of tenth rib backfat, longissimus muscle area, and last rib backfat were obtained
at three week intervals from 10 to 22 wk of age. From these estimates, body composition
estimates were calculated at each of the three week intervals fi'om 10 to 22 wk of age.
Random regression analyses were performed on these growth data to calculate random
regression terms for each animal. The BLUP animal intercept and animal linear
regression coefficients were predicted and included as traits for QTL discovery. The
QTL analyses for these growth traits revealed 55 QTL for 22 of the 29 measured traits
that were significant at the 5% chromosome-wise significance threshold, and 33 QTL for
15 of the 16 random regression BLUP animal terms that were significant at the 5%
chromosome-wise significance threshold. Highly significant regions with putative QTL

were discovered for tenth rib and last rib backfat on SSC 6, body composition traits on

123

SSC 9, backfat and lipid composition traits on SSC 11, tenth rib backfat and total body
fat tissue on SSC 12, and linear regression of body weight, longissimus muscle area, and
tenth rib backfat on SSC 18.

Carcass composition phenotypes included primal cut weights, skeletal
characteristics, backfat thickness, muscle pH, and carcass temperature. Meat quality data
collected on boneless longissimus muscle chops included objective and subjective color
information, marbling and firmness scores, and drip loss. Additionally, chops were
analyzed for moisture, protein, and fat composition as well as cook yield and Warner-
Bratzler shear force measurements. Palatability of chops was determined by a trained
sensory taste panel. A total of 94 QTL for 35 of the 38 carcass composition and meat
quality traits were found to be significant at the 5% chromosome-wise significance
threshold. Highly significant regions with putative QTL were discovered for 45 min pH
and pH decline on SSC 3, marbling score and carcass backfat on SSC 6, carcass length
and number of ribs on SSC 7, marbling score on SSC 12, and color measurements and
tenderness score on SSC 15.

While important QTL were uncovered for both grth and carcass merit traits,
the antagonistic relationship of these traits makes discovery of regions that control both
sets of traits valuable. An example of this relationship was illustrated by QTL affecting
differing fat depots of backfat and intramuscular fat. Many putative QTL in this study
affected measures of backfat. One region on the genome that affected backfat
measurements on the live animal and carcass backfat measurements was located on SSC
6. While these QTL for measures of subcutaneous fat were located in similar positions

with QTL for marbling on this chromosome, marbling also had a significant QTL on SSC

124

12. Discovering QTL in different regions of the genome for distinct fat depots may
improve the ability to specifically select for amounts of fat in these separate regions.

This selection scheme would allow concurrent genetic improvement in both backfat and
intramuscular fat traits. In general, growth and meat quality QTL were not located in
similar regions of the genome. This presents the opportunity to select for certain
advantageous alleles for one class of traits in one region while selecting for advantageous
alleles for the other class of traits in another region, which would allow improvement to
be made in both types of traits at the same time. These QTL discovery results will
facilitate fine mapping efforts to identify genes controlling growth, body composition,
and meat quality traits that can be incorporated into marker-assisted selection programs to

accelerate genetic improvement in pig populations.

125

LITERATURE CITED

126

LITERATURE CITED

Affentranger, P., C. Gerwig, G. J. F. Seewer, D. Schworer, and N. Kunzi. 1996. Growth
and carcass characteristics as well as meat and fat quality of three types of pigs
under different feeding regimens. Livest. Prod. Sci. 45: 187-196.

AMSA. 1995. Research guidelines for cookery, sensory evaluation and instrumental
measurements of fresh meat. American Meat Science Association and National
Livestock and Meat Board, Chicago, IL.

Andersen, S., and B. Pedersen. 1996. Growth and food intake curves for group-housed
gilts and castrated male pigs. Anim. Sci. 63:457-464.

Andersson, L., C.S. Haley, H. Ellegren, S.A. Knott, M. Johansson, K. Andersson, L.
Andersson-Eklund, I. Edfors-Lilja, M. Fredholm, I. Hansson, J. Hakansson, and

K. Lundstrom. 1994. Genetic mapping of quantitative trait loci for growth and
fatness in pigs. Science. 263:1771-1774.

Andersson-Eklund, L., L. Marklund, K. Lundstrom, C.S. Haley, K. Andersson, I.
Hansson, M. Moller, and L. Andersson, 1998. Mapping quantitative trait loci for

carcass and meat quality traits in a wild boar x Large White intercross. J. Anim.
Sci. 76:694-700.

AOAC. 2000. Official Methods of Analysis (17th Ed.) Association of Official Analytical
Chemists, Gaithersburg, MD.

Barnes, D. J ., J. A. Snedegar, and D. W. Rozeboom. 1999. Comparison of Farmweld
model CR42SASO and Crystal Spring model WR3050 early-wean feeders. MSU
Pork Quarterly. 4(4): 1-3.

Beeckmann P., J. Schroffel Jr., G. Moser, H. Bartenschlager, G. Reiner, and H.
Geldermann. 2003. Linkage and QTL mapping for Sus scrofa Chromosome 3. J.
Anim. Breed. Genet. 120:20-27.

Bertram, H.C., J .S. Petersen, and H.J. Andersen. 2000. Relationship between RN'
genotype and drip loss in meat from Danish pigs. Meat Sci. 56:49-55.

Bidanel, J .P., D. Milan, N. Iannuccelli, Y. Amigues, M.Y. Boscher, F. Bourgeois, J .C.
Caritez, J. Gruand, P. Le Roy, H. Lagant, R. Quintanilla, C. Renard, J. Gellin, L.
Ollivier, and C. Chevalet. 2001. Detection of quantitative trait loci for growth and
fatness in pigs. Genet. Sel. Evol. 33(3):289-309.

Blanchard, P.J., C.C. Warkup, M. Ellis, M.B. Willis, and P. Avery. 1999. The influence
of the proportion of Duroc genes on growth, carcass and pork eating quality
characteristics. Anim. Sci. 68:495-501.

127

Blasco, A., P. Gou, M. Gispert, J. Estany, Q. Soler, A. Diestre and J. Tibau. 1994.
Comparison of five types of pig crosses. 1. Growth and carcass traits. Livest. Prod.
Sci. 40:171-178.

Brewer M.S., J. Jensen, A.A. Sosnicki, B. Fields, E. Wilson, and PK. McKeith. 2002.
The effect of pig genetics on palatability, color and physical characteristics of
fresh pork loin chops, Meat Sci. 61:249-256.

Cassady, J. P., L. D. Young, and K. A. Leymaster. 2002. Heterosis and recombination
effects on pig growth and carcass traits. J. Anim. Sci. 80:2286—2302

Cepica, S., A. Stratil, M. Kopecny, P. Blazkova, J. Schréffel Jr, R. Davoli, L. Fontanesi,
G. Reiner, H. Bartenshlager, and G. Moser. 2003. Linkage and QTL mapping for
Sus scrofa chromosome 4. J. Anim. Breed. Genet. 120:28-37.

Chen, P., T.J. Baas, J .W. Mabry, J .C.M. Dekkers, and K.J. Koehler. 2002. Genetic
parameters and trends for lean growth rate and its components in US. Yorkshire,
Duroc, Hampshire, and Landrace pigs. J. Anim Sci. 80:2062-2070.

Churchill, G.A., and R.W. Doerge. 1994. Empirical threshold values for quantitative trait
mapping. Genetics. 138:963-971.

Cisneros, P., M. Ellis, F.K. McKeith, J. McCaw, and R.L. Fernando. 1996. Influence of
slaughter weight on growth and carcass characteristics, commercial cutting yields,

and meat quality of barrows and gilts from two genotypes. J. Anim. Sci. 74:925-
933.

de Koning, D.J., L.L.G. Janss, A.P. Rattink, P.A.M. Oers, B.J. de Vries, M.A.M.
Groenen, J .J . van der Poel, P.N. de Groot, E.W. Brascamp, and J .A.M. van
Arendonk. 1999. Detection of quantitative trait loci for backfat thickness and
intramuscular fat content in pigs (Sus scrofa). Genetics. 152:1679-1690.

de Koning, D.J., A.P. Rattink, B. Harlizius, M.A.M. Groenen, E.W. Brascamp, and
J .A.M. van Arendonk. 2001. Detection and characterization of quantitative trait
loci for growth and reproduction traits in pigs. Livest. Prod. Sci. 72(3): 1 85-198.

Dekkers, J .C.M. 2003. Principles of QTL mapping. Proc. National Swine Improvement
Federation Mtg. Des Moines, IA.

Dekkers, J .C.M., J .J . Kim, M. Malek, H. Thomsen, H. Lee, H.H. Zhao, and M.
Rothschild. 2003. A genome scan to detect QTL affecting growth, composition,
and meat quality traits in a Berkshire x Yorkshire cross. Proc. National Swine
Improvement Federation Mtg. Des Moines, IA.

Dragos-Wendrich, M., G. Moser, H. Bartenschlager, G. Reiner, and H. Geldermann.
2003. Linkage and QTL mapping for Sus scrofa chromosome 10. J. Anim. Breed.
Genet. 120282-88.

128

Edwards D.B., R.O. Bates, and W.N. Osbum. 2003. Evaluation of Duroc- vs. Pietrain—
sired pigs for carcass and meat quality measures. J. Anim. Sci. 81:1895-1899.

Edwards, D.B. 2005a. Analysis of a Duroc x Pietrain F2 pig resource population for
quantitative trait loci affecting growth, body composition, and meat quality traits.
Ch. 111. PhD. Diss., Michigan State Univ., East Lansing.

Edwards, D.B., R.J. Tempelman, and RD. Bates. 2006. Evaluation of Duroc- vs Pietrain-
sired pigs for growth and composition. J. Anim. Sci. (In press).

Edwards, SA. 2003. Intake of nutrients from pasture by pigs. Proc. Nutr. Soc. 62:257-
265.

Edwards, S.A. 2005b. Product quality attributes associated with outdoor pig production.
Livest. Prod. Sci. 9425-14.

Ellis, M., A.J. Webb, P.J. Avery, and I. Brown. 1996. The influence of terminal sire
genotype, sex, slaughter weight, feeding regime and slaughter-house on growth

performance and carcass and meat quality in pigs and on the organoleptic
properties of fresh pork. Anim. Sci. 62:521-530.

Fernando, R.L., and M. Grossman. 1989. Marker assisted selection using best linear
unbiased prediction. Genet. Sel. Evol. 21:467-477.

Fischer T.M., J .H.J . Van der Werf, R.G. Banks, and A.J. Ball. 2004. Description of lamb
growth using random regression on field data. Livest. Prod. Sci. 89: 175-185.

Fortin, A., J .D. Wood, and OR Whelehan. 1987. Breed and sex effects on the
development and proportions of muscle, fat and bone in pigs. J. Agric. Sci.
108239-45.

Fujii, J ., K. Otsu, F. Zorzato, S. deLeon, V. Khanna, J .E. Weiler, P.J. O'Brien, and DH.
MacLennan. 1991. Identification of a mutation in porcine ryanodine receptor
associated with malignant hypothermia. Science. 253:448-451.

Garcia-Macias, J .A., M. Gispert, M.A. Oliver, A. Diestre, P. Alonso, A. Munoz-Luna, K.
Siggens, and D. Cuthbert-Heavens. 1996. The effects of cross, slaughter weight

and halothane genotype on leanness and meat and fat quality in pig carcasses.
Anim. Sci. 63:487-496.

Geldermann H., E. Muller, G. Moser, G. Reiner, H. Bartenschlager,S. Cepica, A. Stratil,
J. Kuryl, C. Moran, R. Davoli, and C. Brunsch. 2003. Genome-wide linkage and
QTL mapping in porcine F2 families. J. Anim. Breed. Genet. 120:363-393.

Gentry, J .G., J .J . McGlone, J .R. Blanton, Jr., and M.F. Miller. 2002. Alternative housing
systems for pigs: Influences on growth, composition, and pork quality. J. Anim.
Sci. 80:1781-1790.

129

Green, P., K. Falls, and S. Crooks. 1990. Documentation for CRIMAP, version 2.4.
Washington Univ. School of Medicine, St. Louis, MO.

Grindflek, E., J. Szyda, Z.T. Liu, and S. Lien. 2001. Detection of quantitative trait loci for
meat quality in a commercial slaughter pig cross. Mamm. Genome. 12:299-304.

Haley, CS, and SA. Knott. 1992. A simple regression method for mapping quantitative
trait loci in line crosses using flanking markers. Heredity. 69:315-324.

Haley, C.S., S.A. Knott, and J .M. Elsen. 1994. Mapping quantitative trait loci in crosses
between outbred lines using least squares. Genetics. 136:1195-1207.

Hamilton, D.N., M. Ellis, B.F. Wolter, A.P. Schinckel, and ER. Wilson. 2003. The
growth performance of the progeny of two swine sire lines reared under different
floor space allowance. J. Anim. Sci. 81:1126—1135.

Harlizius, B., A.P. Rattink, D.J. de Koning, M. Faivre, R.G. Joosten, J .A.M. van
Aredonk, and M.A.M. Groenen. 2000. The X chromosome harbors quantitative

trait loci for backfat thickness and intramuscular fat content in pigs. Mamm.
Genome. 11:800-802.

Henderson, Jr., OR. 1982. Analysis of covariance in the mixed model: Higher level,
nonhomogeneous, and random regressions. Biometrics. 38:623-640.

Higbie A.D., T.D. Bidner, J .0. Matthews, L.L. Southern, T.G. Page, M.A. Persica, M.B.
Sanders, and OJ. Monlezun. 2002. Prediction of swine carcass composition by
total body electrical conductivity (TOBEC). J. Anim. Sci. 80:113-122.

Hill, W.G. 1999. Advances in quantitative genetics theory. Pages 35-46 in Proc. ISU
Genetics Visions Conf.-From Jay L. Lush to Genomics: Visions for Animal
Breeding and Genetics, Iowa State Univ., Ames.

Houghton, PL, and L.M. Turlington. 1992. Application of ultrasound for feeding and
finishing animals: A review. J. Anim. Sci. 70:930-941.

Hu, Z.L., S. Dracheva, W. Jang, D. Maglott, J. Bastiaansen, M.F. Rothschild, and J .M.
Reecy. 2005. A QTL resource and comparision tool for pigs: PngTLDB. Mamm.
Genome. 15:792-800. Data available: http://www.animalgenome.org/QTLdb.
Accessed Sept. 27, 2005.

Huisman, A.E., R.F. Veerkamp, and J .A.M. van Arendonk. 2002. Genetic parameters for

various random regression models to describe the weight data of pigs. J. Anim.
Sci. 80:575-582.

J amrozik, J ., and LR. Schaeffer. 1997. Estimates of genetic parameters for a test day

model with random regressions for yield traits of first lactation Holsteins. J. Dairy
Sci. 80:762-770.

130

J amrozik, J ., L.R. Schaeffer, and J .C.M. Dekkers. 1997. Genetics and breeding: Genetic

evaluation of dairy cattle using test day yields and random regression model. J.
Dairy Sci. 80:1217-1226.

Jeon J .T., O. Carlborg, A. Tomsten, E. Giuffra, V. Amarger, P. Chardon, L. Andersson-
Eklund, K. Andersson, I. Hansson, K. Lundstrom, and L. Andersson. 1999. A
paternally expressed QTL affecting skeletal and cardiac muscle mass in pigs maps
to the IGF2 locus. Nat. Genet. 21:157-158.

Jeremiah, L.E., J.P. Gibson, L.L. Gibson, R.O. Ball, C. Aker, and A. Fortin. 1999. The
influence of breed, gender, and PSS (Halothane) genotype on meat quality,
cooking loss, and palatability of pork. Food Res. 32:59-71.

Johnson, R.K., E.P. Berg, R. Goodwin, J .W. Mabry, R.K. Miller, O.W. Robison, H.
Sellers, and MD. Tokach. 2004. Evaluation of procedures to predict fat-free lean
in swine carcasses. J. Anim. Sci. 82:2428-2441.

Kanis, E., G.J. Nieuwhof, K.H. de Greef, W. van der Hel, M.W.A. Verstegen, J.
Huisman, and P. van der Wal. 1990. Effect of recombinant procine somatotropin
on growth and carcass quality in growing pigs: Interactions with genotype, gender
and slaughter weight. J. Anim. Sci. 68:1193-1200.

Kennedy, B.W., V.M. Quinton, and C. Smith. 1996. Genetic changes in Canadian

performance-tested pigs for fat depth and growth rate. Can. J. Anim. Sci. 76:41-
48.

Knap, PW. 2000. Time trends of Gompertz growth parameters in "meat-type" pigs.
Anim. Sci. 70:39-49.

Knott, S.A., J .M. Elsen, and CS. Haley. 1996. Methods for multiple marker mapping of
quantitative trait loci in half-sib populations. Theor. Appl. Genet. 93:71-80.

Knott, S.A., L. Marklund, C.S. Haley, K. Andersson, W. Davies, H. Ellegren, M.
Fredholm, I. Hansson, B. Hoyheim, K. Lundstrom, M. Moller, and L. Andersson.
1998. Multiple marker mapping of quantitative trait loci in a cross between
outbred wild boar and large white pigs. Genetics. 149:1069-1080.

Knott S.A., P.E. Nystrom, L. Andersson-Eklund, S. Stern, L. Marklund, L. Andersson,
and CS. Haley. 2002. Approaches to interval mapping of QTL in a

multigeneration pedigree: The example of porcine chromosome 4. Anim. Genet.
33:26-32.

Kuhlers, D.L., S.B. Jungst, and J .A. Little. 1994. An experimental comparison of

equivalent terminal and rotational crossbreeding systems in swine: Pig
performance. J. Anim. Sci. 72:2578-2584.

131

Kuhlers, D.L., K. Nadarajah, S.B. Jungst, B.L. Anderson, and BE. Gamble. 2003.
Genetic selection for lean feed conversion in a closed line of Duroc pigs. Livest.
Prod. Sci. 84:75-82.

Kuhn, G., I. F iedler, K. Ender, and W. Matthes. 2005. Duroc and Pietrain breedings in
comparison. F leischwirtschaft. 85:92-96.

Lander, ES, and D. Botstein. 1989. Mapping Mendelian factors underlying quantitative
traits using RFLP linkage maps. Genetics. 121:185-199.

Langlois, A., and F. Minvielle. 1989a. Comparisons of three-way and backcross swine: 1.
Growth performance and commercial assessment of the carcass. J. Anim. Sci.
67:2018-2024.

Langlois, A., and F. Minvielle. 1989b. Comparisons of three-way and backcross swine:
11. Wholesale cuts and meat quality. J. Anim. Sci. 67:2025-2032.

Lean, I.J., M.K. Curran, J .E. Duckworth, and W. Holmes. 1972. Studies on Belgian
Pietrain pigs. 1. A comparison of Pietrain, Landrace and Pietrain Landrace crosses
in growth, carcass characteristics and meat quality. Anim. Prod. 15:1-9.

Lee, H.K., S.S. Lee, T.H. Kim, G.J. Jeon, H.W. Jung, Y.S. Shin, J .Y. Han, B.H. Choi,
and LC. Cheong. 2003. Detection of imprinted quantitative trait loci (QTL) for
growth traits in pigs. Asian-Aust. J. Anim. Sci. 16:1087—1092.

Malek, M., J .C.M. Dekkers, H.K. Lee, T.J. Baas, and M.F. Rothschild, 2001a. A
molecular genome scan analysis to identify chromosomal regions influencing

economic traits in the pig. 1. Growth and body composition. Mamm. Genome
12(8):630-636.

Malek, M., J .C.M. Dekkers, H.K. Lee, T.J. Baas, K. Prusa, E. Huff-Lonergan, and M.F.
Rothschild, 2001b. A molecular genome scan analysis to identify chromosomal

regions influencing economic traits in the pig. 11. Meat and muscle composition.
Mamm. Genome 12(8):637-645.

Marklund, L., P.E. Nystrom, S. Stern, L. Andersson-Eklund, and L. Andersson. 1998.
Confirmed quantitative trait loci for fatness and growth on pig chromosome 4.
Heredity. 82:134-141.

McKay, R.M., W.E. Rempel, S.G. Cornelius, and CE. Allen. 1985. Growth and feeding
characteristics of three breeds of swine and their crosses. Can. J. Anim. Sci.
65:929-936.

McLaren, D.G., D.S. Buchanan, and R.K. Johnson. 1987a. Growth performance for four

breeds of swine: Crossbred females and purebred and crossbred boars. J. Anim.
Sci. 64:99-108.

132

McLaren, D.G., D.S. Buchanan, and R.K. Johnson. 1987b. Individual heterosis and breed
effects for postweaning performance and carcass traits in four breeds of swine. J.
Anim. Sci. 64:83-98.

McLaren, D.G., F.M. McKeith, and J. Novakofski. 1989. Prediction of carcass
characteristics at market weight from serial real-time ultrasound measures of
backfat and loin eye area in the growing pig. J. Anim. Sci. 67:1657-1667.

Meilgaard, M., G.V. Civille, and B.T. Carr. 1991. Sensory Evaluation Techniques. 1St ed.
CRC Press Inc., Boca Raton, FL.

Meyer, K. 1998. Estimating covariance functions for longitudinal data using a random
regression model. Genet. Sel. Evol. 30:221-240.

Meyer, K. 2004. Scope for a random regression model in genetic evaluation of beef cattle
for growth. Livest. Prod. Sci. 86:69-83.

Mignon-Grasteau, S., C. Beaumont, E. Le Bihan-Duval, J. P. Poivey, H. de Rochambeau,
and F. H. Ricard. 1999. Genetic parameters of growth curve parameters in male
and female chickens. Br. Poult. Sci. 40:44-51.

Mikawa, S., T. Hayashi, M. Nii, S. Shimanuki, T. Morozumi, and T. Awata. 2005. Two
quantitative trait loci on Sus scrofa chromosome 1 and 7 affecting the number of
vertebrae. J. Anim. Sci. 83:2247-2254.

Milan, D., J .P. Bidanel, N. Iannuccelli, J. Riquet, Y. Amigues, J. Gruand, P. Le Roy, C.
Renard, and C. Chevalet. 2002. Detection of quantitative trait loci for carcass
composition traits in pigs. Genet. Sel. Evol. 34:705-728.

Moeller, S.J., L.L. Christian, and RN. Goodwin. 1998. Development of adjustment
factors for backfat and loin muscle area from serial real-time ultrasonic
measurements on purebred lines of swine. J. Anim. Sci. 76:2008-2016.

Mrode, R.A., and B.W. Kennedy. 1993. Genetic variation in measures of food efficiency
in pigs and their genetic relationships with growth rate and backfat. Anim. Prod.
56:225-232.

Nezer, C., L. Moreau, D. Wagenaar, and M. Georges. 2002. Results of a whole genome
scan targeting QTL for growth and carcass traits in a Pietrain x Large White
intercross. Genet. Sel. Evol. 34:371-387.

NPPC. 1991. Procedures to Evaluate Market Hogs. 3rd ed. National Pork Producers
Council, Des Moines, IA.

NPPC. 2000. Pork Composition & Quality Assessment Procedures. 1st ed. Natl. Pork
Prod. Counc., Des Moines, IA.

133

NRC. 1998. Nutrient Requirements of Swine. 10th rev. ed. National Academy Press,
Washington, DC.

Oliver, M.A., P. Gou, M. Gispert, A. Diestre, J. Arnau, J .L. Noguera, and A. Blasco.
1994. Comparison of five types of pig crosses. 11. Fresh meat quality and sensory
characteristics of dry cured ham. Livest. Prod. Sci. 40:179-185.

Ovilo, C., A. Clop, J .L. Noguera, M.A. Oliver, C. Barragan, C. Rodriguez, L. Silo, M.A.
Toro, A. Coll, J .M. Folch, A. Sanchez, D. Babot, L. Varona, and M. Pérez-
Enciso. 2002a. Quantitative trait locus mapping for meat quality traits in an
Iberian x Landrace F2 pig population. J. Anim. Sci. 80:2801-2808.

Ovilo, C., A. Oliver, J .L. Noguera, A. Clop, C. Barragan, L. Varona, C. Rodriguez, M.
Toro, A. Sanchez, M. Pérez-Enciso, and L. Silio. 2002b. Test for positional

candidate genes for body composition on pig chromosome 6. Genet. Sel. Evol.
34:465-479.

Paszek A.A., P.J. Wilkie, G.H. Flickinger, G.A. Rohrer, L.J. Alexander, C.W. Beattie,
and LB. Schook. 1999. Interval mapping of grth in divergent swine cross.
Marnm Genome. 10:117-122.

Paszek, A.A., P.J. Wilkie, G.H. Flickinger, L.M. Miller, C.F. Louis, G.A. Rohrer, L.J.

Alexander, C.W. Beattie, and LB. Schook. 2001. Interval mapping of carcass and
meat quality traits in a divergent swine cross. Anim. Biotechnol. 12: 155-165.

Pérez-Enciso, M., and I. Misztal. 2004. prak: a versatile mixed model application for
genetical genomics and QTL analyses. Bioinformatics. 20:2792-2798.

Pe'rez-Enciso, M., and L. Varona. 2000. Quantitative trait loci mapping in F2 crosses
between outbred lines. Genetics. 155:391-405.

Pérez-Enciso M., A. Clop, J .L. Noguera, C. Ovilo, A. Coll, J .M. Folch, D. Babot, J.
Estany, M.A. Oliver, 1. Diaz, and A. Sanchez. 2000. A QTL on pig chromosome 4

affects fatty acid metabolism: Evidence fiom an Iberian by Landrace intercross. J.
Anim. Sci. 78:2525-2531.

Quiniou, N., and J. Noblet. 1995. Prediction of tissular body composition from protein
and lipid deposition in growing pigs. J. Anim. Sci. 73: 1567-1575.

Quintanilla, R., D. Milan, and J .P. Bidanel. 2002. A further look at quantitative trait loci
affecting growth and fatness in a cross between Meishan and Large White pig
populations. Genet. Sel. Evol. 34: 193-210.

Rathje T.A., G.A. Rohrer, and R.K. Johnson. 1997. Evidence for quantitative trait loci
affecting ovulation rate in pigs. J. Anim. Sci. 75:1486-94.

134

Rattink, A.P., D.J. de Koning, M. Faivre, B. Harlizius, J .A.M. van Arendonk, and
M.A.M. Groenen. 2000. Fine mapping and imprinting analysis for fatness trait
QTLs in pigs. Mamm. Genome. 11:656-661.

Rauw W.M., L. Varona, L.G. Raya, and J .L. Noguera. 2003. Meat production using four
terminal pig lines. J. Sci. Food Agric. 83: 1504-15 10.

Robison, O. W. 1976. Growth patterns in swine. J. Anim. Sci. 42:1024-1035.

Rohrer, G.A., and J .W. Keele. 1998a. Identification of quantitative trait loci affecting
carcass composition in swine: 1. Fat deposition traits. J. Anim. Sci. 76:2247-2254.

Rohrer, G.A., and J .W. Keele. 1998b. Identification of quantitative trait loci affecting

carcass composition in swine: II. Muscling and wholesale product yield traits. J.
Anim. Sci. 76:2255-2262.

Rohrer G.A., L.J. Alexander, Z.L. Hu, T.P.L. Smith, J .W. Keele, and CW. Beattie. 1996.
A comprehensive map of the porcine genome. Genome Res. 6:371-391.

Sato, S., Y. Oyamada, K. Atsuji, T. Nade, S.I. Sato, E. Kobayashi, T. Mitsuhashi, K.
Nirasawa, A. Komatsuda, Y. Saito, S. Terai, T. Hayashi, and Y. Sugimoto. 2003.

Quantitative trait loci analysis for growth and carcass traits in a Meishan x Duroc
F2 resource population . J. Anim Sci. 81 :2938-2949.

Schaeffer, LR, and J .C.M. Dekkers. 1994. Random regressions in animal models for
test-day production in dairy cattle. Proc. 5th World Congress on Genetics Applied
to Livestock Production, Guelph, ON, Canada. 182443-446

Schinckel, AP, and C.F.M. de Lange. 1996. Characterization of growth parameters
needed as inputs for pig growth models. J. Anim. Sci. 74:2021-2036.

Schinckel, A.P., J .R. Wagner, J .C. Forrest, and ME. Einstein. 2001. Evaluation of
alternative measures of pork carcass composition. J. Anim. Sci. 79: 1093-1 1 19.

Schinckel, A.P., C.T. Herr, B.T. Richert, J .C. Forrest, and ME. Einstein. 2003.
Ractopamine treatment biases in the prediction of pork carcass composition. J.
Anim. Sci. 81:16-28.

Seaton, G., GS. Haley, S.A. Knott, M. Kearsey, and RM. Visscher. 2002. QTL Express:
mapping quantitative trait loci in simple and complex pedigrees. Bioinformatics.
18:339-340.

Sellier, P. 1998. Genetics of Meat and Carcass Traits. Pages 463-510 in The Genetics of

the Pig. M. F. Rothschild and A. Ruvinsky, ed. CAB International, New York,
NY.

135

Siemens, A.L., T.B. Erickson, R.J. Lipsey, H.B. Hedrick, D.L. Seevrs, R.O. Bates, J.
Williams, EL, and SW. Yokley. 1989. Predictive equations for estimating lean

cuts, fat standardized lean, chemical composition, bone and value of pork
carcasses. J. Anim. Sci. 67:2033-2039.

Stearns, T.M., J .E. Beever, B.R. Southey, M. Ellis, F.K. McKeith, and S.L. Rodriguez-
Zas. 2005. Evaluation of approaches to detect quantitative trait loci for growth,
carcass, and meat quality on swine chromosomes 2, 6, 13, and 18. I. Univariate
outbred F2 and sib-pair analyses. J. Anim. Sci. 83:1481-1493.

Su Y.H., Y.Z. Xiong, Q. Zhang, S.W. Jiang, L. Yu, M.G. Lei, R. Zheng, and CY. Deng.
2002. Detection of quantitative trait loci for growth in Large White x Meishan
intercross. Yi Chuan Xue Bao. 29:607-611.

Su Y.H., Y.Z. Xiong, S.W. Jiang, Q. Zhang, M.G. Lei, R. Zheng, and CY. Deng. 2004.
Mapping quantitative trait loci for meat quality traits in a Large White x Meishan
cross. Acta Genetica Sinica. 31 :132-136.

Szyda J ., Z. Liu, E. Grindflek, and S. Lien. 2002. Application of a mixed inheritance
model to the detection of quantitative trait loci in swine. J. Appl. Genet. 43:69-83.

Szyda J ., E. Grindflek, Z. Liu, and S. Lien. 2003. Multivariate mixed inheritance models
for QTL detection on porcine chromosome 6. Genet. Res. 81 :65-73.

Terry, C.A., J .W. Savell, H.A. Recio, and HR. Cross. 1989. Using ultrasound technology
to predict pork carcass composition. J. Anim. Sci. 67:1279-1284.

Thompsen, H., H.K. Lee, M.F. Rothschild, M. Malek, and J .C.M. Dekkers. 2004.
Characterization of quantitative trait loci for growth and meat quality in a cross
between commercial breeds of swine. J. Anim. Sci. 82:2213-2228.

USDA. 1998. Hogs, Monthly Slaughter 1974-97. Available at:
http://usda2.mannlib.comell.edu/data—sets/livestock/94 145/11gmo7497.csv.
Accessed Feb. 2, 2001.

USDA. 2005a. Livestock Slaughter 2004 Summary. Available at:
http://usda.mannlib.comell.edu/reports/nassr/livestock/pls-bban/lsan0305.pdf.
Accessed Nov. 27, 2005.

USDA. 2005b. Swine Genome Maps. Available: http://www.marc.usda.gov/genome/
swine/swine.html. Accessed July 25, 2005.

Varona, L., C. Moreno, L.A. Garcia-Cortes, and J. Altarriba. 1997. Multiple trait genetic

analysis of underlying biological variables of production functions. Livest. Prod.
Sci. 47:201-210.

136

Varona, L., C. Ovilo, A. Clop, J .L. Noguera, M. Pérez-Enciso, A. Coll, J .M. Folch, C.
Barragan, M.A. Tom, D. Babot, and A. Sanchez. 2002. QTL mapping for growth
and carcass traits in an Iberian by Landrace pig intercross: Additive, dominant
and epistatic effects. Genet. Res. 80:145-154.

Visscher, P.M., R. Thompson, and CS. Haley. 1996. Confidence intervals for QTL
locations using bootstrapping. Genetics. 143:1013-1020.

Wada, Y., T. Akita, T. Awata, T. Furukawa, N. Sugai, Y. Inage, K. Ishii, Y. Ito, E.
Kobayashi, H. Kusumoto, T. Matsumoto, S. Mikawa, M. Miyake, A. Murase, S.
Shimanuki, T. Sugiyama, Y. Uchida, S. Yanai, and H. Yasue. Quantitative trait
loci (QTL) analysis in a Meishan x Gdttingen cross population. 2000. Anim.
Genet. 3 1 :376-3 84.

Wagner, J .R., A.P. Schinckel, W. Chen, J .C. Forrest, and BL. Coe. 1999. Analysis of
body composition changes of swine during growth and development. J. Anim.
Sci. 77: 1442-1466.

Walling G.A., A.L. Archibald, J .A. Cattermole, A.C. Downing, H.A. Finlayson, D.
Nicholson, P.M. Visscher, C.A. Walker, and CS. Haley. 1998. Mapping of
quantitative trait loci on porcine chromosome 4. Anim. Genet. 29:415-424.

Wang L., T.P. Yu, C.K. Tuggle, H.C. Liu, and M.F. Rothschild. 1998. A directed search
for quantitative trait loci on chromosomes 4 and 7 in pigs. J. Anim. Sci. 76:2560-
2567.

Weller, J .I. 2001. Quantitative trait loci analysis in animals. 1St ed. CABI Publishing,
New York, NY. .

Whittemore, C.T., D.M. Greene, J .D. Wood, A.V. Fisher, and GP. Schofield. 2003.
Physical and chemical composition of the carcass of three different types of pigs
grown from 23 to 115 kg live weight. Anim. Sci. 77:235-245.

Wimmers, K., E. Murani, S. Ponsuksili, M. Yerle, K. Schellander. 2002. Detection of
quantitative trait loci for carcass traits in the pig by using AFLP. Mamm.
Genome. 13:206-210.

Wood, J .D. 1985. Consequences of changes in carcass composition of meat quality.
Pages 157-166 in Recent Advances in Animal Nutrition. W. Haresign and D.J.A.
Cole, ed. Butterworths, London, England.

Young, L.D. 1992a. Effects of Duroc, F engjing, and MinZhu boars on productivity of
mates and growth of first-cross progeny. J. Anim. Sci. 70:2020-2029.

Young, L.D. 1992b. Effects of Duroc, Meishan, Fengjing, and Minzhu boars on carcass
traits of first-cross barrows. J. Anim. Sci. 70:2030-2037.

137

Yu, T.P., C.K. Tuggle, C.B. Schmitz, and M.F. Rothschild. 1995. Association of Pit-1
polymorphism with growth and carcass traits in pigs. J. Anim. Sci. 73:1282—1288.

Zuo, B., Y.Z. Xiong, Y.H. Su, C.Y. Deng, R. Zheng, and S. Jiang. 2003. Mapping
quantitative trait loci for meat quality on pig chromosome 3, 4, and 7. Asian-Aust.
J. Anim. Sci. 16:320-324.

138