Download Mapping of QTL for body conformation and behavior in cattle

Journal of Heredity 2003:94(6):496–506
DOI: 10.1093/jhered/esg090
Ó 2003 The American Genetic Association
Mapping of QTL for Body Conformation
and Behavior in Cattle
S. HIENDLEDER, H. THOMSEN, N. REINSCH, J. BENNEWITZ, B. LEYHE-HORN, C. LOOFT, N. XU,
I. MEDJUGORAC, I. RUSS, C. KÜHN, G. A. BROCKMANN, J. BLÜMEL, B. BRENIG, F. REINHARDT,
R. REENTS, G. AVERDUNK, M. SCHWERIN, M. FÖRSTER, E. KALM, AND G. ERHARDT
Address correspondence to Stefan Hiendleder, Department of Molecular Animal Breeding and Biotechnology,
Ludwig-Maximilian University Munich, Hackerstraße 27, D-85764 Oberschleißheim, Germany, or
email: [email protected].
Abstract
Genome scans for quantitative trait loci (QTL) in farm animals have concentrated on primary production and health traits,
and information on QTL for other important traits is rare. We performed a whole genome scan in a granddaughter design
to detect QTL affecting body conformation and behavior in dairy cattle. The analysis included 16 paternal half-sib families
of the Holstein breed with 872 sons and 264 genetic markers. The markers were distributed across all 29 autosomes and the
pseudoautosomal region of the sex chromosomes with average intervals of 13.9 cM and covering an estimated 3155.5 cM.
All families were analyzed jointly for 22 traits using multimarker regression and significance thresholds determined
empirically by permutation. QTL that exceeded the experiment-wise significance threshold (5% level) were detected on
chromosome 6 for foot angle, teat placement, and udder depth, and on chromosome 29 for temperament. QTL
approaching experiment-wise significance (10% level) were located on chromosome 6 for general quality of feet and legs
and general quality of udder, on chromosome 13 for teat length, on chromosome 23 for general quality of feet and legs, and
on chromosome 29 for milking speed. An additional 51 QTL significant at the 5% chromosome-wise level were distributed
over 21 chromosomes. This study provides the first evidence for QTL involved in behavior of dairy cattle and identifies
QTL for udder conformation on chromosome 6 that could form the basis of recently reported QTL for clinical mastitis.
Genome-wide screening with molecular markers in farm
animal populations has successfully focused on the detection of quantitative trait loci (QTL) for primary
production and health traits such as milk yield and clinical
mastitis in dairy cattle (Georges et al. 1995; Heyen et al.
1999; Klungland et al. 2001; Zhang et al. 1999). After fine
mapping and identification of causative mutations in
positional candidate genes, these QTL have the potential
to achieve additional genetic and economic gains by
incorporating marker-assisted selection into breeding
programs (Farnir et al. 2002; Grisart et al. 2002; Winter et
al. 2002).
496
Breeding programs in cattle and other livestock species
aim at selecting animals with the highest combined economic
value for the next generation. Therefore most breeding
programs select for a combination of production and
nonproduction traits, for example, body conformation and
behavior. Nonproduction traits are either correlated with
traits of economic importance or have a direct economic
value. In cattle, body conformation traits such as stature and
body depth affect feed intake and thus milk production
(Veerkamp 1998), while udder traits correlate with the
incidence of clinical mastitis (Nash et al. 2000) and the length
of productive life (Larroque and Ducrocq 2001). Tempera-
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
From the Institut für Tierzucht und Haustiergenetik der Justus-Liebig-Universität, D-35390 Giessen, Germany (Hiendleder,
Leyhe-Horn, and Erhardt), Institut für Tierzucht und Tierhaltung der Christian-Albrechts-Universität, D-24118 Kiel, Germany
(Thomsen, Reinsch, Bennewitz, Looft, Xu, and Kalm), Institut für Tierzucht der Ludwig-Maximilians-Universität, D-80539
München, Germany (Medjugorac, Russ, Förster), Forschungsinstitut für die Biologie Landwirtschaftlicher Nutztiere,
Forschungsbereich Molekularbiologie, D-18196 Dummerstorf, Germany (Kühn, Brockmann, and Schwerin), Institut für die
Fortpflanzung landwirtschaftlicher Nutztiere, D-16321 Schoenow, Germany (Blümel), Tierärztliches Institut der
Georg-August-Universität, D-37073 Göttingen, Germany (Brenig), Vereinigte Informationssysteme Tierhaltung w.V.,
D-27283 Verden/Aller, Germany (Reinhardt and Reents), and Bayerische Landesanstalt für Tierzucht, D-85586 Grub,
Germany (Averdunk).
Hiendleder et al. Mapping of QTL for Body Conformation and Behavior in Cattle
Materials and Methods
Resource Population and Recorded Traits
The resource population was established within the bovine
genome mapping project of the German Cattle Breeders
Association (ADR, Bonn, Germany). German artificial
insemination (A.I.) organizations contributed 872 semen
samples from 16 paternal half-sib families of the German
Holstein breed. The number of sons per sire ranged from 19
to 127 (average 54.5). In addition, semen from three greatgrandsires was available. A detailed description of this
granddaughter design is given in Thomsen et al. (2001).
Phenotypic data were obtained from the United Data
Systems for Animal Production (VIT, Verden, Germany)
for 22 body conformation and behavior traits (Table 1 and
Figure 1). Trained inspectors or dairy farmers scored the
traits subjectively, the majority (body conformation and
behavior linear traits) on a linear scale from 1 to 9, divided
into nine classes, where 1 and 9 represent biological extremes
and 5 the population mean. Four traits (conformation
general characteristics) were graded using a point system with
a maximum of 100. Estimated breeding values (EBVs) for
body conformation and behavior traits of sons were
calculated by using a Best Linear Unbiased Prediction
(BLUP) animal model. Genetic parameters and breeding
values for QTL analyses were estimated from data of the
same population. All data were taken from the national
breeding value evaluation (November 1999). Detailed
information on traits, data standardization, and the estimation of heritability and breeding values can be found at
http://www.vit.de/English/Homepage.html. The number
of daughters per sire with trait values for body conformation
traits and for behavior traits was 123 on average. Deregressed EBVs were used in QTL analysis as a unit of
phenotypic measurement. The breeding values were standardized to 100 with a standard deviation of 12. Deregression of the EBVs was carried out according to Lien et al.
(1995). The measurement represents the average daughter
deviation of each son from the population mean after
accounting for the fixed effects in the model. Although not
all sons in the evaluation run were provided for genotyping,
they were retained in the QTL analyses to contribute to the
calculation of the fixed effect of the grandsire in the
regression model.
The different number of daughters contributing to the
calculation of deregressed breeding values was accounted for
by a weighting factor, W,
ne
ð1Þ
W ¼
1 þ ðne 1Þ1=4h2 ;
in which the number of effective daughters of each sire
contributing to the calculation of the deregressed breeding
values equals ne, and h2 corresponds to the heritability of the
respective trait.
Marker Genotyping and Linkage Mapping
A whole-genome scan covering all autosomes (chromosomes
1–29) and the pseudoautosomal region of the sex chromosomes (X/Yps) was performed with 264 genetic markers. The
marker panel consisted of 247 microsatellite markers, 8
single-strand conformation polymorphisms (SSCP), 4 protein
polymorphisms, and 5 erythrocyte antigen loci from published marker maps (Barendse et al. 1994, 1997; Bishop et al.
1994; Kappes et al. 1997; Thomsen et al. 2000; Weikard et al.
1997). Microsatellite and SSCP genotypes were determined
by automated fragment analysis (A.L.F. express, AmershamPharmacia; ABI377, Perkin-Elmer). Commercial blood typing laboratories determined erythrocyte antigen genotypes
and protein polymorphisms according to standard procedures. All marker data were transferred to the ADR database
and checked for typing errors by using software that analyzed
the scored genotypes for Mendelian segregation (Reinsch
1999). Marker maps were calculated using the multipoint
option of CRIMAP (Green et al. 1990). Markers, typing
procedures, linkage analysis, and maps have previously been
reported by Thomsen et al. (2000).
Genome-wide Search for QTL
A multiple marker regression approach (Knott et al. 1996)
was used for a QTL scan over 3155.5 cM. QTL mapping was
497
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
ment, defined as the animal’s behavioral responses to
handling by humans (Burrow 1997), is of special importance,
as it determines the amount of time spent in handling
animals. In addition, nervous or aggressive animals have
a decreased milk flow and yield as a result of reduced oxytocin
secretion during stress (Rushen et al. 1999, 2001).
Like primary production traits, body conformation and
behavior traits are of a quantitative nature and display varying
amounts of genetic variation. Heritability estimates in dairy
cattle range from 0.25 to 0.60 for body conformation traits
and from 0.15 to 0.25 for behavior (i.e., temperament)
(Schrooten et al. 2000; Visscher and Goddard 1995).
Selection for these traits has relied solely on phenotypic
and pedigree data, using statistical methods for partitioning
the phenotypic performances of individuals into their
additive genetic values plus environmental contributions.
Previous attempts to detect nonproduction trait QTL in
cattle were limited in power by relatively small family size
(Spelman et al. 1999) or a limited number of markers
(Ashwell et al. 2001). So far only one large genome-wide
search for QTL affecting body conformation and behavior
has been completed with a sufficiently dense marker map
(Schrooten et al. 2000). This study, however, excluded the sex
chromosomes, and ‘‘chromosome-wise’’ significance thresholds were derived from a single chromosome, which was also
used to calculate ‘‘genome-wise’’ significance levels. In the
present study we performed a comprehensive whole-genome
scan with empirically determined chromosome-wise and
experiment-wise significance thresholds to identify bovine
chromosome regions harboring QTL for economically
important body conformation and behavior traits as a first
step toward dissecting the quantitative genetic nature of these
traits and to identify candidate genes for further studies.
Journal of Heredity 2003:94(6)
Table 1.
Heritability (h2)a and description of bovine body conformation and behavior traits investigated for QTL
Trait
h2
Description
Body conformation general
characteristicsb
Body
0.28
Dairy character
0.28
Quality of feet and legs
0.17
Quality of udder
0.22
Conformation and size of the animals trunk in comparison to
an idealized model of a dairy cow
Dairy cow–like appearance in comparison to an idealized
model of a dairy cow
Overall quality of the feet and legs in comparison to an
idealized model of a dairy cow
Overall quality of the udder in comparison to an idealized
model of a dairy cow
Angularity
0.24
Body depth
Foot angle
0.24
0.12
Fore udder attachment
0.21
Hocks
Rear leg set rear view
Rear leg set side view
0.15
0.15
0.15
Rear udder height
Rump angle
0.22
0.26
Rump width
0.28
Stature
Strength
Suspensory ligament
0.41
0.18
0.13
Teat length
Teat placement
0.25
0.22
Udder depth
0.26
Conformation of the silhouette of the animals back, withers
and neck from rear view
Chest height from side view
Height and angle of the horn shoe on the claws
from side view
Angle formed between the fore udder and abdomen,
side view
Diameter of the tarsal joint from rear view
Curvature of the hind legs from rear view
Angle formed at the tarsal joint by upper and lower
hind leg, side view
Distance of rear udder tissue from vulva
Slope of a straight line between cranial and caudal pelvis bone
protrusions, side view
Distance between both caudal pelvis bone protrusions,
rear view
Animals height at the sacrum
Chest width from front view
Upward extension and depth of the udder suspensory
ligament
Length of the front teats from side view
Placement of teats along a straight line from rear to fore
udder from rear view
Distance between udder and tarsal joint from side view
Milking speed
Temperament
0.18
0.07
Time required for milking in relation to yield
Nervous/aggressive or docile behavior during milking
Body conformation linear
characteristicsc
Behavior linear characteristicsc
a
Population data (http://www.vit.de/English/Homepage.html).
b
Subjectively graded by a trained inspector on a point system with maximum 100.
c
Subjectively scored by a trained inspector or dairy farmer on a linear scale from 1 to 9.
performed for each trait separately, as described by Reinsch
et al. (1999). The most likely marker haplotype for each
chromosome of each grandsire was determined from marker
genotypes of all typed animals in the granddaughter design,
that is, all typed progeny of each grandsire and, where
available, the marker genotypes of the great-grandsires. The
number of genotyped progeny ranged from 19 to 127 per
family. In all three-generation families, the use of greatgrandsire marker information reduced the number of
possible grandsire haplotypes considerably. Using these
haplotypes, the probability that a sire inherited the first allele
of a possible QTL was derived from the smallest informative
marker interval available for each sire every centiMorgan on
each chromosome. For the analysis of 872 progeny, about 2.7
million QTL transition probabilities were required. A joint
498
analysis of all families was undertaken by fitting every
centiMorgan the following regression model to the data:
ð2Þ
yijk ¼ gsi þ bik pijk þ eijk ;
where yijk is the trait value for the jth son of the ith grandsire,
gsi is the fixed effect of the ith family (grandsire), bik is the
regression coefficient for the ith family at the kth position,
pijk is the probability that the jth son has received the first
QTL allele from the ith grandsire at the kth chromosomal
position, and eijk is the random residual. For each single
position in the genome, the null hypothesis that all bik are
equal to zero was tested. The test statistic was the F ratio of
pooled mean squares due to regression within grandsires to
residual mean square. The peak of the test statistic on
a chromosome was considered to be the most likely position
of a QTL. Chromosome-wise and experiment-wise signifi-
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
Trait category
Hiendleder et al. Mapping of QTL for Body Conformation and Behavior in Cattle
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
Figure 1. Graphical representation of body conformation traits described in Table 1 as classified according to the linear
scale used in phenotyping. The scale ranges from 1 to 9, classes 1 and 9 representing biological extremes.
499
Journal of Heredity 2003:94(6)
cance thresholds were derived for each trait by a permutation
test (Churchill and Doerge 1994) with 10,000 shuffles of the
original data. The need for the estimation of unique
permutation thresholds to determine critical a levels was
previously stressed by others (e.g., Cassady et al. 2001). All
computations were done with the BIGMAP and ADRQTL
programs, which are directly connected to the ADR database
(Reinsch 1999), on a SUN sparc SUNW, Ultra-1 workstation. The programs are available upon request (reinsch@
fbn-dummerstorf.de).
Table 2.
Average Average no.
Chromosome marker of informative
No.
interval meioses/
length
of
marker
(cM)
Chromosome markersa (cM)b
500
14
11
14
9
10
10
10
11
5
8
12
9
13
8
10
8
9
7
12
5
9
6
15
8
5
5
8
5
5
2
264
190.5
155.2
140.9
163.0
136.9
127.7
140.2
146.0
93.0
95.0
113.0
121.6
142.0
139.4
109.2
95.0
92.0
117.5
130.1
71.0
88.0
83.5
84.0
97.0
73.0
49.0
53.4
35.0
64.4
9.0
3155.5
14.7
15.5
10.8
20.4
15.2
14.2
15.6
14.6
23.2
13.6
10.3
15.2
11.8
15.4
12.1
13.6
11.5
19.6
11.8
17.8
11.0
16.7
6.0
13.9
18.3
12.3
7.6
8.8
16.1
9.0
13.9
450.9
455.1
449.1
534.3
330.3
384.4
466.1
511.9
527.4
572.3
449.9
524.5
587.9
371.4
398.9
662.5
524.9
593.7
454.3
492.0
606.0
423.3
490.8
378.6
468.6
512.4
484.8
395.2
445.2
422.5
479.0
a
Detailed marker information is given in Thomsen et al. (2000).
b
Map distance was calculated using the Haldane map function.
cM, on chromosome 18 at 105/109 cM, on chromosome 29
at 20/20 cM, and on chromosome X/Y at 9/9 cM or QTL for
hocks and rear leg set rear view (rG ¼ 0.19) on chromosome
11 at 61/61 cM, on chromosome 13 at 53/54 cM, and on
chromosome 21 at 12/12 cM (Table 3 and Figure 2).
Discussion
We have detected 60 QTL significant at the 5% chromosome-wise level for 22 body conformation and behavior traits
with estimated heritabilities ranging from 0.07–0.41 (Tables 1
and 3), demonstrating that a considerable number of loci
with effects on body conformation and behavior are still
segregating in the Holstein cattle population. These traits
are of economic importance and have been under phenotypic selection with differing intensity for several generations. Therefore negative, albeit nonsignificant (P . .05),
regression coefficients for the number of QTL per trait on
heritability estimates of the population are expected (b ¼ 2.66
for 60 QTL at the 5% chromosome level, and b ¼ 3.29 when
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
X/Yps
Total
Results
The genetic marker map covered 3155.5 cM of the bovine
genome across all autosomes and the pseudoautosomal
region of the sex chromosomes. The average marker interval
was 13.9 cM and ranged from 6.0 cM (chromosome 23) to
23.2 cM (chromosome 9) on individual chromosomes (Table
2). A total of 60 QTL with chromosome-wise significance
(5% level) were identified on 20 autosomes and the sex
chromosomes. Significant QTL were detected for all but one
(rear leg set side view) of the 22 investigated body
conformation and behavior traits (Table 3). The QTL for
foot angle, teat placement, and udder depth on chromosome
6 and for temperament on chromosome 29 exceeded the
experiment-wise significance threshold at the 5% level. QTL
for general quality of the feet and legs and general quality of
the udder on chromosome 6, for teat length on chromosome
13, for general quality of the feet and legs on chromosome 23,
and for milking speed on chromosome 29 approached
experiment-wise significance at the 10% significance level
(Figure 2). Only nine chromosomes (3, 4, 12, 14, 15, 19, 24, 26,
and 28) did not harbor any QTL, but four chromosomes (6, 5,
18, and X/Y) with 10, 7, 6, and 5 QTL, respectively, contained
almost 50% of the total QTL detected. Chromosomes with
and without QTL did not differ in average marker interval
(14.1 ^ 0.85 versus 13.4 ^ 1.3, P ¼ .662) or the average
number of informative meioses per marker (492.4 ^ 16.5
versus 446.1 ^ 25.2, P ¼ .135). This is also evident from
comparisons between marker characteristics of individual
chromosomes, for example, comparisons between the QTLrich chromosomes 5 or 6 with the QTL-poor chromosomes 3
or 12 (Table 2). QTL detection was therefore not related to
marker number, marker distribution, or marker information
content. Chromosome 6 showed a remarkable QTL cluster
for eight body conformation traits at positions 85–89 cM, five
of them exceeding (5% level) or approaching (10% level)
experiment-wise significance. The QTL cluster includes three
very diverse groups of genetically correlated traits. These are
suspensory ligament, teat placement, udder depth, and
general quality of the udder (rG ¼ 0.25–0.74), foot angle
and general quality of the feet and legs (rG ¼ 0.58), and body
and rump width (rG ¼ 0.22) (Table 3 and Figure 2). In addition
to the QTL on chromosome 6, other chromosomes also
showed QTL for correlated traits to be present in the same
chromosome regions. An example are QTL for temperament
and milking speed (rG ¼ 0.53) on chromosome 5 at 136/136
Characteristics of the genome coverage
Hiendleder et al. Mapping of QTL for Body Conformation and Behavior in Cattle
Table 3. Chromosomal location, test statistic (F values), and significance (P values) of putative QTL for body conformation and
behavior in dairy cattle
Chromosome
1
Traita
Position
(cM)
Test statistic
(F value)
Chromosome-/
experiment-wise
thresholds (F values)b
Chromosome-/
experiment-wise
P valuesc
No. of
heterozygous
siresd
Body
Rump width
Rear udder height
135
142
145
2.035
2.392
3.244
2.029/2.454
2.008/2.444
3.207/3.630
0.049
0.006
0.046
3
4
1
Angularity
17
1.705
1.699/2.121
0.048
3
Foot angle
Teat placement
Suspensory ligament
Quality of the feet and legs
Udder depth
Milking speed
Temperament
72
97
98
101
106
136
136
2.323
2.134
2.100
2.305
2.093
2.153
2.312
2.026/2.570
1.903/2.405
2.101/2.603
1.965/2.521
2.075/2.627
1.922/2.551
2.014/2.661
0.014
0.015
0.050
0.010
0.046
0.019
0.015
4
4
4
3
2
2
3
6
Stature
Strength
Body
Rump width
Suspensory ligament
Teat placement
Foot angle
Quality of udder
Quality of the feet and legs
Udder depth
66
70
85
87
88
88
88
89
89
89
2.018
1.746
2.297
2.144
2.474
2.646
3.062
2.544
2.585
3.105
1.872/2.375
1.762/2.301
1.936/2.454
1.923/2.444
2.082/2.603
1.886/2.405
2.009/2.570
1.965/2.491
1.979/2.521
2.098/2.627
0.023
0.054
0.006
0.017
0.007
0.002/0.029
0.000/0.008
0.000/0.008
0.003/0.072
0.000/0.008
3
2
4
3
3
5
4
3
4
5
7
8
9
10
Teat length
Body depth
Angularity
Milking speed
18
68
59
31
2.307
2.032
1.600
1.947
2.232/2.769
1.942/2.429
1.590/2.121
1.924/2.551
0.036
0.034
0.048
0.046
2
2
2
3
11
Hocks
Rear leg set rear view
61
61
2.516
2.639
2.350/3.029
2.641/3.427
0.029
0.050
4
4
13
Teat length
Strength
Hocks
Rear leg set rear view
39
51
53
54
2.773
1.970
2.502
3.009
2.256/2.769
1.879/2.301
2.383/3.029
2.665/3.427
0.006/0.098
0.032
0.033
0.017
3
3
3
4
16
17
Rear udder height
Foot angle
Rump angle
Quality of the feet and legs
92
5
34
34
3.167
2.110
2.506
1.962
2.628/3.630
1.981/2.570
2.043/2.634
1.939/2.521
0.010
0.029
0.005
0.046
1
3
2
2
18
Body
Stature
Teat placement
Dairy character
Temperament
Milking speed
0
0
43
82
105
109
1.938
2.260
1.841
1.874
2.364
2.254
1.891/2.454
1.862/2.375
1.855/2.405
1.854/2.389
2.002/2.661
1.901/2.551
0.040
0.005
0.053
0.046
0.012
0.010
3
4
2
2
4
6
20
21
Teat placement
Foot angle
Hocks
Rear leg set rear view
2
6
12
12
1.983
2.217
2.352
2.670
1.771/2.405
2.014/2.570
2.307/3.029
2.609/3.427
0.018
0.019
0.042
0.042
2
1
4
4
22
23
Rump width
Teat length
Rear udder height
Quality of the feet and legs
Foot angle
52
82
84
84
84
2.007
2.647
2.535
2.526
2.190
1.929/2.444
2.149/2.769
2.506/3.630
1.945/2.521
2.008/2.570
0.037
0.006
0.045
0.004/0.098
0.022
2
4
4
1
2
25
Body depth
Teat length
33
33
2.322
2.027
1.777/2.429
2.047/2.769
0.003
0.054
3
3
27
29
Fore udder attachment
Milking speed
Temperament
34
20
20
2.663
2.514
2.869
2.452/3.508
1.773/2.551
1.844/2.661
0.028
0.002/0.116
0.000/0.044
2
3
4
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
2
5
501
Journal of Heredity 2003:94(6)
Table 3.
Continued
Chromosome
Traita
Position
(cM)
XY
Body
Rump width
Stature
Temperament
Milking speed
0
0
4
9
9
Test statistic
(F value)
Chromosome-/
experiment-wise
thresholds (F values)b
Chromosome-/
experiment-wise
P valuesc
No. of
heterozygous
siresd
1.986
1.696
1.788
1.376
1.418
1.651/2.454
1.712/2.444
1.599/2.375
1.401/2.661
1.384/2.551
0.013
0.053
0.021
0.054
0.043
2
2
4
a
For a detailed description of traits see Table 1 and Figure 1.
b
Thresholds were calculated using the method described by Churchill and Doerge (1994).
2
c
Only chromosome-wise P values less than 0.05 and experiment-wise P values less than 0.12 are shown.
d
Out of a total number of 16 sires.
502
for milking speed at 57/72 cM on chromosome 2 only
approached chromosome-wise significance at P , .072 (data
not shown, available upon request) in our study, stressing the
need to report QTL at suggestive thresholds or make
nonsignificant data available. A QTL for stature was mapped
to chromosome 6 at 11 cM by Schrooten et al. (2000) and to
66 cM in the present study. The QTL for ‘‘chest width’’
reported by Schrooten et al. (2000) on chromosome 2 at 139
cM could be identical to the QTL for the trait ‘‘body depth’’
reported at 82 cM in the present study (chromosome-wise
significance P , .094; data not shown, available upon
request). The data presented by Ashwell et al. (2001) did not
show any indication of QTL shared with our experiment,
perhaps due to different traits or trait definitions.
While this manuscript was processed, Boichard et al.
(2003) reported QTL for nonproduction traits in three dairy
cattle breeds. This granddaughter design combined 1548 sires
of the Holstein, Montbéliarde, and Normande breeds in 14
families and is similar to our study in the Holstein breed.
Boichard et al. (2003) used an approximately 30% lower
marker density than the present study, but nevertheless
detected 86 QTL for 15 body conformation and one
behavior trait at chromosome-wise and/or experiment-wise
significance levels determined empirically by permutation.
Although the overlap in traits and especially trait definitions
used for the QTL searches by Boichard et al. (2003) and the
present study is limited, a considerable number of QTL were
identified in both studies and are therefore confirmed. A
QTL involved in chest width/depth was mapped by
Schrooten et al. (2000) to chromosome 2 at 139 cM, at 82
cM in the present study (see above), and at 40 cM by Boichard
et al. (2003). The QTL for milking speed at 72 cM on
chromosome 5 detected in our study was also identified by
Boichard et al. (2003) at 97 cM. The QTL for stature detected
by Schrooten et al. (2000) on chromosome 6 at 11 cM and at
66 cM in the present study was mapped to 54 cM by Boichard
et al. (2003). On chromosome 6, four additional QTL
reported by Boichard et al. (2003) were confirmed in our
study. These are QTL for rump width at 62/87 cM, for
stature at 54/66 cM, for teat placement at 87/88 cM, and for
udder depth at 147/89 cM (position reported by Boichard
et al. (2003) presented first). Further QTL reported by
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
an additional 40 putative QTL at the 10% chromosome level
are included) because of fixation of favorable alleles. The
proportion of genetic variance explained by nine QTL with
a genome-wise P value of less than 10% ranged from 1.6%
to 2.5% (Table 4). It has to be considered, however, that in
our experimental design these estimates are derived from
a combined analysis of families with heterozygous and
homozygous grandsires, and the number of heterozygous
sires among the total of 16 grandsires (Table 3) was generally
small (mean 2.8, range 1–6). Average allele substitution
effects of these QTL ranged from 0.71 to 3.60 genetic
standard deviations (Table 4) and are in line with allele
substitution effects for nonproduction trait QTL reported by
others (e.g., Boichard et al. 2003).
The number of QTL for body conformation and
behavior identified in previous studies (Ashwell et al. 2001;
Schrooten et al. 2000; Spelman et al. 1999) is not directly
comparable to the present study because of differences in the
number and type of traits analyzed, differences in experimental design, and differences in significance thresholds
employed. Using a similar spectrum of 17 traits in a study
with less power because of a smaller number of families and
a smaller number of sons per family analyzed, Spelman et al.
(1999) reported a single QTL for stature on chromosome 14
at the 15% experiment-wise error level. The highly genetically
correlated trait of body weight peaked at the same position,
but did not reach the suggestive significance level. Other
studies (Ashwell et al. 2001; Schrooten et al. 2000) and our
study have not confirmed this QTL. Schrooten et al. (2000)
reported a total of nine QTL exceeding an experiment-wise
significance threshold of 10% and an additional 36 QTL
exceeding a chromosome-wise threshold of 0.0345% for 20
traits scored very similar to the traits used in our study.
However, the experiment-wise significance threshold was not
derived empirically for each trait with all chromosomes, but
was calculated from permutation data of a single ‘‘average’’
chromosome, which was then used to calculate ‘‘genomewise’’ significance levels. Chromosome-wise thresholds were
also derived from the single ‘‘average’’ chromosome
permutation data. Nevertheless, our study confirmed a QTL
for udder depth on chromosome 5 at 109/106 cM (position
reported by Schrooten et al. (2000) presented first). A QTL
Hiendleder et al. Mapping of QTL for Body Conformation and Behavior in Cattle
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
Figure 2. Across-sire family F values for QTL effects on body conformation and behavior traits in Holstein cattle along
chromosomes 6, 13, 23, and 29. Only QTL significant at the experiment-wise threshold (P , .05) or approaching experiment-wise
significance (P , .10) are shown. The 5% threshold is indicated. Thresholds for foot angle and general quality of the feet and
legs on chromosome 6 (F values 2.49 and 2.52) have been merged. All significance thresholds were determined empirically by
permutation (Churchill and Doerge 1994).
Boichard et al. (2003) that are likely to be identical and thus
confirmed in the present study are found on chromosome 9
(overall quality of udder at 48/63 cM), chromosome 13 (teat
length at 0/39 cM), chromosome 17 (rump angle at 8/34
cM), chromosome 18 (foot angle at 74/36 cM, rear udder
height at 98/52 cM, teat placement at 115/43 cM), and
chromosome 20 (suspensory ligament at 9/0 cM, overall
quality of udder at 48/9 cM). Although three confirmed QTL
503
Journal of Heredity 2003:94(6)
504
Table 4. Estimates of the proportion of genetic variance
(r2g) explained and average substitution effects (rg) of QTL
exceeding or approachinga an experiment-wise significance
threshold of 10%
Trait
Teat placement
Foot angle
Quality of udder
Quality of the feet
and legs
Udder depth
Teat length
Quality of the feet
and legs
Milking speed
Temperament
a
Position
(chromosome/cM)
r2g (%)
rg
06/88
06/88
06/89
06/89
2.1
2.2
1.7
1.6
0.71
1.40
0.76
1.00
06/89
13/39
23/84
2.5
2.1
2.0
0.75
1.10
0.96
29/20
29/20
1.9
2.3
2.91
3.60
Milking speed, P , .116.
therefore reduced milking speed as a result of reduced
oxytocin secretion (Rushen et al. 1999, 2001).
Relatively few coding genes are yet mapped in cattle, and
no compelling candidate genes are evident for the QTL
affecting udder traits on chromosomes 6 and 13 or leg traits
on chromosomes 6 and 23. However, new comparative
maps, such as the high-resolution integrated and comparative
map of bovine chromosome 6, which accesses information
from orthologous human and mouse chromosomes (Weikard et al. 2002), should enable identification of candidate
genes. On chromosome 29, the tyrosinase (TYR) gene maps
to the QTL region for temperament and milking speed with
the highest test statistic (Schmidtz et al. 2001; Schmutz and
Moker 1999). Tyrosinase is a multifunctional enzyme
involved in metabolism of the neurotransmitter dopamine.
It catalyzes the conversion of tyrosine to dihydroxyphenylalanine (DOPA), the precursor of dopamine, and oxidizes
DOPA to form DOPA quinone. Tyrosinase also oxidizes
dopamine to form melanin via dopamine quinone, which has
been shown to inactivate tyrosine hydroxylase, the ratelimiting enzyme for dopamine synthesis (Higashi et al. 2002;
Xu et al. 1998). The tyrosinase gene is therefore a QTL
candidate for temperament in cattle.
Acknowledgments
This study was supported by the German Ministry of Education, Science,
Research and Technology (BMBF) and the German Cattle Breeders
Federation (ADR). The technical assistance of the staffs of all participating
laboratories is gratefully acknowledged.
References
Ashwell MS, Van Tassell CP, and Sonstegard TS, 2001. A genome scan to
identify quantitative trait loci affecting economically important traits in a US
Holstein population. J Dairy Sci 84:2535–2542.
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
(teat placement and udder depth on chromosome 6, teat
length on chromosome 13) were significant at the genomewise 10% threshold in our study, 7 of 11 additional QTL
assumed to be confirmed reached only the suggestive
chromosome-wise significance level (P , .10). This again
stresses the need to report putative QTL at suggestive
thresholds.
Klungland et al. (2001) recently reported a QTL for
clinical mastitis with experiment-wise significance on bovine
chromosome 6 in the same approximate chromosome region
where we identified several highly significant QTL affecting
udder conformation traits (teat placement, udder depth,
overall udder quality) and leg conformation traits (overall
quality of the feet and legs, foot angle). In addition, we
detected QTL for udder conformation traits at the
chromosome-wise suggestive level (P , .10) on chromosomes 8 and 27 (data not shown), two chromosomes
reported by Klungland et al. (2001) to harbor QTL for udder
health at the chromosome-wise significance level. Recent
heritability estimates for the incidence of clinical mastitis and
the number of clinical episodes range from 0.11 to 0.25 and
0.15 to 0.45, respectively, and relationships between clinical
mastitis and udder conformation traits have been well
established (Nash et al. 2000). The QTL for udder
conformation traits could therefore form the basis for the
mastitis QTL on chromosomes 6, 8, and 27. QTL for leg
conformation on chromosome 6, however, might also
contribute to mastitis susceptibility, since animals with poor
feet could be more prone to infections, perhaps because of
longer bedding times and associated microbial udder
contamination.
Previous studies attempting to localize QTL for temperament in dairy cattle (Schrooten et al. 2000; Spelman et al.
1999) were unsuccessful. We have detected one QTL for
temperament significant at the experiment-wise level and two
QTL significant at the chromosome-wise level. An additional
three QTL (chromosomes 10, 20, and X/Y) were significant
at the chromosome-suggestive level (P , .10, data not
shown). This is the first evidence for QTL affecting behavior
in dairy cattle. Schmutz et al. (2001) reported QTL for
behavior traits in beef cattle on chromosomes 1, 5, 9, 11, 14,
and 15. Although these data are not directly comparable
because of differences in experimental design and a different
trait definition for temperament (response to isolation), it is
possible that the QTL for temperament on chromosome 5 in
dairy cattle is identical to the QTL on chromosome 5 for
temperament in beef cattle. Schrooten et al. (2000) reported
two suggestive QTL for milking speed on chromosomes 3
and 23, but neither chromosome showed QTL evidence for
milking speed or temperament in our study. In agreement
with the high genetic correlation between both traits, we
detected QTL for milking speed at an almost equal frequency
and, when suggestive QTL with a chromosome-wise P value
of less than .10 are included, always at identical or nearly
identical chromosome positions. Milk secretion is directly
dependent on oxytocin, a hormone sensitive to various types
of emotional stress (Wellnitz and Bruckmaier 2001), and
stressed or nervous animals have a delayed milk flow and
Hiendleder et al. Mapping of QTL for Body Conformation and Behavior in Cattle
Barendse W, Armitage SM, Kossarek LM, Shalom A, Kirkpatrick BW, Ryan
AM, Clayton D, Li L, Neibergs HL, Zhang N, Grosse WM, Weiß J,
Creighton P, McCarthy F, Ron M, Teale AJ, Fries R, McGraw RA, Moore
SS, Georges M, Soller M, Womack JE, and Hetzel DJS, 1994. A genetic
linkage map of the bovine genome. Nat Genet 6:227–235.
Klungland H, Sabry A, Heringstad B, Olsen HG, Gomez-Raya L, Våge DI,
Olsaker I, Ødegård J, Klemetsdal G, Schulman N, Vilkki J, Ruane J,
Aasland M, Rønningen K, and Lien S, 2001. Quantitative trait loci affecting
clinical mastitis and somatic cell count in dairy cattle. Mamm Genome
12:837–842.
Barendse W, Vaiman D, Kemp SJ, Sugimoto Y, Armitage SM, Williams JL,
Sun HS, Eggen A, Agaba M, Aleyasin SA, Band M, Bishop MD, Buitkamp
J, Byrne K, Collins F, Cooper L, Coppieters W, Denys B, Drinkwater RD,
Easterday K, Elduque C, Ennis S, Erhardt G, Ferretti L, Flavin N, Gao Q,
Georges M, Gurung R, Harlizius B, Hawkins G, Hetzel J, Hirano T, Hulme
D, Jorgensen C, Keßler M, Kirkpatrick BW, Konfortov B, Kostia S, Kühn
C, Lenstra JA, Leveziel H, Lewin HA, Leyhe B, Li, Martin Burriel I,
McGraw RA, Miller JR, Moody DE, Moore SS, Nakane S, Nijman IJ,
Olsaker I, Pomp D, Rando A, Ron M, Shalom A, Teale AJ, Thieven U,
Urquhart BDG, Vage D-I, Van de Weghe A, Varvio S, Velmala R, Vilkki J,
Weikard R, Woodside C, Womack JE, Zanotti M, and Zaragoza P, 1997. A
medium-density genetic linkage map of the bovine genome. Mamm
Genome 8:21–28.
Knott SA, Elsen JM, and Haley CS, 1996. Methods for multiple-marker
mapping of quantitative trait loci in half-sib populations. Theor Appl Genet
93:71–80.
Larroque H and Ducrocq V, 2001. Relationship between type and longevity
in the Holstein breed. Genet Sel Evol 33:39–59.
Lien S, Gomez-Raya L, Steine T, Fimland E, and Rogne S, 1995.
Associations between casein haplotypes and milk yield traits. J Dairy Sci
78:2047–2056.
Nash DL, Rogers GW, Cooper JB, Hargrove GL, Keown JF, and Hansen
LB, 2000. Heritability of clinical mastitis incidence and relationships with
sire transmitting abilities for somatic cell score, udder type traits, productive
live, and protein yield. J Dairy Sci 83:2350–2360.
Reinsch N, 1999. A multiple species, multiple project database for
genotypes at codominant loci. J Anim Breed Genet 116:425–435.
Boichard D, Grohs C, Bourgeois F, Cerqueira F, Faugeras R, Neau A, Rupp
R, Amigues Y, Boscher MY, and Leveziel H, 2003. Detection of genes
influencing economic traits in three French dairy cattle breeds. Genet Sel
Evol 35:77–101.
Reinsch N, Thomsen H, Xu N, Brink M, Looft C, Kalm E, Brockmann
GA, Grupe S, Kühn C, Schwerin M, Leyhe B, Hiendleder S, Erhardt G,
Medjugorac I, Russ I, Förster M, Reents R, and Averdunk G, 1999. A QTL
for the degree of spotting in cattle shows synteny with the KIT locus on
chromosome 6. J Hered 90:629–634.
Burrow HM, 1997. Measurements of temperament and their relationships
with performance traits of beef cattle. Anim Breed Abstr 65:477–495.
Cassady JP, Johnson RK, Pomp D, Rohrer GA, Van Vleck LD, Spiegel EK,
and Gilson KM, 2001. Identification of quantitative trait loci affecting
reproduction in pigs. J Anim Sci 79:623–633.
Rushen J, De Passille AM, and Munksgaard L, 1999. Fear of people by cows
and effects on milk yield, behavior, and heart rate at milking. J Dairy Sci
82:720–727.
Churchill GA and Doerge RW, 1994. Empirical threshold values for
quantitative trait mapping. Genetics 138:963–971.
Rushen J, Munksgaard L, Marnet PG, and De Passille AM, 2001. Human
contact and the effects of acute stress on cows at milking. J Dairy Sci
82:720–727.
Cooper JR, Bloom FE, and Roth RH, 1991. Dopamine. In: The
biochemical basis of pharmacology (Cooper JR, Bloom FE, and Roth
RH, eds). New York: Oxford University Press; 285–337.
Schmidtz BH, Buchanan FC, Plante Y, and Schmutz SM, 2001. Linkage
mapping of the tyrosinase gene to bovine chromosome 29. Anim Genet
32:119–120.
Farnir F, Grisart B, Coppieters W, Riquet J, Berzi P, Cambisano N, Karim
L, Mni M, Moisio S, Simon P, Wagenaar D, Vilkki J, and Georges M, 2002.
Simultaneous mining of linkage and linkage disequilibrium to fine map
quantitative trait loci in outbred half-sib pedigrees: revisiting the location of
a quantitative trait locus with major effect on milk production on bovine
chromosome 14. Genetics 161:275–287.
Schmutz SM, and Moker JS, 1999. In situ hybridization mapping of TYR
and CCND1 to cattle chromosome 29. Anim Genet 30:241–242.
Georges M, Nielsen D, Mackinnon M, Mishra A, Okimoto R, Pasquino AT,
Sargeant LS, Sorensen A, Steele MR, Zhao X, Womack JE, and Hoeschele
I, 1995. Mapping quantitative trait loci controlling milk production in dairy
cattle by exploiting progeny testing. Genetics 139:907–920.
Green P, Falls K, and Crooks S, 1990. Documentation of CRI-MAP,
Version 2.4. St. Louis: Washington University School of Medicine.
Grisart B, Coppieters W, Farnir F, Karim L, Ford C, Berzi P, Cambisano N,
Mni M, Reid S, Simon P, Spelman R, Georges M, and Snell R, 2002.
Positional candidate cloning of a QTL in dairy cattle: Identification of
a missense mutation in the bovine DGAT1 gene with major effect on milk
yield and composition. Genome Res 12:222–231.
Heyen DW, Weller JI, Ron M, Band M, Beever JE, Feldmesser E, Da Y,
Wiggans GR, VanRaden PM, and Lewin HA, 1999. A genome scan for
QTL influencing milk production and health traits in dairy cattle. Physiol
Genomics 1:165–175.
Schmutz SM, Stookey JM, Winkelman-Sim DC, Waltz CS, Plante Y, and
Buchanan FC, 2001. A QTL study of cattle behavioral traits in embryo
transfer families. J Hered 92:290–292.
Schrooten C, Bovenhuis H, Coppieters W, and Van Arendonk JAM, 2000.
Whole genome scan to detect quantitative trait loci for conformation and
functional traits in dairy cattle. J Dairy Sci 83:795–806.
Spelman RJ, Huisman AE, Singireddy SR, Coppieters W, Arranz J, Georges
M, and Garrik DJ, 1999. Quantitative trait loci analysis on 17 nonproduction traits in the New Zealand dairy population. J Dairy Sci 82:2514–
2516.
Thomsen H, Reinsch N, Xu N, Bennewitz J, Looft C, Grupe S, Kuhn C,
Brockmann GA, Schwerin M, Leyhe-Horn B, Hiendleder S, Erhardt G,
Medjugorac I, Russ I, Forster M, Brenig B, Reinhardt F, Reents R, Blumel J,
Averdunk G, and Kalm E, 2001. A whole genome scan for differences in
recombination rates among three Bos taurus breeds. Mamm Genome
12:724–728.
Higashi Y, Asanuma M, Miyazaki I, and Ogawa N, 2000. Inhibition of
tyrosinase reduces cell viability in catecholaminergic neuronal cells. J
Neurochem 75:1771–1774.
Thomsen H, Reinsch N, Xu N, Looft C, Grupe S, Kühn C, Brockmann
GA, Schwerin M, Leyhe-Horn B, Hiendleder S, Erhardt G, Medjugorac I,
Russ I, Förster M, Brenig B, Reinhardt F, Reents R, Blümel J, Averdunk G,
and Kalm E, 2000. A male bovine linkage map for the ADR granddaughter
design. J Anim Breed Genet 117:289–306.
Kappes SM, Keele JW, Stone RT, McGraw RA, Sonstegard TS, Smith TP,
Lopez-Corrales NL, and Beattie CW, 1997. A second-generation linkage
map of the bovine genome. Genome Res 7:235–249.
Veerkamp RF, 1998. Selection for economic efficiency of dairy cattle using
information on live weight and feed intake: a review. J Dairy Sci 81:1109–
1119.
505
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
Bishop MD, Kappes SM, Keele JW, Stone RT, Sunden SL, Hawkins GA,
Toldo SS, Fries R, Grosz MD, Yoo J, and Beattie CW, 1994. A genetic
linkage map for cattle. Genetics 136:619–639.
Journal of Heredity 2003:94(6)
Visscher PM and Goddard ME, 1995. Genetic parameters for milk yield,
survival, workability, and type traits for Australian dairy cattle. J Dairy Sci
78:205–220.
CoA:diacylglycerol acyltransferase (DGAT1) with variation at a quantitative trait locus for milk fat content. Proc Natl Acad Sci USA 99:9300–
9305.
Weikard R, Goldammer T, Kühn C, Barendse W, and Schwerin M, 1997.
Targeted development of microsatellite markers from the defined region of
bovine chromosome 6q21-31. Mamm Genome 8:836–840.
Xu Y, Stokes AH, Roskoski R, and Vrana KE, 1998. Dopamine, in the
presence of tyrosinase, covalently modifies and inactivates tyrosine
hydroxylase. J Neurosci Res 54:691–697.
Weikard R, Kühn C, Goldammer T, Laurent P, Womack JE, and Schwerin
M, 2002. Targeted construction of a high-resolution, integrated, comprehensive, and comparative map for a region specific to bovine chromosome
6 based on radiation hybrid mapping. Genomics 79:768–776.
Zhang Q, Boichard D, Hoeschele I, Ernst C, Eggen A, Murkve B, PfisterGenskow M, Witte LA, Grignola FE, Uimari P, Thaller G, and Bishop MD,
1999. Mapping quantitative trait loci for milk production and health of dairy
cattle in a large outbred pedigree. Genetics 149:1959–1973.
Wellnitz O and Bruckmaier RM, 2001. Central and peripheral inhibition of
milk ejection. Livestock Prod Sci 70:135–140.
Winter A, Kramer W, Werner FA, Kollers S, Kata S, Durstewitz G,
Buitkamp J, Womack JE, Thaller G, and Fries R, 2002. Association of
a lysine-232/alanine polymorphism in a bovine gene encoding acyl-
Received October 10, 2002
Accepted July 31, 2003
Corresponding Editor: Bruce Weir
Downloaded from http://jhered.oxfordjournals.org/ by guest on May 24, 2013
506