Journal of the Scientific Agricultural Society of Finland
Vol. 45: 17-119 1973
Maataloustieteellinen Aikakauskirja

BOARS FOR BREEDING: A STUDY OF METHODS
OF EVALUATION USED AT TESTING STATIONS
Selostus: Karjujen koeasematestauksen tuloksellisuudesta

ELSI ETTALA

Department of Animal Breeding. University of Helsinki 1 )

Academic dissertation
To BE PRESENTED, WITH THE PERMISSION OF
the Faculty of Agriculture and Forestry
of the University of Helsinki, for public
CRITICISM IN THE SMALL FESTIVAL HALL ON
May 2nd, 1973, at 12 o’clock.

!) Present address: Agricultural Research Centre,
Department of Animal Husbandry, Tikkurila, Finland

SUOMEN MAATALOUSTIETEELLINEN SEURA HELSINKI

https://www.c-info.fi/en/info/?token=sBwbTABEH1vEvmeR.ZNqhHuCclqfG9tkoj_VqBg.0f9destuFHlo1kUgCHX-efvRRkB2jCU39Ma5lQI-3FlU-_KU0fP0y1qyg5UUaq7HwE-n4gxwXjdi7E7oPKW10Y1h3VPXeJGI441qh0isRZyLf2D9d5gPOoAUwWLDrjiWhMFplNgK0xUD3QGMDu6Lus8b-N58s_L4dQ




19

Preface

The first part of the study was carried out at the Pig Husbandry Testing
Station of North Finland, Haapajärvi, in 1967—1969, and the second part at
the Agricultural Research Centre, Tikkurila, in 1969—1972.

I wish to express my gratitude to Professor Mikko Varo, Principal of the
Department of Animal Breeding, University of Helsinki, who gave me this
research topic, and much valuable advice and inspiration during the course
of the work.

I wish to thank Professor Martti Lampila, Principal of the Department of
Animal Husbandry, Professor Kalle Maijala, Principal of the Department
of Animal Breeding, both of the Agricultural Research Centre, Tikkurila and
Mr. Johannes Partanen, M.Sc., Principal of the Pig Husbandry Testing
Station, Hyvinkää, for providing the facilities for the second part of the study.

I wish to thank Mr. Erkki Nenonen, M. A., and Mr. Veijo Vilva, B. A.,
for preparing the statistical programmes, and Mrs. Liisa Mattila, M. A., for
assistance in their use. I also wish to thank Mr. Unto Uusisalmi, B.Sc., for
many valuable suggestions concerning the analysis of the data.

I thank Mr. David Homer, B.Sc., and Dr. Ulf Lindström for linguistic
corrections to the manuscript in its translated form.

I thank the then-functioning Board of the Pig Husbandry Testing Station
of North Finland for support received during the first part of the study. I
wish to offer particularly warm thanks to one Board member, Dr. Kerttu
Saalasti, for continued interest and encouragement at all the stages of the
work.

I thank the executives and staff of the Artificial Insemination Society of
North Finland and Salpausselkä for all assistance, without which this study
would have been impossible. I wish, in particular, to thank Mrs. Irma Holopai-

nen, animal husbandry technician for her help in performing the tests on the
progeny. I also thank the staff of the Pig Husbandry Testing Stations of North
Finland and Hyvinkää for their conscientious work. I also wish to thank all
those persons who, in slaughterhouses, on farms etc., assisted me in the study.

I wish to thank the Department of Pig Husbandry of the Finnish Animal
Breeding Association and the Artificial Insemination Societies which provided
economic support in purchasing the test boars.

I wish to thank the August Johannes and Aino Tiura Agricultural Research
Foundation and the Fund for the Advanced Training of Agronomists for scholar-
ships.

Tikkurila, January, 1973
Elsi Ettala





21

CONTENTS
Page

I. Introduction 23
11. Review of literature 25

A. Ultrasonic measurement of the fat thickness and the meatiness of live pigs 25
1. Accuracy 25
2. Fat thickness and cross-sectional area of longissimus muscle as indicators of

meatiness and fat content of carcass 28
B. Rate of growth and feed efficiency as criteria in the phenotypetesting of boars

... 30
C. Effects of environmental factors on phenotypetesting 32
D. Genetic factors in phenotype evaluation 33

111. Author’s studies 38
A. Material and methods 38

1. Test boars and their rearing 38
2. Evaluation, selection and use in breeding of test boars 39
3. Progeny and their rearing 44
4. Progeny evaluation 45
5. Statistical methods 47

B. Results 49
1. Test boars 49

a. Phenotype evaluation: results and factors affecting them 49
b. Selection results 52
c. Interrelationships between characteristics 53

2. Progeny 56
a. Phenotype evaluation; results and factors affecting them 56
b. Carcass evaluation and lean cuts results of progeny 63

3. Results of phenotypeand carcass evaluation and lean cuts analysis of progeny,
classified in accordance with grade of sire 67

4. Heritability of characteristics, and phenotypic and genetic correlations
between characteristics 74
a. Sire-progeny correlations 74
b. Heritability estimates 77
c. Phenotypic and genetic relations between characteristics 78

5. Investigations on the improvement of station testing of boars 83
a. Phenotype and carcass characteristics as indicators of meatiness of carcass 83
b. Potential for early selection of boars 88
c. Leg strength of boars on abundant feeding 92
d. Additional information from barrow sibs in assessment of breeding value

of boars 96
IV. Discussion 100

A. Efficiency of phenotype evaluation of boars 100
1. Influence of animal material on results 100
2. Effects of environment 102
3. Heritability estimates of, and phenotypic correlations between, characteristics 103

B. Appraisal of means for the improvement of station testing of boars 104
1. Potential for early selection of boars 104
2. Leg strength of boars on various levels of feeding 105
3. Predicting carcass quality from results of phenotype testing 106
4. Additional information about variation in boars' carcass quality from barrow
sibs' lean cuts analysis 107

V. Summary 109
References 11l
Selostus 118





23

Ettala, Elsi 1973. Boars for breeding: a study of methods of evaluation
used at testing stations. J. Scient. Agric. Soc. Finl. 45: 00—000.

Abstract. The phenotype testing of highly selected boars was studied. Two
groups, each of 30 boars, were tested centrally at stations for growth rate and
ultrasonically measured fat thickness. According to test points, made up of a
combination of these two traits, the 5 best boars, 5 average boars and the 5 poorest
boars where selected for progeny evaluation. In all 26 boars and 441 progeny
were tested.

The progeny evaluation showed that ultrasonic measurement of the fat thickness
of the boars gave a very reliable estimate of the meatiness of their progeny. Those
boars, as a group, giving the poorest carcass quality could be distinguished with
statistical reliability from the other groups.

The correlations between sires and progeny were significant for both daily
gain (period 20 —BB kg) and feed efficiency. The importance of rate of growth
and feed efficiency has been neglected in selection for breeding as the test points
used for the selection of boars depended almost entirely on fat thickness.

More than 30 % of both boars and progeny boars suffered from some form
of difficulty in walking. 13 % of the boars were eliminated because of leg faults.
Leg faults in progeny were mainly inherited or caused through injury.

A phenotype evaluation of progeny boars accounts much more effectively
for the variation in their carcass value than does a full barrow sib evaluation
alone. For best results, a progeny boar phenotype evaluation should be combined
with a full barrow sib evaluation.

I. Introduction

Theoretical calculations show that for highly heritable traits (h > 0.25)
phenotype testing of pigs produces more rapid genetic progress than does pro-
geny or full sib evaluation (Hartmann and Fewson 1967, Jonsson 1971 b,
Kirsch et al. 1962, Lush 1947, Standal 1968). The brevity of the generation
cycle, the possibility of selection prior to breeding and the extensive range of
selection are benefits provided by phenotype testing. Consequently, phenotype
testing has rapidly gained ground, particularly after the introduction of ultra-
sonic measurement made it possible to evaluate the meat and fat contents of
live animals.



24

In on-the-farm testing, attention is paid mostly to determining the thick-
ness of the fat or, in some countries, to determining the cross-sectional area
of the longissimus dorsii muscle. Attention to growth and feed utilisation has
been relegated to second place, for such characteristics are difficult to determine
with reliability under varying farm conditions. Consequently boars are to
an increasing extent tested at stations, in order to ascertain these economically
very important characteristics and to increase the reliability of carcass
evaluation. The method of regular station testing of boars has thus been adopted
in many countries. Interest has also increased in a concurrent sib evaluation
in which the sibs of the boars are tested under experiment station conditions
(Blendl 1970, Kalm 1972, Moen 1972).

The number of boars tested at stations is generally low in comparison with
the number tested on farms. It is the objective of station testing to produce
superior boars to be used as parents of the next generation and in AI (artificial
insemination). In central testing of boars, originating from the best parents,
the differences between the boars are minimal. In order to measure these diffe-
rences it is necessary to ensure that the genetic characteristics of the boars are
fully expressed. A requirement for this is believed to be abundant feeding
which, however, is usually avoided in testing, because of the leg weaknesses
often encountered. Inaccuracies in the ultrasonic measuring of very thin layers
of fat, health disorders during growth, differing rearing conditions of the piglets
and other environmental factors may mask any significant differences among
the boars and render an expensive station test worthless. Moreover, progress
in breeding depends also upon the reliability of the criteria or indices used in
selection.

The present study was carried out in 1967—1972 to determine the value
of station tests of boars and to establish what factors influence the success of
the tests.

The major subjects of the study were:
1) The selection results of the testing stations, as evaluated through progeny

testing.
2) The heritability of the growth and carcass characteristics, and their pheno-

typic and genetic associations.
3) The influence of environmental factors in testing the individual animal.
4) Studies for the improvement of station testing:

a) Phenotype test and carcass measurements as indicators of the
meatiness and fat layer of the carcass.
b) possibilities of early selection of boars.
c) leg strength of boars on a high plane of nutrition, and
d) additional information on the breeding value of boars as provided
by castrated sibs.



25

11. Review of literature

A. Ultrasonic measurement of the fat thickness and the meatiness of live pigs

7. A ccuracy

As far as known, the principle of ultrasonic reflection in predicting the
carcass quality of pigs was first used by the Pig Marketing Board in England
in 1954. The technique of ultrasonic measurement obtained wider publicity
in 1956 at a demonstration to the Commission on Pig Production of the European
Association for Animal Production (Lauprecht 1960).

The mechanical measurement of fat (live probe) was developed slightly
earlier (Hazel and Kline 1952), and the use of the electrical probe as a gauge
for fat at roughly the same time (Andrews and Whaley 1955). These methods
of determination, however, have not gained such popularity in Europe as has
ultrasonic measurement.

Early studies showed that a good estimate of the thickness of the fat of
pigs could be obtained through ultrasonic measurement (Dumont 1957, ref.
Lauprecht 1960, Kliesch et al. 1957, Lauprecht et al. 1957, Price et al.
1958), the accuracy of which has subsequently been the subject of a great deal
of study. The precision has been checked by repeated ultrasonic measurements
at the same point, and the accuracy by comparing the results with the respective
measurements obtained from the carcasses (Table 1). The most commoh
positions for ultrasonic measurement have been the rear of the shoulder blade,
the midback and the loin, along the midline of the back. The places of measure-
ment of the side fat have varied a great deal (Table 1).

The repeatability of ultrasonic fat measurements obtained in various studies
varies between 0.41 and 1.00, and the correlation coefficients with the carcass
measurements between 0.39 and 0.95 (Table 1). The absolute differences
between the ultrasonic measurements of fat and the measurements of the
carcass have usually been under 2 mm (Clausen 1959, Lauprecht et al. 1957,
Peter 1962, Rittler et al. 1964, Rittler 1968 a, b, Skjervold 1962). The
accuracy of the ultrasonic measurement has depended primarily on the location
of the fat and muscle layers at the various points of measurement, the skill of
the operator, the movements and position of the animal at the momet of
measurement, and whether the points of measurement were marked and the
control measurements made in exactly the same spot.



To allow for the between-animal variation in the distribution of the fat
layers, it has been recommended that ultrasonic measurement be made at
3 5 positions (Gerlach 1967, 1970, Lauprecht et al. 1965, Rittler 1964,
1968 a, b, Schoen 1962, Scholz 1965). Measurements on both sides of the
animal, as opposed to measurements on one side only, do not markedly improve
the reliability of the results (Horst 1964, Rittler et al. 1964, Schoen 1962),
nor do a large number of measurements made at the same place (Gerlach
1967, Lauprecht et al. 1965, Rittler 1964, 1968 a, Rittler etal. 1964, Schoen
1962).

Table 1. Accuracy of ultrasonic measurement of fat thickness in live pigs: repeatability, and
correlation with carcass measurements.

Number Location .
..... ~

of pigs (MB = midline of back) epea ai i y r )Source

Clausen (1959) MB, 3 points
8 cm from MB

0.79-0.89
0.93

Gerlach (1967) 163 6 8 cm from MB,
I—s points, I—3 repeats

Hazel & Kline (1959) 56 5 cm from MB, 3 points
Hofmann & Peter 109 MB, 3 points

1—5 points, I—3 repeats 0.96 1.00
56 5 cm from MB, 3 points 0.66 0.70

109 MB, 3 points 0.91 0.93
109 90° to MB at midback, 2 points 0.88 0.90
109 Haunch, 3 points 0.77 0.91

(1964)

44 6 cm from MB on both sides 0.92
44 5 and 10 cm from MB, 5 repeats 0.74 0.90
30 5, 8 and 10 cm from MB

Horst (1964)
Kliesch & Horst
(1961)

1 repeat 0.92 0.97
Lauprecht et al. (1965) 132 MB, 1— 5 points, 1— 2 repeats 0.49—0.84

132 6 cm from MB, 1—5 points.
I—2 repeats 0.61—0.90

Mennerich (1967) 228 MB, 3 points 0.58 0.81
228 90° to MB at midback, 3 points 0.84 0.89

Otto & Sieg (1963) 109 MB, 3 points 0.92
Peter (1962)
Price et al. (1960)

Rittler (1964)

392 6 cm from MB, 5 points 0.98(1968 a)
- (1968 b) 82 6 cm from MB, 5 points 0.96

35 6 8 cm from MB, one or bothRittler et al. (1964)
sides, 1—5 points, 1—3 repeats 0.41—0.86

100 7 cm from MB, one or bothSchoen (1962)
sides 1— 6 points, I—6 repeats 0.71 0.95 0.85

563 MB, 3 points 0.39Scholz (1965)
SkArman (1960)

120 90° to MB at midback 0.64
Skjervold et al. (1960) 320 MB, 3 points
SUNDGREN (1969)

Uusisalmi (1969 a)

236 8 cm from MB 0.79
Vockert (1969)

Weniger et al. (1967) 101 MB, 3 points
101 90° to MB at midback

30 MB, 3 * 0.54-0.89
74 MB, 3 » 0.82
84 MB, 3 » 0.88

152 6 8 cm from MB,
1—5 points, 1— 3 repeats 0.72-0.93

0.39-0.52
120 MB, 3 * 0.55-0.71

0.47-0.68
190 MB. 3 points 0.49-0.95
190 8 cm from MB at midback 0.73 0.75
236 MB, 3 points 0.56-0.76
236 3, 6 and 9 cm from MB 0.64 0.75

95 MB, 4 points 0.76-0.90 0.49-0.58
95 6—B cm from MB, 3 points 0.87 0.94 0.51—0.59

0.55
0.58

*) r = coefficient of correlation between thickness of fat in live animal measured ultra-
sonically and thickness of fat measured on carcass.

26



27

Table 2. Accuracy of ultrasonic determination of thickness or area of cross-section of the lon-
gissimus muscle in live pigs: repeatability, and correlation with carcass measurements.

Number Location _

of pigs (MB = midline of back) Repeatability r>)Source

Assadi (1967) 163 Meat area, ultrasonograph2) 0.67 0.60
146 Fat area ultrasonograph2 ) 0.77 0.46

io2 ssar k
163 Thickness of muscle 6—B cm

from MB, 1-5 points, 1-3
repeats 0.92-0.99

44 Thickness of muscle 6 cm
from MB on both sides2 ) 0.78

98 Meat area, midback and rump,
ultrasonograph 0.59

98 Fat area, ultrasonograph 0.93
30 Meat area, ultrasonograph 0.61
30 Fat area, ultrasonograph 0.89

120 Meat area, one or both sides,
I—4 repeats, ultrasonograph 0.56 0.88

44 Thickness of muscle 5 10 cm
from MB, 5 repeats 2) 0.66-0.85

30 Thickness of muscle 5, 8 and
10 cm from MB, 1 repeat 0.91-0.96

85 Meat area 3 ) 0.77
85 Fat area3 ) 0.46
47 Meat area3 ) 0.63
47 Fat area 3 ) 0.64

132 Thickness of muscle 6 cm from
MB, 1-5 points, 1-2repeats 0.25-0.71

228 Thickness of muscle 90° to MB
at midback, 3 points 2) 0.21-0.61

108 Thickness of muscle at MB,

152 Thickness of muscle 6—B cm
from midback, 1—5 points,
1-3 repeats 0.40-0.86

82 Thickness of muscle 6 cm from
MB, 5 points 0.78

35 Thickness of muscle 6—B cm
from MB, 1 5 points,
1-3 repeats 0.48-0.88

35 Meat area, 1-3 repeats 0.91-0.97
35 Fat area, 1-3 repeats 0.93-0.98

100 Thickness of muscle, one or
both sides, I—6 repeats 8) 0.78 0.99 0.41

100 Meat area one or both sides,
1-6 repeats 3 ) 0.70-0.98 0.59

100 Fat area one or both sides
1-6 repeats 3 ) 0.86-0.99 0.57

42 Thickness of muscle, ultra-

-42 Meat area, ultrasonograph2) 0.70
236 Thickness of muscle, 3, 6 and

9 cm from MB, 3 points4) 0.03-0.28
95 Thickness of muscle 6 8 cm

from MB, 3 points 0.78-0.82 0.35-0.49

Diekmann (1960)

Gerlach (1967)

Horst (1964)

- (1969)

- (1971)

Kliesch & Horst
(1961)

Lauprecht et al. (1960)

- (1965)

Mennerich (1967)

Otto & Sieg (1963)

Rittler (1964)

- (1968 b)

Rittler et al. (1964)

Schoen (1962)

Stouffer et al. (1961)

Uusisalmi (1971 b)

Vockert (1969)

1) r = coefficient of correlation between thickness or area of cross-section of longissimus
muscle measured ultrasonically and corresponding measurements made on carcass.

2) Measured between 13th and 14th rib.
3) Measured between last and last-but-one rib.
4) Measured behind last rib.



28

In Finland the ultrasonic measurement of fat thickness has usually been
carried out in accordance with Swedish practice (Sundgren 1965, 1967, 1969).
In 1965—6B the measurements were taken along the midline of the back at
the withers, the midback and the loin. Since 1968 measurements have been
made at the midback and at a distance of about 8 cm on either side of the
midback (see Fig. 1); the change was made after tests performed in Sweden
and Finland had shown that the new method gives improved results (Sundgren
1964, 1967); Uusisalmi’s findings were published later (1969 a, 1971 b).

The thickness and cross-sectional area of the longissimus muscle are more
difficult to determine by ultrasonic methods than the thickness of the fat layers.
Series of ultrasonic mesurements made close to the longissimus muscle are used
to construct a cross-sectional diagram of this muscle and its overlaying fat.
Reported repeatabilities of the measurements of muscle area, and the correlation
with carcass measurements, have varied considerably (Table 2). The use of
several locations of measurement and repeated measurement at the same position
have improved the reliability in the determination of muscle thickness, but not
in the determination of muscle area (Lauprecht et al. 1965, Mennerich 1967,
Rittler 1964, Rittler et al. 1964). If the appropriate equipment is available,
an ultrasonic picture of the site of measurement can be used to improve the
reliability of the ultrasonic measurements (Assadi 1967, Horst 1969, 1971,
Kliesch and Horst 1962, Stouffer et al. 1961). The relative contributions
of muscle and fat to the cross-sectional diagram, obtained by the series of
ultrasonic measurements mentioned above, permit calculation of the expected
fat: meat ratio of the carcass. The ratio is generally closely correlated with
the ratio found by direct measurement on the carcass (r = 0.41—0.94)
(Assadi 1967, Diekmann 1960, Horst 1969, Lauprecht et al. 1960, Schoen
1962).

2. Fat thickness and cross-sectional area oflongissimus muscle as indicators
of mealiness and fat content of carcass

The whole purpose in measuring the fat layers and longissimus muscle of the
live pig is to quantitate the meatiness and fat content of the main part of the
carcass obtained at slaughter. The value of these two measurements as indica-
tors of carcass characteristics has been tested by correlating them with the
weights of lean cuts obtained from the carcass; measurements made on the
longitudinally sectioned carcass have been evaluated in the same way (Tables
3 and 4). As regards fat thickness, ultrasonics data and carcass section data are
both well correlated with the lean cuts results (Table 3). With regard to lon-
gissimus area, however, the ultrasonics-versus-lean cuts correlation is not as
good as the carcass section-versus-lean cuts correlation (Table 4). According
to Horst (1971), recent developments in ultrasonics have made a measure of
the muscle obtained by this technique as reliable as that obtained directly
from the carcass, in the evaluation of the meatiness of the carcass.

Measurements of side fat have proved to be the best indicators in predicting
carcass quality in a number of studies (Fender 1962, Hoffmann and Peter



Table 3. Correlation of fat thickness, measured ultrasonically on the live animal or determined directly
on the longitudinally-sectioned carcass, with proportion of lean cuts and fat cuts obtained from carcass.

Number Correlation coefficient, r
Source of Ultrasonic measurement '

'. ! !

pigs % lean cuts % fat cuts onß t
r ° '" '" muscle area

Gerwig (1965 a) 205 Fat, 2-3 points 0.42 to-0.562)
Hazel & Kline 56 Sidefat, 5 cm from midline

Hofmann & Peter 53 Backfat, 3 points —0.48 0.62

I.u'prkcht etal. 132 Backfat, 3 points —0.27 0.47

Mennerich (1967) 101 Backfat, 3 points -0.23 0.51

back, 3 points -0.37 0.68
Peter (1962) 30 Backfat, 3 points -0.60
Price et al. (I960) 74 Sidefat 3.8 cm from mid-

Sundgren (1969) 190 Backfat, 3 points -0.15 to -0.43 0.15-0.55
190 Sidefat 8 cm from midline

of back -0.55 to-0.57 0.61-0.66
Uusisalmi (1971 b) 236 Sidefat 3-9 cm from mid-

line of back, 4 points —0.33 to —0.43

Carcass section measurement
0.72-0,74Cross et ai. (1970) 43 Fat, 3-5 points -0.59 to-0.62 0.72-0.74

Gf.rwig (1965 a) 205 Fat, 2-3 points -0.38 to 0.432 )
Hazel & Kline 96 Fat, 4 points —0.45 —0.41

- (1959) 56 Fat, 3 points -0.85
Holland & Hazel 105 Fat, 8 points —0.72 0.81 —0.29

Kline & Goll 50 Fat, 8 points -0.35 to-0.47

Lauprecht et al. 132 Backfat, 3 points —0.31 0.50

Mennerich (1967) 101 Backfat, 3 points —0.35 0.54
101 Sidefat, to side from

-0.41

midback, 3 points —0.54
MOller-Haye 78 Backfat, 3 points —0.73
(1965) 78 Sidefat -0.77
Pearson et al. 195 Fat —0.68
(1958 a)
- (1958 b) 195 » -0.47
- (1959) 292 » -0.80- (1959) 292 » -0.80
Pedersen (1968) 320 +356 Backfat, 3 points —0.33 t 0—0.63

320 +356 Sidefat -0.64 to —0.71
Peter (1962) 30 Backfat, 3 points
Price et al. (1960) 74 Fat, 3 points —0.74

84 » 3 * -0.80
Rittler et al. 228 Backfat —0.38
(1965) 228 Sidefat -0.45228 Sidefat -0.45
Sundgren (1969) 190 Backfat, 3 points —0.20 t 0—0.48

190 Sidefat 8 cm to side —0.43 to —0.49
Uusisalmi (1971 a) 153 Sidefat —0.57
Zobrisky et al. 207 Backfat, 3 points
(1959)

0.72
0.77
0.79

-0.50

-0.78

0.43
0.49
0.23-0.50
0.64-0.66
0.73
0.67

*) Reported methods of obtaining lean cuts vary slightly.
2 ) Results expressed in weight units, not % units.

29



30

Table 4. Correlation between the cross-sectional area of the longissimus muscle, measured
ultrasonically on the live animal or directly on the longitudinally sectioned carcass, and %

lean cuts and % fat cuts.

Number Ultrasonic o'o %

of pigs measurement of muscle area lean cuts fat cuts

Diekmann (1960) 102 between ribs 13 and 14 0.41
Horst (1971) 111 ultrasonograph 3 points 0.17 0.42
Lauprecht et al. 132 3—5 points 0.36 0.42 —0.29 t 0—0.42
(1965)
Mennerich (1967) 101 between ribs 13 and 14 0.16 -0.34
Schoen (1962) 97 • I I » • 0.32 -0.28

Carcass section
measurement of muscle area

Cross et al. (1970) 43 0.48 -0.37
Horst (1971) 111 2 points 0.50-0.52
Kline & Hazel 23 between last and last- 0.65
(1965) but-one rib

23 behind rib 10 0.66
Lauprecht et al. 132 between ribs 13 and 14 0.60 —0.50
(1965)
Mennerich (1967) 101 • • • • » 0.69 —0.51
Muller-Haye 78 » » » » » 0.59 —0.53
(1965)
Pearson et al. 102 between last and last- 0.53
(1956 b) but-one rib

102 behind rib 10 0.52
Pedersen (1968) 320+ 356 between last and last- 0.30 0.49

but-one rib
Price et al. (1960) 84 behind rib 10 0.62
Rittler et al. 228 0.43 -0.40
(1965)
Uusisalmi (1971 c) 97 behind last rib 0.61

1964, Mennerich 1967, Rittler et al. 1965, Sundgren 1969), despite the
rather variable results reported earlier (Buck et al. 1962, Hazel and Kline
1952, Robison et al. 1960). When carcass meatiness was determined by ultra-
sonic measurement of both fat and longissimus muscle, the muscle data added
little to the value of the fat data (Lauprecht et al. 1965, Schoen 1962, Uusi-
salmi 1971 b).

B. Rate of growth and feed efficiency as criteria in the phenotype
testing of boars

The rate of growth of boars is easily determined by weighing, and the feed
efficiency by controlled feeding of the individual animals. However, little
account has been taken of such characteristics in phenotype evaluation. The
varying conditions of on-farm testing make it difficult to produce comparable
results (Flock et al. 1970, Mennerich 1967). In station testing consideration
of rate of growth and feed efficiency have depended on the selection indices



31

used and the absolute results on the feeding standards practised (Blendl 1970,
Ettala 1971 a. b, Fewson et al. 1962, Minkema et al. 1964, Persson and
Lindh£ 1972).

Standal (1962) and Sundgren (1965, 1967) have quantitated the relation-
ships between weight and age, making it possible to compare animals of diffe-
rent age. Attempts have been made to reduce differences due to environmental
conditions by combining the growth and fat data. Growth nomograms pre-
pared by Sundgren have been used in Finland (Figures 3 and 5), and the se-
lection of boars at farms and testing stations has been made in accordance
with test points calculated from growth and fat data (Anon. 1968). The rate
of growth, however, has accounted for only a very small part of the variation
in test points, i.e. 9.2 % for boars tested at stations in Finland in 1965 68,
while the thickness of the fat has accounted for 85.2 % of the variation (Ettala
1971 b). Moen (1968) reported that rate of growth and feed efficiency together
account for more than 20 % of the variation in a Norwegian index. In Sweden
the system of test points has been abandoned in favour of a station index, in
which the importance of feed efficiency is emphasised (Persson and Lindhe
1972). In Finland feed efficiency has not been as a separate directly taken into
account; the association with growth rate was analysed by Ettala (1971 b).

The interrelationships between characteristics naturally have a bearing
upon their usefulness in selection and breeding. There are very close phenotypic
and genetic correlations (r p ranges from 0.49 to —0.92 and r g from —0.57
to —1.02) between rate of growth and feed efficiency (Biedermann 1971,
Flock 1970, Jonsson 1963, Krippl et al. 1965, Langholz 1966, Persson
and Lindhe 1972, Schmitten 1967, Varo 1962, Zagozen and Schröder
1970). Evidently these figures are due in part to auto-correlation, for rapid
attainment of final weight requires improvement of both rate of growth and
feed efficiency (Biedermann 1971, Jonsson 1963, Krippl et al. 1965, Schmit-
ten 1967).

According to some studies, are of growth and feed efficiency are positively
genetically correlated with low fat-thickness and meatiness, i.e. meaty pigs
with a thin layer of fat grow rapidly and have a good feed efficiency. In these
studies the correlation coefficient between rate of growth and fat thickness
ranged from —0.31 to —0.68 ,and that between rate of growth and area
of longissimus muscle from + 0.28 to + 0.39 (Biedermann 1971, Flock
1970, Gerlach 1967). The correlation coefficient (r g ) between feed efficiency
and (a) fat thickness is -j-0.20 to +0.79, and (b) longissimus muscle area is
—0.27 to —0.76 (Biedermann 1971, Ettala 1971 b, Flock 1970, Jonsson
1963, Krippl et al. 1965, Schmitten 1967). However, rate of growth and feed
efficiency do not always correlate well with carcass characteristics (Buchen-
auer 1970, Englisch 1969, Ettala 1971 b, Fewson et al. 1962, Langholz

1966, Mennerich 1967, Persson and Lindhe 1972, Scholz 1965, Wussow
and Grosse 1959). The latter reports indicate that the assumption that growth
and feed-utilisation characteristics of pigs would improve automatically when
selection for breeding is made on the basis of low fat-thickness and meatiness is
erroneous. Consequently it is important that rate of growth and feed efficiency
be taken into account, at least when selecting boars for artificial insemination.



32

C. Effects of environmental factors on phenotype testing

Phenotype evaluation is sensitive to environmental factors, which mainly
affect rate of growth and feed efficiency, and only slightly meatiness (Horst
1969, Minkema et al. 1964, Persson and Lindh£ 1972).

F e e d is one of the most important of the environmental factors (Blendl
1970, Brunstad and Fowler 1959, Cole and Hardy 1971, Cook et al. 1972,
Glodek et al. 1971, Lodge et al. 1972, Schierbaum 1961, Sundgren 1972).
Feeding levels have varied from restricted to ad libitum (Blendl 1970, Ettala
1971 a, Fewson et al. 1962, Persson and Lindhe 1972, Minkema et al. 1964,
Zagozen and Schröder 1970). With pigs, individual feeding is preferred to
group feeding, in order that both feed efficiency and rate of growth can be
measured (Fredeen and Jonsson 1957, Hofmann 1965, Jonsson 1959).

The extent to which abundant feeding increases leg weakness is not yet
known. A correlation between rapid growth and leg weakness was observed
(Schmid 1970), but leg weakness was common also among slow-growing boars
on restricted feed. It has generally been necessary to eliminate 20—40 %

of boars under test because of this tendency (Ettala 1971 a, Nebe 1969,
Persson 1972, Wright 1967). Diet, genetic factors, lack of exercise and in-
fection of leg injuries may play a part in causing leg weakness (Gerwig 1965 b,
Kangasniemi 1971 a, Melrose 1967, Smith 1966).

Live-weight and fat thickness. Fat thickness as determined
by ultrasonic measurement is closely related to live weight, so that live-weight
corrections to the thickness data are required. The correction factors, which
depend on breed, sex, position of measurement and desired testing weight,
vary from 0.07 to 0.24 mm/kg (Englisch 1969, Gerlach 1967, Glodek 1964,
Hofmann and Peter 1964, Hofmann et al. 1965, Langlet et al. 1968, Laup-

recht et al. 1967, Lindhe and Sundgren 1969, Mennerich 1967, Otto and
Sieg 1965, 1966, Rittler 1964, 1968 a, Schumm et al. 1966, Uusisalmi 1971
d). The relationship between fat thickness and live weight is linear or almost
linear (Gerlach 1967, Hofmann and Peter 1964, Lauprecht et al. 1967,
Mennerich 1967, Quijandria and Robison 1971, Rittler 1964, 1968 a).
If the fat is measured at a live weight close to a predetermined value, the effect
of live weight upon the results is not significant (Uusisalmi 1971 b). The
thickness of the longissimus muscle and the fat-to-meat ratio calculated from
the ultrasonic measurements are not closely related to live weight (Assadi
1967, Gerlach 1967, Horst 1969, Lauprecht et al. 1967, Mennerich 1967,
Rittler 1964, 1968 a).

Age and fat thickness. According to the studies cited above,
consideration of the age of the animal, in addition to its weight, at the time of
ultrasonic measurement does not improve the accuracy of the results; these
findings contrast with those of Ettala (1971 a), Plonka et al. (1969), Qui-
jandria and Robison (1971), Robison (1962) and Skjervold et al. (1960).
Using points scales for fat thickness evaluation, Standal (1962) corrected
for both age and weight; Sundgren (1965) made a correction for age only, but
later (Sundgren 1967) corrected for weight only (Figures 4 and 6 are examples
of Sundgren’s points scales).



The effects of the initial weight and age on the test results have
been determined in many progeny trials. Pigs that have reached the required
starting weight at a normal age usually grow slightly faster during the testing
period than pigs that have reached this weight at an earlier age (Jonsson 1963,
Langholz 1965, Persson and Lindhe 1972, Scholz 1965). The effects of star-
ting age on feed efficiency and carcass characteristics are slight; both positive
and negative effects have been reported (Buchenauer 1970, Jonsson 1963,
Langholz 1965, Persson and Lindhe 1972).

Seasonal effects on the rate of growth, feed efficiency and carcass
quality have been studied in progeny and phenotype tests, and conflicting
results have been obtained. Biedermann (1971), Buchenauer (1970), Fredeen
and Jonsson (1957), Johansson and Korkmann (1950), Krippl et al. (1965),
Persson and Lindhe (1972), Rittler (1968 b) and Schmitten (1967) found
that season had a significant effect on growth, feed efficiency or carcass
characteristics, whereas Horst (1969), Langholz (1965), Mennerich (1967),
Schmidt (1964) and Vockert (1969) found that season did not cause any syste-
matic differences. This divergence in results is probably due to differences in
rearing conditions. Particular attention should consequently be paid to the
temperature and the ventilation of the animal sheds.

D. Genetic factors in phenotype evaluation

The accuracy of phenotype evaluation of boars has been studied by cal-
culating heritabilities. Heritability (h2) estimates have been calculated from half
sib correlations for boars tested in the field or for their progeny reared at testing
stations, and from regressions of boars tested in the field on their progeny
tested either in the field or at a testing station (Flock et al. 1970, Glodek
1964, Horst 1969, Langlet et al. 1968, Lauprecht et al. 1967, Mennerich
1967, Rittler 1964, 1968 a, b, Vockert 1969). In USA efficiency of pheno-
type selection was checked by selecting pigs for high and low fat thickness
for 5 to 10 generations, and was comparing these with an unselected control
strain (Berruecos et al. 1970, Gray et al. 1968, Hetzer and Harvey 1967).
Fat thickness was determined with the live probe or by the »leanmeter».

The h2-estimates of the above investigations are given in Table 5. The esti-
mates are not completely comparable because the number of animals and the
environmental conditions considerably influence the results. Similarly, the
number of farms at which the progeny have been reared, and the relationships
between the animals in particular have affected the h2-estimates (Lauprecht
et al. 1967, Mennerich 1967, Rittler 1964, 1968b, Vockert 1969). Systematic
differences between farms are confounded with the differences between boars,
and increase the h2-estimates considerably. There is a pronounced tendency
towards over-estimation when the progeny are on one or two farms only
(Mennerich 1967, Vockert 1969). Despite differences in conditions and po-
pulations, the h2-estimates for thickness of fat reported by most investigators
are comparable. The average fat thickness of several measurements, has a
greater heritability than that of individual measurements (Mennerich 1967,

3 33



34

Table 5. Heritability estimates of carcass characteristics determined by performance tests.

No. of pigs Heritability estimates (h 2)

Source Method of evaluation -.,
.

,
Thickness of „,,

c . -r, Imckness , . . Fat/meatSires Progeny . -
. loneissimus° J of fat 6

, ratiomuscle

Berruecos et al. 55 483 Cumulative selection
(1970) difference over several

generations 0.27
Glodek (1964) 21 1096 Half-sib correlations

of on-farm tested boars 0.59 0.35 0.48
Gray et al. (1968) 67 1828 Cumulative selection

difference over
several generations 0.23 0.39

67 1828 Dam-progeny regres-
sion 0.39-0.51

Hetzer & Harvey 346 3556 Cumulative selection
(1967) difference over

several generations 0.38 0.48
Parent-progeny re-
gression 0.43 0.70

Horst (1969) 193 Correlation between

an! teste?
at station 0.11-0.32 0.32-0.561) 0.44-0.75

Langlet et al. 26 283 Selection difference
(1968) between progeny of

high and moderate
grade sires 0.49 0.26

Lauprecht et al. 578 3229 Half-sib correlation
(1967) of farm-tested boars 0.37-0.44 0.01-0.09 0.32-0.45

274 3240 Regression between
farm-tested sires
and progeny 0.34-0.42 0.02-0.12 0.29-0.37

Mennerich (1967) 268 5064 Half-sib correlation
of farm-tested boars 0.46-0.74 0.24-0.35 0.53-0.65

30 Regression between
farm-tested sires
and progeny 0.26-0.87 0.40-0.56 0.17-0.20

144 2091 Half-sib correlation
of farm-tested boars 0.34-0.46 0.31 0.38

63 965 • 0.28-0.55 0.30 0.65
Rittler (1964) 815 I 0.27-0.31 0.06-0.24 0.25-0.26
(1968 a) 141 1457 I 0.25-0.31 0.08-0.13 0.16-0.22

77 340 Regression between
farm-tested sires
and progeny 0.33-0.35 0.30-0.40 0.37 0.43

(1968 b) 1332 Half-sib correlation
of station-tested pigs 0.34-0.38 0.26-0.34 0.33 0.38

Vockert (1969) 55 440 » (70 kg) 0.34-0.71 0.27-0.58 0.29-0.58
I (90 kg) 0.20-0.82 0.15-0.44 0.54-0.70

x) Figures for muscle area.



35

Rittler 1968 b, Vockert 1969). Hetzer and Harwey (1967), in their selection
experiment stretching over B—lo generations, got higher h2-estimates for the
fat thickness of the first generation and the low fat strains than for the last
generation and the high fat strains.

In a number of investigations the h2-estimates for the ultrasonically measured
cross-sectional area of the longissimus muscle were lower than those for the
ultrasonically measured fat thickness (Lauprecht et al. 1967, Rittler 1964,
1968 a) (Table 5). Using an improved ultrasonographic method, however,
Horst (1969) obtained higher h2-estimates for the muscle area than for the
thickness of the fat. The h2-estimates for the fat:meat ratio have usually been
similar to those for the thickness of the fat (Table 5).

By comparing heritabilities of characteristics measured ultrasonically on live
animals with the corresponding ones of the carcasses the reliability of the ultra-
sonic measurements can be determined. Rittler (1968 b) obtained similar
heritabilities for ultrasonic and carcass measurements of pigs reared at a testing
station. Generally speaking, the h2-estimates for carcass-measured fat thickness
in pigs reared at experimental stations has varied from 0.15 to 0.73 (more
frequently 0.30—0.55), for longissimus muscle area from 0.09 to 0.56, and for
the fat:meat ratio from 0.28 to 0.72 (Biedermann 1971, Enfield and What-
ley 1961, Flock 1970, Fredeen and Jonsson 1957, Jensen et al. 1967, Jo-
hansson and Korkmann 1950, Jonsson 1971 a, Kangasniemi 1971 b, Lang-
holz 1966, Maijala and Vainikainen 1962, Schmitten 1967, Siers and
Thomsom 1972, Varo and Partanen 1965, Weiss 1967). The h2-estimates for
growth rate in these studies were 0.14 0.67 and those for feed efficiency 0.12
0.72. The h2-estimates for growth rate in phenotype evaluation studies have
been 0.03—0.46 and those for feed efficiency 0.14—0.39 (Buchenauer 1970,
Gerlach 1967, Mennerich 1967, Scholz 1965). Norwegian h2-estimates for
combined fat+growth scores were 0.55—0.63 (Skjervold 1962).

In boar (testing station) progeny studies it has not been possible, on
account of the limited amount of data, to calculate heritabilities, but Fewson
et al. (1962) observed that the difference between progeny of thin-fat and
thick-fat boars was according to expectation, and Minkema et al. (1964), in
the station-testing of boars, got almost twice as reliable results as those obtained
in full sib testing. In a preliminary report Persson and Lindhlj (1972) give
the following correlations between results of station tested boars and their
progeny (at 90 kg): fat thickness measured at several points and in material
from several years r = 0.13 0.76, growth rate r = 0.24—0.52 and feed ef-
ficiency r = 0.40. The correlations between the results of boars tested in the
field and their progeny tested at stations were: fat thickness r = 0.11—0.61,
longissimus muscle area 0.27 to —0.48 and fat:meat ratio 0.23—0.72 (Flock
et al. 1970, Glodek 1964, Mennerich 1967). No statistically significant cor-
relations between rate of growth for sires and progeny have been found (Flock
1970, Mennerich 1967).

Genetic correlations between fat thickness and longissimus muscle thickness
(both measured ultrasonically), and between these and the fat:meat ratio,
are given in Table 6. The table also shows the corresponding correlation coef-
ficients for carcass characteristics of progeny-tested animals. The correlations



36

for ultrasonic and carcass measurements are of about the same order. Fat
thickness measured ultrasonically is more closely correlated with the fat:meat
ratio than the ultrasonic measurements of the longissimus muscle. However,

Table 6. Genetic coefficients of correlation (r g ) between thickness of fat and longissimus muscle,
determined by ultrasonics, and measurement of carcass section and carcass characteristics
determined by lean cut analysis.

Carcass characteristics

Source Ultrasonic measurement of live pig Longissimus Fat/meat
muscle thickness ratio

T _g rg

Gerlach (1967) Thickness of fat, 5 points 1.04
Lauprecht et al. » » » 3 » —0.45 0.99
(1967)
Mennerich (1967) Thickness of fat towards side from

midback, 3 points —0.48 0.93
Rittler (1964) Thickness of fat along side, 5 points —0.40 0.96
- (1968 a) » » » » » 5 » -0.04 0.95
Weniger et al. » » » » » 3 » -0.07 0.96
(1967)
Lauprecht et al. Thickness of longissimus muscle,
(1967) 3 points -0.60
Mennerich (1967) Thickness of longissimus muscle

towards side from midback, 3 points —0.75
Rittler (1964) Thickness of longissimus muscle

along side, 5 points —0.71
(1968 a) Thickness of longissimus muscle

along side, 5 points —0.27
Weniger et al. Thickness of longissimus muscle
(1967) along side, 3 points —0.31

Carcass section measurement

Krippl et al. (1965) Thickness of fat —0.41
Biedermann (1971) • I • -0.44 to -0.50 0.64-0.80
Flock (1970) I • » -0.34 0.73
Jensen et al. (1967) I i » -0.06
Biedermann (1971) Area of longissimus muscle —0.78
Flock (1970) » » » » -0.82

the cross-sectional area of the longissimus muscle as measured on the carcass
is correlated with the fat:meat ratio as closely as is fat thickness. Jensen
et al. (1967), Rittler (1968 a) and Weniger et al. (1967) found no genetic
correlation between fat thickness and the measurements of the longissimus
muscle. The genetic correlations between growth and feed-utilisation characte-
ristics and between these and carcass characteristics have already been given
(page 31).

From a breeding point of view the genetic correlation of the ultrasonically



37

measured fat and longissimus muscle with the lean cuts results gives the best
indication of the value of ultrasonic measurements. There are few such studies.
The genetic correlation coefficients rg obtained by Rittler (1968 b) are as
follows:

measured ultrasonically on live animal —0.83 to —0.86 0.74 to 0.78
measured on carcass —0.65 to —0.74 0.63 to 0.81

Thickness of longissimus muscle
measured ultrasonically on live animal 0.55 to 0.57 —0.26 to —0.32
measured on carcass 0.61 —0.45

Fat: meat ratio
measured ultrasonically on live animal —O.Bl to —0.90 0.62 to 0.78
measured on carcass —0.84 0.69

Rittler’s figures are comparable to those of Jensen et al. (1967): rg = —O.Bl
for fat thickness (carcass) vs lean cuts, and r g

= 0.49 for longissimus muscle
area (carcass) vs lean cuts. Thus, according to the results of Rittler, ultrasonic
measurement of the live pig is as reliable as carcass measurement in breeding
for meatiness.



38

111. Author’s studies

A. Material and methods

1. Test hoars and their rearing

Housing for 30 boars was put up in the vicinity of the Pig Husbandry
Experiment Station of North Finland (Haapajärvi 01). Group A was reared
in 1967 and Group B in 1968 69, making a total of 60 boars.

In order to speed up progeny evaluation, one half of the boars selected were
of Yorkshire and one half of Landrace breed.

23 boar piglets were selected from each of 24 litters in various parts of
Finland. Their sires (19 in all) and dams were of the highest quality available.
From the very best sires piglets were taken from several litters. In making
the selection the aim was to follow a procedure corresponding to a situation
where animals were selected for the only boar testing station in the country.

The 30 boar piglets of Group A were brought to the testing station over the
shortest possible period, so that they could be reared and evaluated at the same
time. The piglets were born between June 2nd and July 24th 1967. After
rearing, the boars were retained for varying periods while semen was being
collected from them. When a place became vacant it was filled with a Group
B piglet. The piglets of Group B were born between January 27th and October
25th 1968.

Individual feeding of the boars, 2 or 3 of which were housed in each pen,
was arranged by lowering partitions at feeding time. Each boar was fed twice
daily with a restricted amount of feed depending on its weight (Table 7).
Restricted feeding was employed in order to avoid leg weakness. The level of
feeding for Group A was generally the same as that used for test boars in Finland
(Ettala 1971 a). In Group B, the amount of feed proved to be too restricted
during the first weeks and was increased by 0.1 feed units per day until the ani-
mals weighed 60 kg, after which the amounts followed the 1967 standard up to
a liveweight of 80 kg, and were then 0.1 feed units less than the 1967 standard.
The composition of the feed was uniform throughout, being the same as that
used for all progeny, up to a weight of 50 kg, tested for breeding purposes
in Finland (Table 8). The animals were given water ad libitum, and the boars
were weighed once a week.



39

Table 7. Feeding standards for the boars.
FU = Scandinavian feed unit.

Live weight of boar FU/boar/day

k 8 1967 (Group A) 1968-69 (Group B)

20.0-24.9 0.8 0.9

Table 8. Composition of feed of test boars and 1968 progeny.

Group A Group B
Composition of feed Aug. —Oct. 1967 1968—69

% %

Barley meal 52.0 52.0
Maize meal 26.0 28.5
Dried milk 10.0 0.0
Fish meal 4.5 12.0
Soyabean meal 3.0 3.0
Yeast 2.0 2.0
Mineral mixture1 ) 2.0 2.0
Vitamin preparation2) 0.5 0.5
Energy values FU/kg3) .... 1.01 0.97-1.00
True protein %3 ) 15.2 16.3-19.3

b Composition of mineral mixture, %: fodder phosphate 40.0, ground limestone 40.0,
common salt 18.5, zinc sulphate 0.75, iron sulphate 0.60, copper sulphate 0.065, manganese
sulphate 0.10, cobalt sulphate 0.01 and potassium iodide 0.005.

2) Composition of vitamin preparation: vitamin A 600 000 i.u., vitamin D6O 000 i.u. and
vitamin E 2000 mg per kg feed.

3) Analysis done at the State Institute of Agricultural Chemistry, on each batch of feed.
FU

t
= Scandinavian feed unit.

2. Evaluation, selection and use in breeding of test hoars

In the phenotype evaluation of the boars, rate of growth, feed efficiency,
thickness of fat, length of side and leg condition, as well as the ham measurement
of the boars in Group B, were determined.

Growth and feed efficiency were calculated for the weight range 20—88 kg.
If the initial and final weights diverged from these, correction coefficients, uni-
form with those employed in progeny testing, were used. The coefficients per
kg liveweight were 2.58 days and 2.30 feed units at the beginning of the test.



40

and 1.25 days and 4.19 feed units at the end of the test. Another growth
characteristic used was the age in days at a weight of 88 kg. When the boars
were selected for progeny testing the characteristic used was the number of
growth points determined by using weight-age nomograms (Figs. 3 and 5)
prepared by Sundgren (1965, 1967).

Ultrasonic measurement of the fat thickness of the boars was done by staff
of the Finnish Pig Breeding Society. The measurement was made as near as
possible to a live weight of 88 kg, the long jorneys involved preventing complete
uniformity in this respect. The equipment used was the Krautkrämer USM
1 or USK 5 SF. Measuring was done at points along the midline of the back
at the withers (I), midback (2) and loin (3) (Fig. 1). The thickness of the side
fat, that is the so-called sol fat, was measured at a point 8 cm from the midline
of the back on both sides, at the rear edge of the last rib (4 and 5). The side
length was measured from the rear edge of the front leg to the base of the tail
(Fig. 1). The ham measurement (Group B only) was made as shown in Fig 2
(Uusisalmi 1971 c). The legs were evaluated as follows:

Animal incapable of walking 1 point
Animal incapable of walking unaided 2 points
Visible defects in legs impeding walking 3 »

Visible defects in legs not impeding walking 4 »

No visible defects 5 »

The uncorrected fat thickness average (1,2,3) for the withers, midback and
loin, representing average back fat, was converted to fat points by means
of the scale shown in Fig. 4. The uncorrected average (2,4,5) for the mid-
back and the two sol measurements was converted to fat points by means
of the scale shown in Fig. 6. The points scales are the same as those prepared
by Sundgren (1965, 1967), except that instead of the above-mentioned averages
Sundgren used sol fat measurements. Test points were obtained by summing
growth and fat points. The old test point scales (Figs. 3 and 4) were official in
Finland until October 15th 1968, when new point scales were introduced.
Consequently, the selection of the boars of Group A for progeny evaluation was
based on the old points scales, and that for Group B on the new scales. Test
points (old system) and test points (new system) were calculated for all boars.

The intention was to select for progeny evaluation the 5 best, 5 average and
the 5 poorest boars, in terms of test points, from each of the two groups of 30
boars. Surplus candidates for each sub-group were eliminated when necessary
on the basis of their breed, since it was desired to strive for an overall balance
between the two breeds. The selection could not be made entirely as intended
because some boars had to be rejected on account of leg defects. Later, other
boars had to be eliminated because of low semen quality or difficulties in ob-
taining the ejaculate, and also because of subsequent leg defects and injuries.
In Group A there were insufficient reserve boars, so that for progeny eva-
luation the sub-groups were limited to 3 good, 4 average and 4 poor boars.
A large number of Group B boars were held in reserve, so that the sub-groups
of Group B could be brought up to number 5+5 + 5 as intended. The quality



41

Fig. 1. Locations of ultrasonic measurements of fat on test boars, and manner in which length
of side was measured. (Boar »Rukki», see Table 14).
1 = withers, behind shoulder blades
2 = midback, thinnest place
3 = loin, thinnest place
4 = sol fat, right side, at rear edge of last rib, about 8 cm from midline of back (on pit; weighing

about 88 kg).
Side length is the distance between the vertical lines.

Fig. 2. Method of measuring the ham.
(Photo U. Uusisalmi)



42

Fig. 3. Assessment of Group A (1967) boars: weight-age nomogram giving growth points (old
system).

Fig. 4. Assessment of Group A (1967) boars; fat thickness-age nomogram giving fat points
(old system). Back fat thickness is the mean of the fat thickness at withers, midback and loin.



43

Fig. 5. Assessment of Group B (1968/69) boars: age-weight nomogram giving growth points
(new system).

Fig. 6. Assessment of Group B (1968/69) boars: fat thickness-weight nomogram giving fat
points (new system). Fat thickness is the mean of the thickness of the sol fat (right side plus left
side) and the midback fat.



44

of the reserve boars was only slightly above or slightly below average, for
boars of average grade were in the majority. The introduction of reserve boars
naturally reduced the differences between the sub-groups. A further weakness
is that the selection of the Group B boars was made over a period of time,
i.e. several months.

When the boars of Group A attainedbreeding age, a boar station, subordinate
to the North Finland Artificial Insemination Society, was established at the
testing station. Inseminations were done within a region of approximately 100
km radius. In most of this area sows had not been artificially inseminated pre-
viously, and initially the demand for semen was low. The semen was used
chiefly at small piggeries. Three of the test boars were housed at neighbouring
farms in order to make room for other boars. When the Group B boars selected
for progeny evaluation reached breeding age they were transported to the boar
station of the Salpausselkä AI Society. Here, in the Lahti area, AI of pigs was
already widespread.

3. Progeny and their rearing

For the progeny evaluation, the intention was to take one barrow piglet
from each of 16 litters sired by Group A boars, and one boar and one barrow
from each of 10 litters sired by the Group B boars. As AI was used the progeny
were randomly distributed to all parts of the two districts, except for the pro-
geny of the three Group A boars mentioned above, which came from 36 farms
only. Also in these cases matings to closely related sows were avoided. The
piglet or piglets from each litter were randomly picked.

The Group A progeny were reared at the North Finland Pig Husbandry
Experimental Station, and the progeny of Group B at the Hyvinkää Pig Hus-
bandry Experimental Station. The animals were housed 4 to a pen, and were
group fed. The pigs in any one pen were always the progeny of a single boar;
their ages and starting dates were not the same, because the pigs came from
2—4 litters. The feed portions were based on the average weight of the pigs
in the pen, got by weekly weighing. Amounts and composition of feed are given
in Tables 8 (1968 69), 9 and 10. At the beginning of 1969, during the rearing
of the progeny of Group A boars, the feeding levels and feed quality used in
Finland were changed, the former being raised for liveweights up to 65 kg. The
increase amounted to 0.05 0.4 feed units per animal per day, after which
the feeding level approximated to appetite. The feed mixture for the progeny
in 1968 was the same as that for the test boars (Table 8). From the beginning
of 1969 onward, each pig was given, separately, 150 g milk powder per day,
the rest of the feed being composed mainly of cereals (Table 10). This revised
feeding was used throughout the test of progeny of Group B boars.



Table 9. Feeding standards for progeny of test boars.

1968 (Group A) 1969-71 (Group B)

Weight of pig Weight of pig FU/pig/dayFU/pig/daykg kg

20.0-24.9 0.90
25.0-29.9 1.10
30.0-34.9 1.30
35.0-39.9 1.60
40.0-44.9 1.80
45.0-49.9 2.10
50.0-54.9 2.25
55.0-59.9 2.45
60.0-64.9 2.65
65.0-69.9 2.80
70.0-74.9 2.90
75.0-79.9 3.00
80.0-84.9 3.10
85.0- 3,20

20.0-22.4 1.20
22.5-24.9 1.30
25.0-27.4 1.40
27.5-29.9 1.50
30.0-32.4 1.60
32.5-34.9 1.70
35.0-37.4 1.80
37.5-39.9 1.90
40.0-42.4 2.00
42.5-44.9 2.10
45.0-47.4 2.20
47.5-49.9 2.30
50.0-54.9 2.40
55.0-59.9 2.60
60.0-64.9 2.70
65.0-69.9 2.80
70,0-74.9 2.90
75.0-79.9 3.00
80,0-84.9 3.10
85.0 3.20

Table 10. Composition of feed of boar progeny 1969—7l.

Barley meal 90.0 %

Fish meal 8.0 %

Mineral mixture 1) 1.5 %

Vitamin preparation2 ) 0,5 %

Dried milk 150 g/pig/day

') Composition of mineral mixture, %: fodder phosphate 38.0, ground limestone 38.8,
common salt 20.0, zinc oxide 0.65, iron sulphate 0.85, copper sulphate, manganese sulphate
0.85, cobalt sulphate 0.033 and potassium iodide 0.007.

2 ) Composition of vitamin preparation: A 800 000 i.u., D 3 160000 i.u., E 4000 mg, riboflavin
800 mg, D-pantothenic acid 3000 mg, niacin 5000 mg, vitamin 812Bl2 4mg per kg feed.

4. Progeny evaluation

The rate of growth and average feed efficiency of all progeny were determined,
and carcass characteristics were evaluated by measurements on the carcass
section and lean cuts analysis. Additional measurements were taken for the
Group B progeny, i.e. ultrasonic measurement of fat thickness, ham measure-
ment, and leg assessment.

The rate of growth of the progeny, as with the boars, was expressed in terms
of (i) daily growth between 20 and 88 kg liveweight and (ii) age at the weight
of 88 kg. Variation in initial and final weight was corrected for by interpolation,
with the aid of a computer. Determination of age at weight intervals of 10 kg
was also done by interpolation. Feed efficiency averages per pen and per boar

45



Fig. 7. Positions for fat thickness measurement of
porcine carcass section (Partanen 1965).
1 = withers, 2 = midback, 3a = fore loin, 3b =

mid loin (corresponding to loin fat measured
ultrasonically), 3c = rear loin, 4 = side fat.
The longissimus muscle is measured along the
horizontal line shown, and the side length is the
length of the short vertical line.

Fig. 8. Porcine carcass, showing the
most important parts (Uusisalmi
1969a).
1 = ham, 2 = loin, 3 = back, 4 =

foreback, 5 = shoulder.

46



47

were calculated for the liveweight interval 20 to 88 kg. Variation in initial and
final weight was corrected for, as with the test boars, by means of the factors
2.30 and 4.19 feed units per kg liveweight. Carcass section measurements and
lean cuts analysis were made by ordinary methods (Fig. 7, Partanen 1965;
Fig. 8, Uusisalmi 1969 a).

The ultrasonic measurements of fat thickness were made with Krautkrämer
USK 4 apparatus at the same loxations as for the test boars (Fig. 1). The
measurements were made twice, 8 weeks after the start of the test and just be-
fore slaughter, when the liveweight was close to 88 kg. Data obtained at the
latter stage were converted to fat points, as had been done with the test boars
by using the nomograms shown in Figs. 4 and 6. Growth points were determined
similarly (Figs. 3 and 5). Test points (growth -j- fat) according to the old
system were obtained by summing the growth and fat points got from the
nomograms in Figs. 3 and 4, and the test points according to the new system,
correspondingly, by using the nomograms in Figs. 5 and 6. The ham measure-
ments and leg assessments were made in the manner employed with the test
boars (p. 40, Fig. 2). Detailed records of leg defects were kept.

5. Statistical methods

Statistical analyses were carried out at The Agricultural Research Centre
using an IBM 1130 computer.

The effects of the linear regression— and class variables on the characteris-
tics of the animals were tested by least squares analyses (Harvey 1966, Neno-
nen 1972, Scheffe 1961). The regression variables were initial weight and age
or final weight and age. The class variables were year, testing station, feeding
level, castration, breed, group (A or B) and sire. Because variables often are
interrelated, the proportion of the total variation accounted for by the model
was also calculated. The general model (Draper and Smith 1966, p. 128) was
as follows;

y= a 0 +aj + ... +an +b, xx + ... +bm x m +e.
in which y = dependent variable, a 0 = constant, a 1...an = class variables,
bj. .. b m = regression coefficients, xx . .. x m = regression variables, and e. =

normally distributed random error with expected average of 0.
The significance of the differences between means was determinedby Tukey’s

test (Speel and Torrie 1960). Certain effects of the variables were eliminated
through the use of pooled estimates (Nenonen 1972): the sums of squares re-
quired for the calculation of the correlation matrix are calculated separately
for each subclass average. By adding up the sums of squares within each class,
estimates of the sums of squares, independent of the differences between the
classes, are obtained (compare with »Analysis of variance» Hard 1952, p. 412).
These sums of squares are divided by their degrees of freedom to give unbiassed
estimates of the covariances, which can be used in stead of the original covarian-
ces. The effects of testing station, feeding and difference between boars and
barrows on the various characterics were eliminated in this way.



48

The interrelationships between characteristics were studied by means of
correlation analysis, using both weight-corrected and uncorrected values. The
corrections were made by eliminating the weight variable (Nenonen 1972).
The correlations calculated from the corrected data are actually partial cor-
relations. The models best describing the dependent variables were obtained
by means of stepwise multiple regression analysis (Draper and Smith 1966,
Nenonen 1971). The relative and absolute importance of each independent
variable was also determined by the programme. The form of the expression
was as follows:

y=a0 + a lXl + a 2x 2 +•••+ a n xn +e,
in which y dependent variable, a 0 = constant, a x ... an = regression
coefficients, xx .. .xn = dependent variables, and e, = normally distributed
random error with expected average of 0.

Missing observations were automatically excluded from the analysis.
The heritability estimates of the characteristics were calculated both from

sire-progeny regressions (h2 = 2b) and half sib correlations of the progeny,
the regressions being calculated by means of the Nenonen (1971) programme
described above.

Calculation of h 2 from the half sib correlations was made by means of
hierarchic variance analysis (Sokal and Rohlf 1969) in accordance with a
programme prepared by Ming-Pi Mi (1962). In analysing data for all progeny
(Group A plus Group B), the average between-group sum of squares was cal-
culated as the first level. Within Group B the average boar-barrow sum of
squares was calculated, and within this the average sire sum of squares for the
boar-progeny and barrow-progeny. It was not possible to obtain an average
between-dam sum of squares, because in calculating the barrow- and boar-
progeny separately there was only one piglet per litter. The heritability estimate
in this case is given by

4 a;
h« =-,

■■■ a »

+ °l
where cr 2 = between-sire variance component, and a\ error variance com-
ponent.

The standard errors were calculated by means of the programme (Ming-
Pi Mi 1962), using the following relationship:

I / 32 h 1s e - (hI) " [' dT+r
where s.e. (h 2)= standard error, df = degrees of freedom

In order to determine the genetic correlations, both hierarchic variance
analysis and hierarchic covariance analysis (Ming-Pi Mi 1962) were carried
out. The genetic coefficient of correlation, r g , is given by

gsl 2rg = ]fä7A ’



where asl 2 = between-sire covariance component for variables 1 plus 2,
trs j = between-sire variance component, variable 1, crj = between-sire
variance component, variable 2.

Estimates of the standard error of rg were calculated by means of the pro-
gramme, using the following relationship:

v g g; y 2h* h*
where s.e. (hj) and s.e. (h|) represent the mean standard errors of the herita-
bilities for variables 1 and 2.

B. Results

1. Test hoars

a. Phenotype evaluation: results and factors affecting them

The averages for the various characteristics of the 60 test boars used in the
study are given in Table 11. Generally, the testing was done at live weights
close to the desired range (21.6—87.9 kg), though the range of single values was
considerable: initial weight 17.0—28.4 kg, weight at ultrasonic measurement
70—109 kg. The initial age range was 51—87 days, and that at ultrasonic mea-
surement 143—214 days.

The boars grew relatively slowly and had relatively thin layers of fat, the
restricted feeding and the feed mixture employed evidently being the cause.
The fat was thickest at the withers and thinnest at the midback. The fat points
and standard deviations in the old system were smaller than those in the
new system. The boars scored the same growth points, on average, according
to both old and new scales. The standard deviation of the new-scale growth
points (Fig. 5) was smaller than that of the old-scale growth points (Fig. 3).

8 of the boars (13 %) had to be withdrawn from the testing on account of
weak legs; 5 of these were from two litters. Leg weakness occurred in both
good and low-grade boars, rated according to test points. Of the 30 selected
boars, 4 had to be excluded on account of poor semen quality or difficulty in
obtaining semen.

Initial weight had a significant effect on the age at 88 kg, on the
weight 8 weeks after the beginning of the test and on feed consumption per kg
weight increase (Table 12). The boars starting the test at a relatively high
initial weight had a lower feed efficiency than those starting at a smaller
weight. The effect of initial weight on fat thickness was slight, the primary
effect being on the age at 88 kg. The early rate of growth of the piglet thus
had a considerable influence on the rate of growth measured during the test.

An increase in the weight at ultrasonic measurement
was associated with a highly significant increase in fat thickness (Table 12).
It can be seen from the linear regression coefficients that an increase of one kg
in the weight at ultrasonic measurement resulted in an increase in fat thickness

4 49



50

Table 11. Characteristics of the 60 test boars used in the study (Group A + Group B).

Standard
Average

,deviationCharacteristics

Initial weight, kg 21.6 2.8
Initial age, days 68.5 8.2
Weight at ultrasonic measurement, kg 87.9 10.4
Age at ultrasonic measurement, days 179.1 12.6
Weight after 8 weeks, kg 48.7 5.2
Weight increase per day (20 —BB kg), g 597.0 70.9
Age at 88 kg, days 179.8 14.5
FU/kg weight increase (20 —BB kg) 2.87 0.4
Fat thickness, mm

withers (1) 21.2 4.0
midback (2) 16.0 3.3
loin (3) 17.5 3.6
sol, right (4) 17.7 3.4
sol, left (5) 17.5 3.4
mean of 1,2, 3 18.2 3.4
mean of 2,4, 5 17.1 3.2
mean of 1,2, 3, i-ti 18.1 3.32

Growth points, old system 52.8 3.4
Fat points, » » 49.3 3.5
Test points (growth + fat), old system 102.1 2.9
Growth points, new system 52.8 1.4
Fat points, » * 51.4 5,4
Test points (growth + fat), new system 104.2 5.1
Side length, cm 80.8 4.3
Ham measurement1), cm 107.4 5.3
Legs, points 3.7 0.5

x) Hams measured on 30 boars only; for method, see Fig. 2.

of 0.19 0.23 mm. The greater was the age at ultrasonic measurement, the
thinner was the fat, but the effect of age was much less than that of weight.
When only the weight at ultrasonic measurement was taken as the regression
variable in the least squares analysis, fat thickness (at different measuring
points) per kg live-weight, was as follows

(1) withers 0.19 mm/kg
(2) midback 0.17 »

(3) loin 0.19 »

(4) sol 0.18 »

(5) x (l + 2 +3) 0.18 *

(6) x (2+4 +5) 0.18 »

These coefficients were used to correct the fat thickness of the boars selected
for progeny testing to a liveweight of 88 kg.

The weight at ultrasonic measurement had a highly significat effect on the
test points (old system), in which the thickness of the fat is corrected for age
but not for weight (see Figs. 4 and 6). The feed consumption per kg weight



51

Tabic
12.

Least
squares

analyses
of

relationship
between
initial
weight
and
age,
weight
and
age
at

ultrasonic
measurement,

groupand
breed,
and

growth
and

fat
thickness
of
the
test

boars
(n

=

60).

Regression
variables

Class
variables
Total

r
,

.

.

..Weight
at
ultrasonic
Age
at
ultrasonic

Lnaractenstic
Initial
weight

Initial
age

Group

Breed

°

measurement
measurement
r

b

R2%F

b

R2%F

b

R2%F

b

R2%F

R2%F

R2%F

R2%
Age
at
88
kg,
days

.

-1.91
8.4
9.l**
+1.28
32.2
34.9**-1
)--1
)-
0.1

0.1

25.7
27.8***
49.2

Growth
per

day,

(20-88
kg),
g

....
-1.04
0.1
0.1

-1.52
1.9
1.8-1
)--1
)-
0.0

0.0

32.6
30.7***
41.7

Weight
8
weeks

after
start
of

experiment,
kg

...
+0.86

13.1
9.5**
-0.02
0.1
0.1-1
)--1
)-
0.1

0.1

14.9
10.9**

24.3

FU/kg
growth

+0.04
4.4

B.6**
-0.01
2.0
4.0

-0.03
35.3
69.9***
+0.02

13.7
27.2***
5.1

10.1**
3.1
6.1*

73.3

Fat
thickness,

mm

withers
(1)

-0.12
0.4
0.5

-0.15
4.2
4.8*
+0.19

15.5
17.8***

-0.01
0.0
0.0

10.0
11.6**

8.0
9.3**
53.9

midback
(2)

-0.22
1.9
3.3

-0.03
0.2
0.4

+0.19
23.3
39.1***
-0.05
1.9
3.1

22.2
37.2***
1.0
1.6

68.4

loin
(3)

-0.05
0.1
0.1

-0.01
0.0
0.0

+0.23
29.4
42.2***
-0.08
4.5
6.5*

8.1

11.6**
2.2
3.2

63.0

sol
(4)

-0.26
2.5
3.2

-0.02
0.1
0.1

+0.22
27.6

36.2***
-0.07
3.8
5.0*

5.5

7.3**
5.3

7.0*

59.6

meanof
1,2,
3

...

-0.12
0.5
0.8

-0.06
1.0
1.5

+0.20
23.8
36.8***
-0.05
1.5

2.4

14.0
21.6***
4.2
6.4*

65.7

meanof
2,4,
5
2
)

.

-0.26
2.9
4.3*
-0.02
0.1
0.2

+0.21
28.0
41.4***
-0.06
3.0
4.5*

11.8
17.4***

3.6
5.4*

64.2

mcanof
1,2,
3,
ii

0.17
1.1
1.8

-0.05
0.7
1.1

+0.21
26.8
42.8***
-0.05
2.0
3.2

12.7
20.2***
4.3
6.9*

66.8

Old
points
for
fat

...

+0.13
0.6
0.9

+0.06
0.8
1.3

-0.20
21.9
34.8***
+0.13

11.7
18.5***

12.9
20.5***
4.2
6.6*

66.6

New
points
for
fat
.

+0.57
4.7
4.9*
+0.03
0.1
0.1

-0.18
7.5

7.7**
+0.13
5.2
5.3*

17.5
18.0***

4.3
4.5*

48.4

Old
test
points

+O.lB
1.7
1.8

+0.05
0.8
0.8

+0.13
13.1

13.5***
-0.03
0.8
0.8

21.8
22.3***
3.7
3.8

48.3

New
test

points
....

+0.57
5.3
4.9*
+0.04
0.2
0.2

-0.08
1.7

1.5

+0.03
0.2
0.2

20.2
18.5***

4.9
4.5*

42.3

Length
of
side,

cm
.

+0.15
0.5
0.4

10.04
0.2
0.2

+0.12
5.7

4.9*

+O.OB
2.8
2.4

9.6

8.2*

2.1
1.8

37.8

I*
P
<

0.05,
**

P
<

0.01,
***
P
<

0.001.

l
)

Weight
and

age
at
ultrasonic

measurement
are

themselves
indices
of

rate
of

growth
and
other
growth

characteristics.

2
)

(5)
is
the
left
sol
fat
thickness.



52

gain decreased with increasing weight at ultrasonic measurement, because in
the same litter the fastest growing boars were naturally the biggest, and the
fast-growing individuals were efficient feed converters. The age at measurement
was positively related to the quantity of feed required per kg of growth; that is,
slow-growing boars consume a relatively large amount of feed.

The two groups of boars (A and B) differed significantly (P<
0.05 or PcO.001) in characteristics other than rate of growth (Table 12). The
mean fat thickness for Group B was 3.1 mm less than that for Group A. This
appeared to be due mainly to the fact that when the piglets were being selected
the requirements were stricter for Group B than for Group A, because of the
longer period during which the selection for B was made. The score for the
parents was 363.3 for Group A and 610.1 for Group B. Admittedly the scores
are not quite comparable, because of the slight difference in the way in which
the fat thickness data affect the score.

Breeds differed primarily in respect of rate of growth: the Yorkshires
grew faster than the Landrace boars, the difference being highly significant,
and being found in Group B only (Tables 13 and 14). In the Yorkshires the fat
thickness at withers and sides was significantly greater than in the Landrace
boars. There was no significant breed difference in length of carcass.

b. Selection results
The test points distributions were as follows:

Group A
Test points
(old system) 96 97 98 99 100 101 102 103 104 105
No of boars 1216239411
Group B
Test points
(new- system) .... 94 99 101 102 103 105 106 107 108 109 110 114 115 116
No of boars 11242332322221

The test points (old system) showed a distribution closer to normal than did
those according to the new system. The Group A boar with only 96 points
had to be withdrawn on account of leg weakness, as were several others with
low test points.

The breed and parentage of the progeny-evaluated boars are shown in
Tables 13 and 14; in Group A 6 were Yorkshires and 5 were Landrace, and in
Group B the figures were 8 and 7 respectively. On account of the premature
loss of boars »Kantti», »Fakiiri» and »Piiska» of Group A they were each evaluated
on the basis of B—9 progeny only (Table 25). As each of the progeny was from
a different litter the evaluation was regarded as reliable enough for inclusion
in the analysis. The reserve boars of Group B (see p. 40) were »Reka» and
»Hymyri», the last two of sub-group I, and »Rata», the first of sub-group 111.
It is worth noting that the test points for the full and half litter mates were
very similar. As regards the quality, in Group A there was a marked division
on basis of breed, whereas this was not observed in Group B.

Tables 13 and 14 show the grouping of the boars according to test points.
In Group A, the poorest sub-group differed significantly (P<0.05 and P<0.01)



53

from the other sub-groups in respect of test points (old system), rate of growth
and feed efficiency. In Group B, the poorest sub-group differed from the others
(P<0.05 and P<0.01) in respect of test points (new system), fat points and thick-
ness of fat; the best and average sub-groups differed significantly (P<0.05)
only in respect of test points (new system). The rank order of Group A boars
(old system) was entirely due to the differences in growth points; the mean fat
points (old system) of the 3 subgroups were almost the same. This uniformity
in fat thickness is seen mainly in average backfat thickness x (1,2,3), the two
differences being 1.0 and 1.5 mm respectively. The clear differences in test
points between the sub-groups of Group B were due to the fat points (new
system), and this in turn was due to the difference (1.4 and 3.8 mm) between
the average midback- and sol fat x (2,4,5). The large differences in growth
rate (g/day) among the boars of Group B had no effect upon their grading. For
example, »Riku» and »Nakeri», the boars with the slowest growth, were placed
in the best sub-group because of their low fat thickness. The growth points
(new system), in fact, changed very little (50 55) despite the large differences
in rate of growth, which is due to the low gradient of the points scale (Fig. 5).
For the sake of comparison, Tables 13 and 14 include points other than those
used in grading the boars. It can be seen that if these points had been used the
boars of the high and the medium sub-groups would have been interchanged.
However, the same boars would have been allocated to the poorest group
according to either evaluation.

In quantitating the characteristics of the boars, fat thickness was corrected,
to correspond to a liveweight of 88 kg, by using the correction factors given
above (p. 50). Accordingly fat thickness is quoted in mm and mm x 10'1

,

though it was measured ultrasonically in mm only.

c. Interrelationships between characteristics

Old and new test points correlated significantly with all six fat thickness
values (Table 15). On the other hand correlations between test points and rate
of growth (kg/day and age at 88 kg) were low.

The effect of each characteristic on the test points was determined by means
of stepwise multiple regression analysis (see p. 48), in which all characteristics
expressing fat thickness and rate of growth were the independent variables.
The effects of differences in weight were eliminated from the data pooled in
respect of Group and breed. The following variables provided a statistically
significant independent increment of effect:

1 Mean fat thickness x (1,2,3) -22.81*** 90.1
2 Age at 88 kg - 7.o6*** 8.6
3 FU/kg growth (20-88 kg) + 3.20** 1.8

Total effect 91.2
New lest points

1 Mean fat thickness x (2,4,5) -34.57*** 95.4
2 Age at 88 kg -s.B7*** 2.7
3 FU/kg growth (20-88 kg) + 1.99* 0.3

Total effect.! 95.9



54

Table
13.

Test
results
of
boars
selected
for

progeny
evaluation,
Group
A

(1967).

Old
points

New
points

Growth

Thickness
of
fat
at
88
kg,

mm

Sub-group

[

FU/kg

"

-o

Age
w+

Average

and
name

Sire
x

dam

S__
20
—BB.

00

Wi-
Mid-

T

.

Sol

£

Gro-

Gro-

at
88

m

.
Loin

of
boar

ffl

Test

~

Fat
Test

~

Fat
kg,

,

thers
back

right

1,2,3,

wth

wth

k
g-

crease

1

o
i

oin,

.g/da
y

days

W

&

(3)

(4)

££3
2,4,5')
4+5

Sub-group
I.

1.
Nerokas
Nero
x

Hurma
L

105

56

49

104
53

51

638
169

2.57
20.3

14.6
14.3
17.5
16.5
16.2
16.6

2.
Nehru

»

L

103

58

45

104
54

50

645
161

2.44
23.2

18.6
19.2
16.4

20.4
16.8
19.2

3.
Hiiva

HolvixSiina
L

103

55

48

108
53

55

629
169

2.61
21.1

15.3
17.1
14.2
17.9
14.5
16.9

Average

103.7
ac
56.3
47.3

105.3
53.3
52.0
637
a

166
a

2.54c

21.5
16.2
16.9
16.0
18.3
15.8
17.6

Sub-group
11.

4.
Haku

Holvi
x

Miina

L

102

54

48

105
53

52

543
179

2.94
20.7
16.8
17.7
15.7
18.4
16.1
17.7

5.
Hokki

Holvi
x

Tiina

L

102

53

49

108
53

55

537
180

2.90
20.2
16.3
18.2
14.3
18.3
15.4

17.3

6.
Kantti

Kanttori
x

Puntti
Y

102

60

42

102
55

47

667
159

2.44
25.0

18.4
22.0
18.2

21.9
18.3

20.9

7.
Fakiiri

FollexHarava
Y

101

51

50

101
52

49

705
182

2.29
20.3
17.2

18.3
19.3
18.6
18.5
18.8

Average

101.8
a

54.5
47.3

104.0
53.3
50.8
613

ab
175

ab
2.64°
21.5

17.2
19.0
16.9
19.3
17.0
18.7

Sub-group
111.

8.
Piiska

PokerxLiisa
Y

99

50

49

101
52

49

569
194

2.99
24.7
17.5
18.7
19.6

20.3
18.9

20.1

9.
Fiksu

Follex
Hanska
Y

99

51

48

101
52

49

560
193

3.02
21.5
17.4

20.5
19.4

20.1
18.7
19.7

10.
Forkki

•

Y

99

50

49

99

52

47

602
195

2.83
24.1
17.9
19.1

20.0
20.3

19.6
20.4

11.
Suntio

Kanttori
x

Puntti
Y

97

52

45

96

53

43

540
183

3.23
25.4
21.3
21.4
21.4

22.7
21.4
22.4

Average

98.5
bd

50.8
47.8
99.3
52.3
47.0
568
b

191
b

3.02
d

23.9
18.5
19.9

20.1
20.8

19.6
20.6

Overall
mean

(sub-groups
I,
II
and
III)

101.1
53.6
47.5

102.6
52.9
49.7
603
179

2.75
22.4

17.4
18.8
17.8
19.6
17.7
19.1

The
significance
of
the

differences
between
the
sub-group

averageswas
tested
by
variance

analysis
and
Tukeys
test,
a

—b:
P
<

0.05,
c

—d:
P
<

0.01.

Breed,
L

=

Landrace,
Y

=

Yorkshire

*)

(5)
is
the
left
sol
fat
thickness.



Table
14.

Test
results
of

boars
selected
for
progenyevaluation.
Group
B

(1968/69).

Old
points

New
points

Growth

Thickness
of
fat
at
88
kg,

mm

Fu/kg

Sub-group

-c

a

a

°

r

5

Age

wt.

Average

and
name

Sire
x
dam

£

„

„

20-88
„

OQ

.

Wi-
Mid-

Sol

m

Gro-

Gro-

ats»
in-

Loin

of
boar

Test

Fat
Test

wth
Fat
kg,

crease
thers

back

right

1,2,3,

g/day
dasys 5 ys

(1)

(2)

(3)

(4)

1.2.3
2,4,5')

44-5

Sub-group
I.

1.
Rukki

Lukkoxliris
L

116

52

64

105

49

56

580
187

3.17
17.5
12.2
11.5
11.3
13.6
12.0
13.3

2.
Riku

»

L

115

52

63

106

48

58

491
217
3.86
12.9
11.6
13.9
13.7
12.7
12.6
12.9

3.
Nakeri

KalterixNaru
L

114

50

64

103
46
57

436
199

4.05
19.0
12.7
12.0
11.9
14.6
12.1
13.9

4.
Reka

VekaxTaara
Y

109

55

54

106
55

51

644
165

2.77
19.1
14.0
16.1
18.1
16.4
16.4
16.4

5.

Hymyri

Hymypoika
x

Arita
Y

108

55

53

105
58

47

739
156

2.34
20.5

14.6
17.5
16.6
17.6
15.9
15.9

Average

112.4ac
52.8
59.6ac

105.0C

51.2
53.8
578

185
3.24
17.
8
a

13.0C

14.2
a

14.3a

15.0ac
13.8

ac
14.5

ac

Sug-group
11.

6.
Rape

Reipas
x

Rae

Y

108

51

58

103

51

52

597
190

2.77
20.9
12.0
15.9
14.9
16.2
13.2
15.7

7.

Kilperi

Kasper
x

Friisi

Y

107

54

53

105

56

49

687
164

2.59
20.0
13.1
17.0
18.1
16.8
16.1
16.9

8.
Ripa

Reipas
x

Rae

Y

107

50

57

103
49
54

559
206
3.21

18.0
11.0
15.0
16.0
14.7
14.3
15.0

9.
Ranu

KalterixNaru
L

106

52

54

107

52

55

591
171

3.01
10.2
13.3
12.2
16.3
12.0
15.3
13.0

10.
Ruuppo

Kasper
x

Friisi

Y

105

55

50

105
58

47

731
158

2.37
20.9
14.1
18.9
19.0
18.0
17.1
18.1

Average

106.6
b

52.2
54.4

ab
104.6C

53.2
51.4
633
178

2.79
18.0

a

12.7°
15.8ab
16.9

ab
15.
5
a

15.2
ab

15.7»

Sub-group
111.

11.
Rata

Hymypoika
x

Arita
Y

105

55

50

102
56
46

694
163

2.51
22.4

15.5
20.4

18.5
19.5
17.5
19.2

12.
Sapro

Santsox
Nuoli

L

102

52

50

100

50

50

541
191

3.23
22.8

16.7
17.8
18.7
19.0
18.0
19.0

13.
Rooli

»

L

102

54

48

99

53

46

624
170

2.73
24.5
17.4
21.5
20.4
21.1
19.1

20.8

14.
Sapeli

Sake
x

Pirrikka
L

101

52

49

100
50

50

510
198

3.68
22.9
16.7
18.9
19.8
19.5
18.5
19.5

15.
Riski

Vinski
x

Lauha
Y

94

53

41

98

55

43

640
170

2.59
31.3
19.5

20.3
23.4
23.7
22.1
23.6

Average

100.0bd
53.2
47.6bd

99.8
d

52.8
47.0
602
178

2.95
24.8
b

17.2
d

19.8
b

20.
2b

20.6
bd

19.0
bd
20.4bd

Overall
mean

(sub-groups
I,
II
and
III)

106.6
52.8
53.9

103.1
52.4
50.7
604
180

2.99
20.2
14.3
16.6
17.1
17.0
16.0
16.9

Statistical
analyses

as
in
Table
13.

J
)

(5)
is
the
left
sol
fat
thickness.

55



56

Thus, variation in fat thickness accounted for almost all of the variation in
both old and new test points. New test points were more closely dependent on
fat thickness than were old test points. Despite the low correlations, the age
at a weight of 88 kg was of significance in accounting for variation in both
sets of test points, being a characteristic independent of the others. The effect
of rate of growth (age at 88 kg) on test points was greater than that of the
growth per day (not significant), because when the test points were calculated
the total age, not the growth during the testing period, was taken into account
(Figs. 3 and 5). The effect of feed efficiency was significant, but the influence
is undesirable from a breeding point of view: animals with low feed efficiency
obtained high test points, because of the low fat thickness and the high fat
points of these slow-growing boars (see Table 14). Feed quality was obviously
one of the reasons for the slow growth and high consumption of feed of some
boars. The feed mixture in question (Table 8) sometimes caused diarrhoea,
at these and other testing stations.

Feef efficiency and growth per day correlated closely (rp
= —o.B4***),

due partly to autocorrelation. The different fat thickness values, too, were
closely interrelated, and the ultrasonic measurements of fat at different lo-
cations gave highly consistent results (Table 15).

2. Progeny

a. Phenotype evaluation: results and factors affecting them

Altogether 142 (72 Y + 70 L) progeny of the boars of Group A and 299
(161 Y + 138 L) progeny of Group B were evaluated. Those from Group A
were barrows, whereas there were 134 boars and 165 barrows from B. The
imbalance withregard to Group B arose because the replacements were barrows,
only one boar in each litter not being castrated.

The initial weights of the two groups of progeny were almost identical, as
were the final weights, but the progeny from A were older than those from B
at the commencement of testing (Table 16). The slower growth and lower
feed efficiency of the A progeny is probably mainly due in to the poorer feeding
used initially (see p. 44). An attempt was made to eliminate the disrupting
effects of change in feed on the A progeny evaluation by dividing them into
three categories, i.e. those on the old feeding standard (43 pigs), new (59) and
both (40), and then pooling these before data analysis. Every boar of Group A
had progeny in all three categories, but the number of animals in each category
varied. Differences between testing stations did not affect the differences
between boars, for no boar had progeny at both stations.

The progeny grew faster than the sires (Tables 11 and 16), primarily due
to the more abundant feeding and the higher feed quality for the former
(Tables 7 and 10). Like the Yorkshire sires of Group B, the Yorkshire progeny
grew faster than the Landrace progeny (Table 16). Boar-progeny grew faster
than the barrows. Since the progeny were group fed there is no dataconcerning
individual differences in feed consumption.



Table
15.

Coefficients
of
correlations

between
performance

chracteristics
of
test

boars,
n
=

60

Characteristics
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

8.
meanof
1,2,
3

-0.25
+0.04
-0.15
+0.90
+0.85
+0.89
+0.76
+l.OO

9.
meanof
2,4,
51
)

....-0.26
+0.12
-0.15
+0.63
+0.89

+0.74
+0.97
+0.83
+l.OO

growth

-0.84
+0.60
-0.50
+0.21
+0.32
+0.36
+0.33
+0.32
+0.33
+l.OO

growth

-0.80
+0.58
-0.50
+0.20
+0.39
+0.42
+0.35
+0.36
+0.37
+0.95
+l.OO

12.
Old
points
for
fat

...

+0.50
-0.26
+0.33
-0.79
-0.82
-0.88
-0.74
-0.93
-0.80
-0.60
-0.62
+l.OO

13.
New
points
for
fat

..

+0.29
-0.15
+0.17
-0.62
-0.89

-0.74
-0.97
-0.82
-0.99
-0.35
-0.39
+O.Bl
+l.OO

14.
Old
test

points

-0.07
+0.17
-0.00
-0.81
-0.76
-0.79
-0.65
-0.90
-0.73
+O.OB
+O.Ol
+0.75
+0.72
+l.OO

15.
New
test
points

....
+0.07
+O.Ol
+0.04
-0.61
-0.84
-0.67
-0.94
-0.78
-0.96
-0.09
-0.12
+0.69
+0.96
+0.78
+l.OO

16.
Side
length,
cm

+0.06
+0.04
-0.06
-0.01
-0.02

+0.04
-0.04
+O.Ol
-0.07
-0.19
-0.15
+0.04
+0.06
-0.11
+0.02
+l.OO

P
<

0.05,
r
g
0.26,
P
<

0.01,
r
§

0.33,
P
<

0.001,
r
S
0.42

The
correlations

werecalculated
from
weight-corrected
data

pooled
in

respect
of
the

boar
groupsand

breeds
(see
p.

47)

x)

(5)
is
the
left
sol
fat
thickness.

57



58

Initial weight had a significant effect on the rate of growth of the progeny
(Table 17). Pigs starting at a high weight achieved a live-weight of 88 kg at
an early age, even though their growth per day at the station was lower than
that of pigs of low initial weight. For A progeny the negative effect of initial
weight on daily gain was apparent only during the first part of the evaluation
period.

The effect of hereditary factors on rate of growth of B progeny was statisti-
cally highly significant (Table 17). The sires accounted for 17.0 % of the total
variation in daily gain day of the B progeny (Table 17). When the effect of
the sires was taken as class variable in the least squares analysis, it masked
the breed differences completely. Since the boars were selected according to
quality of sire, irrespective of breed, the »breed differences» were, basically,
differences between individual boars.

8 weeks after the start of the testing the weight of the B progeny-boars
was very close to the expected 60 kg, but that of the progeny-barrows was
about 2 kg less (Table 18). The appropriate ultrasonic measurement was for
both categories taken at a weight of about 88 kg. The less restricted feeding,
the inclusion of barrows, and the genetic level of the dams together caused
an increase in the thickness of the fat of the progeny, so that the layer of fat of
the progeny was considerably thicker than that of the sires (Tables 14 and 18).
Therefore the progeny obtained fewer fat points than did the sires. The high
growth points did not make up for the low fat points, so that the progeny’s
test points were lower than the sires’ (Tables 11 and 18). The fat layer of the
barrows was significantly thicker than that of the boars. The difference at a
weight of 88 kg averaged 4.4 mm (Table 18). The boars’ average leg points
was slightly greater than the barrows’. The sires’ average ham measurement
was greater than the progeny’s (107.4 and 101.9 cm) respectively; for the
progeny, the mean for the boars was 103.5 cm and for thebarrows 100.7 cm (Table
18).

Variation in initial weight and age of the progeny had very little effect on
the ultrasonically measured fat thickness (Table 19). The effect of variation
in weight at ultrasonic measurement was highly significant in the early test
but much less in the regular test. The effect of age at ultrasonic measurement,
however, was greater in the regular test than in the early test. This contrast
is due to the fact that the early test was made according to number of days,
whereas the regular test was made according to the animal’s weight. When the
effects of variation in weight and age were eliminated by means of linear reg-
ressions, and the differences between the progeny boars and barrows were
eliminated by pooling, the effects of the sires on the variation in fat thickness
values in the preliminary test were found to be highly significant, while in the
regular test these effects were highly significant only with regard to variation
in loin fat thickness. The effect on loin fat is reflected in the respective averages
of both fat points and test points (Table 19).



59

Table
16.

Growth
and
feed
efficiency
of

progeny.
All
progeny

Progeny
of

Progeny
of
Group
B

Differences

Group
A (barrows)

boars

barrows
Progeny

Progeny
of
B

Characteristics

..„

.

•„

~,

~.

~

n
=

440

n
=

142

n
=

133

n
=

163

of
A

boars-

mean
s.d.

mean
s.d.

mean
s.d.

mean
s.d.
Y

L

Y

L

barrows

Initial
age,
days

60.8
9.7

64.4
13.1

59.4
7.2

58.9
6.8

+l.O
+

2.8

+

0.5

Initial
weight,
kg

19.2
3.8

19.4
4.2

19.3
3.7

19.0
3.6

-0.4
+

0.0

4-
0.3

Final
weight,
kg

90.6
2.4

91.3
2.2

90.6
2.6

90.1
2.3

+O.O
+

0.0

+

0.5

Age
at
20
kg,
days

64.5
11.5

70.2
13.6

61.3
9.7

61.8
9.5

-0.2
+

2.8*

-
0.5

Age
at
60
kg,
days

126.7
14.5
138.5

15.6
120.0
10.2
121.9

9.7

-1.0
+

0.1

+

1.9*

Age
at
88
kg,
days

159.9
15.7
173.0
17.4
151.5

11.8
155.2
11.4

-1.5

-
3.9*

-
3.7***

Growth
per
day

20-60
kg,
g

655.2
78.7
601.0
80.2
690.0
80.8
673.8
75.4
+6.4
+l7.3**
+16.2*

20-88
kg,
g

721.7
79.2
670.7
70.3
761.2
75.7
734.6
68.4
+7.2
+47.7***
+26.6***

60-88
kg,
g

865.0
125.6

826.6
108.6

907.8
127.4

861.3
134.5

-5.6
+Bl.o***
+46.s***

FU/kg
weight
increase

(20-88
kg)...2.83
0.21

2.99
0.14

2.74
+

0.21

Abbreviations:
Y

=

Yorkshire,
L

=

Landrace,
s.d.

=

standard
deviation.

The
significance
of
the

differences
wastested

by
least

squares
analysis
(cf.
p.

47);

•

P
<r
0
05

**

P
<■
0
01

***
P
<r
0
001

r

v.vd,
r
<.

v.vi,

r-

u.wi.



Table 17. Least squares analysis of the effects of initial weight, initial age and sires on the
growth of boar progeny.

Regression variables Class variables

Total
Characteristics Initial weight Initial age Sires

b R 2 % F b R 2 % F R 2 % F R 2 %

Group A, n = 142
Age at 60 kg -1.53 13.4 24.6***
Age at 88 kg -1.84 15.6 28.3***
Growth per day

20-60 kg -0.01 11.8 18.1***
20-88 kg -3.13 2.8 3.8
60-88 kg +O.Ol 5.4 7.5**

Group B, n 299
Age at 60 kg -1.89 40.5 290.4***
Age at 88 kg -1.99 33.5 189.0***
Growth per day

+0.49 12.1 22.1*** 7.81.4 30,6
+ 0.5512.7 23.0*«* 5.61.0 29.8

+O.OO 0.6 0.9 4.6 0.7 17.0
-0.10 0.0 0.0 3.2 0.4 6,5
-0.00 2.1 3.0 2.9 0.4 8.9

+0.90 26.8 192.1*** 6.1 3.l*** 60.8
+0.89 19.4 109.5*** 10.6 4.3*** 50.2

20-60 kg -0.016.1 21.8*** +O.OO 0.31.2 11.1 2.B*** 20.8
20-88 kg -3.773.2 11.7*»* +O.Bl 0.41.6 17.0 4.5»** 23.6
60-88 kg +O.OO 0.51.6 -0.000.0 0.012.3 2.9*** 13.4

The Group A data was pooled according to feeding, and that of Group B according to dif-
ferences between boars and barrows (see p. 47).

• P < 0.05, •* P < 0.01, *** P < 0.001.

60



61

Table 18. Fat thickness as determined by ultrasonic measurement, ham measurement and leg
score of the progeny of Group B: results of preliminary and regular test.

All progeny
n = 299

Boars
n = 134

Barrows
n 165 Differences 1)Characteristics boars-barrows

mean s.d. mean s.d. mean s.d.

Preliminary test
58.7 5.3 59.6 5.1 58.1 5.3 +1.5Weight, kg 58.7 5.3 59.6 5.1 58.1 5.3 +1.5

Age, days 122.1 47.6 122.5 51.3 121.7 44.7 +O.B
Fat thickness, mm

withers (1) 25.7 3.2 25.0 3.3 26.3 3.0 -I.3***
midback (2) 12.1 2.3 11.5 2.1 12.6 2.3 -1.1»»*
loin (3) 12.3 2.5 11.6 2.3 12.8 2.6 -I.2***
sol,right (4) 14.4 2.3 13.6 2.0 15.0 2.3 —l.4***
sol, left (5) 14.4 2.3 13.6 2.0 15.0 2.3 —l.4***
mean of 1, 2. 3 16.7 2.2 16.0 2.2 17,2 2.2 -I.2***
mean of 2,4, 5 13.5 2.1 12.9 1.9 14.1 2.2 -I.2***
mean of 1, 2. 3, 4+5 16.1 2.2 15.4 2.0 16.7 2.2 -I.3***

2
Ham measurement, cm 85.2 13.3 87.2 5.9 83.6 16.9 -J-3,6
Forelegs, points 4.5 0.6 4.5 0.6 4.5 0.6 ±0
Back legs, points 4.1 0,8 4.2 0.7 4.1 0.8 ±O.l
Regular test
Weight, kg 88.1 4.0 88.4 3.9 87.8 4.0 +0.6
Age, days 153.2 12.1 151.4 12.1 154.6 12.0 -3.2
Fat thickness, mm

withers (1) 35.0 4.6 32.5 3.6 37.0 4.4 -4.s***
midback (2) 18.1 3.8 15.8 3.0 20,0 3.3 -4.2***
loin (3) 18.4 4.4 16.0 3.3 20.4 4.1 -4.4***
sol, right (4) 21.9 4.3 19.4 3.7 23.9 3.7 -4.s***
sol, left (5) 21.9 4.3 19.4 3.6 23.9 3.7 - 4.s***
mean of 1,2, 3 23.9 3.8 21.4 2.9 25.8 3.3 -4.4***
mean of 2. 4. 5 20.7 3.9 18.2 3.2 22.6 3.3 -4.4***
mean of 1,2, 3, I+l 23.4 3,7 20.9 2.9 25.3 3.2 -4.4***

2
Ham measurement, cm ... 101,9 11.0 103.5 4.2 100.7 14.2 +2.8
Forelegs, points 4.3 0.7 4.4 0.6 4.2 0.8 +O.l
Back legs, points 3.8 0.8 3.9 0.8 3.7 0.9 +0.2
Test points, old system .... 98.6 4.5 101.4 3.7 96.3 3.7 +s.l***
Growth points, old system 58.0 2.8 58.6 2,9 57.5 2.7 +l.l
Fat points, old system ... 40.6 4.1 42.8 3.5 38.7 3.5 +4.l***
Test points, new system . 100.3 8.0 105.3 6.5 96.3 6.6 +9.o***
Growth points, new system 55.9 1.4 56.1 1.4 55.7 1.4 +0.4
Fat points, new system 44.4 7.6 49.2 6.3 40.6 6.3 +B.6***

*) Least squares analysis; effects of variation in initial weight, initial age, and weight and
age at ultrasonic measurement were eliminated by means of linear regressions, * P < 0.05,
*• P < 0.01, *** P < 0.001.



62

Table
19.

Least
squares

analysis
of
the
effects
of
initial
weight,
initial
age

and
weight
and
age
at

ultrasonic
measurementand

sires
on
fat

thickness

determined
by
ultrasonic

measurementand
fat

scoresof
the

progenyof
B
boars,
n
=

299.

Regression
variables

Class
variable

Weight
at

ultrasonic
Age
at

ultrasonic

Total

Characteristics
Initial
weight

Initial
age

Sires

°

measurement
measurement

b

R2%F

b

R2%F

b

R2%F

b

R2%F

R«
%

F

R2%
Preliminary

test

Fat
thickness withers

(1)

+0.04
0.2

0.8

+0.02
0.2

0.6

+0.24
14.3

53.0***
+O.OO
0.1
0.2

8.3
2.2**

24.7

midback
(2)

+0.05
0.5

1.8

-0.01
0.1

0.4

+0.17
14.6

56.9***
+O.Ol
1.1

4.1*

12.6
3.s***
28.3

loin
(3)

+0.05
0.4

1.6

+O.OO
0.0

0.0

+0.17
13.0

48.8***
+O.OO
0.8
2.8

10.7
2.9***
25.4

sol,
right
(4)

+O.OO
0.0

0.0

+O.Ol
0.0

0.1

+0.19
18.4

74.4***
+O.OO
1.0

4.0*

10.3
3.o***
31.0

meanof
1,2,
3

+0.04
0.4

1.8

+O.OO
0.0

0.0

+0.19
20.1
83.3***
+O.OO
0.6
2.7

11.3
3.3***
32.8

meanof
2,4,
5

+0.03
0.2

0.9

+O.OO
0.0

0.0

+0.17
17.1

69.3***
+O.OO
1.1

4.6*

12.2
3.s***
31.2

Regular
test

Fat
thickness withers

(1)

+0.14
0.8

2.4

+0.06
0.5

1.6

+0.14
1.6

4.8*

+0.04
0.7
2.1

6.3
1.4

9.2

midback
(2)

+0.03
0.0

0.2

-0.03
0.2

0.7

+O.OB
0.9
2.8

+0.03
0.5
1.5

6.4
1.4

9.1

loin
(3)

+0.13
0.8

2.7

-0.08
1.1

3.6

+O.lO
0.9
2.9

+0.07
2.2
7.2**

10.8
2.5**

15.0

sol,
right
(4)

+O.lB
1.5

4.7*

-0.08
1.1

3.3

+0.07
0.4
1.3

+0.07
2.7

B.3**

5.8
1.3

10.5

meanof
1,2,
3

+O.lO
0.7

2.2

-0.06
0.8

2.7

+O.ll
1.6

5.1*

+0.05
1.5

4.7*

8.2
1.9*

12.6

meanof
2,4,
5

+0.12
0.9

2.8

-0.06
0.7

2.1

+O.OB
0.7
2.2

+0.06
1.9

5.8*

5.8
1.3

10.1

meanof
1,2,
3,

....
+0.12
1.0

3.1

-0.06
1.0

3.0

+O.lO
1.4

4.5*

+0.05
1.9

6.1*

7.5
1.7*

12.4

Old
points
for
fat

-0.09
0.4

1.5

+0.06
0.7

2.5

-0.10
1.2

4.1*

+O.OB
3.6

12.5***
7.3
I.B*

19.8

New
points
for
fat

-0.15
0.4

1.1

+0.09
0.4

1.3

+O.lO
0.3
1.0

-0.08
1.2
3.5

5.6
1.2

7.2

Old
test

points

-0.06
0.2

0.8

+0.05
0.5

1.7

+0.26
6.5
24.2***
-0.14
10.1

37.4***
6.5

1.7*

25.1

New
test

points

-0.16
0.4

1.3

+0.09
0.4

1.4

+O.lB
1.0
3.3

-0.20
6.2
20.1***
5.3
1.2

14.9

Group
B
data

was
pooled
according
to

differences
between
boars
and
barrows
(see
p.
47).

*

P
<

0.05,
**

P
<

0.01,
***

P
<

0.001



63

b. Carcass evaluation and lean cuts of progeny

The progeny were slaughtered at an average weight of 90.6 kg (Table 16);
the average carcass weight was 66.9 kg (Table 20). In the conventional carcass
evaluation (Fig. 7) the fat measurements were taken at locations as close as
possible to those at which the ultrasonic measurements of fat had been made.
The location at which the loin fat had been ultrasonically measured corresponded
to 3 b, that is the mid loin. Ultrasonically measured fat thicknesses were
0.8—1.5 mm less than the corresponding thicknesses measured on the
carcass (Tables 18 and 20). The differences in the thickness of carcass fat
between the progeny boars and the progeny barrows were almost identical to
those obtained by ultrasonic measurement. The progeny of Group A had a
higher fat thickness than those of group B. Correspondingly, the cross-sectional
area of the longissimus muscle for A was smaller than for B, and for the barrows
smaller than for the boars. The boars had significantly longer carcasses than
the barrows’ (Table 20).

Variation in initial weight and age of the progeny had no effect on carcass
measurements (Table 21). Variation in carcass weight, however, was highly
significant, accounting for much of the variation in most of the characteristics.
The effects of variation in carcass weight were much greater that those of the
variation in weight at ultrasonic measurement (Tables 19 and 21). The more
accurate fat measurements and weighing results, and the differences in slaughter
losses (Table 20), evidently increased the effect of weight in accounting for va-
riation in the thickness of fat. When the effects of differences in weight were
eliminated by means of linear regressions, and the effects of feed and boar-
barrow differences were eliminated through pooling, the differences between
the sires were found to have a highly significant effect on most of the carcass
fat values (Table 21). The sires accounted for 8.1—29.0 % the overall variation
in the thickness of carcass fat in Group A, and for 9.1 16.7 % in Group B.
With regard to side fat the influence of the sires was the least, and for withers
fat in Group A and for loin fat in Group B the greatest. Sire differences were
much more pronounced in the carcass fat values of the progeny than in the
ultrasonic fat measurements of the live animal (Tables 19 and 21). The sire
differences were highly significant in accounting for variation in side length.
Sire differences accounted for 14.3 % of the total variation in the cross-sectional
area of the longissimus muscle in Group A, and for 17.8 % in Group B (Table
21).

In the lean cuts analysis (Fig. 8) the most valuable parts of the half carcass
(ham, loin, back, foreback and shoulder) were divided into two parts, namely
meat+bone and fat+skin. The total weight of the meat+bone plus fat+skin
is given as a percentage of the weight of the half carcass in Table 22. These
percentages give the best picture of the meatiness and fatness of the carcass.
By dividing the total weight of the meat+bone component by the age in days
the daily yield of valuable parts produced was got. The fatness or meatiness
of the carcass has was also expressed as the weight ratio (fat+skin): (meat +
bone) as totals for the most valuable parts of the carcass. A carcass score, Y,

10 A 2
derived from meatiness and meat yield, was calculated from Y = '

DU



64

Table
20.
Carcass

characteristics
of

progenyof
test

boars.
All

progeny
Progeny
of

Progeny
of
Group
B

Group
A

.„

t,

t,

Differences

All

progeny
Boars

Barrows

Characteristics
.......

.

boars

n
=

441

n
=

142

n
=

299

n
=

134

n
=

160

barrows

mean
s.d.

mean
s.d.

mean
s.d.

mean
s.d.

mean
s.d.

Carcass
weight,
kg

66.9
2.2

65.8
2.2

67.4
2.2

67.7
2.4

67.2
2.1

+0.5

Slaughter
loss,
%

26.2
2.2

27.9
1.9

25.4
1.8

25.3
1.8

25.4
1.8

-0.1

Length
of
carcass,cm

96.3
2.7

96.5
2.6

96.2
2.7

97.1
2.8

95.5
2.7

+l.7***

Length
of
side,

cm

76.2
2.3

75.9
1.9

76.3
2.4

77.1
2.5

75.7
2.4

+l.4***

Area
of

longissimus,
cm2

29.4
3.6

27.2
3.3

30.4
3.5

31.5
3.6

29.5
3.5

+2.o***

Fat
thickness,

mm

withers
(1)

36.6
4.1

38.3
4.6

35.8
3.9

33.5
3.8

37.7
4.0

-4.2***

midback
(2)

20.3
3.3

22.6
3.2

19.1
3.3

16.8
3.0

21.0
3.6

-4.2***

fore
loin
(3a)

26.5
4.1

28.2
4.2

25.7
4.0

23.3
3.7

27.7
4.3

-4.4***

mid
loin
(3b)

20.6
4.1

22.6
4.2

19.6
4.1

16.9
4.0

21.8
4.2

-4.9***

rearloin
(3c)

28.1
5.2

30.4
5.2

27.0
5.2

23.6
5.1

29.7
5.2

-6.2***

side
(4)

24.5
4.3

26.8
4.3

23.4
4.3

20.8
4.1

25.6
4.5

-4.B***

meanof
1,2,
3
1
)

27.3
3.3

29.3
.

3.4

26.4
3.3

23.9
3.2

28.4
3.3

-4.s***

meanof
1,2,
3,
4

26.6
3.4

28.7
3.4

25.6
3.3

23.1
3.3

27.7
3.4

-4.6***

meanof
2,4

22.4
3.5

24.7
3.4

21.3
3.5

18.8
3.3

23.2
3.6

-4.4***

Group
A
data
was

pooled
according
to
feeding,
that
of
Group
B

according
to

boars/barrows,
and
the

all-progeny
data

wassummed
overthe

different
testing
stations
(see
p.

47).
*

P
<

0.05,
**

P
<

0.01,
***

P
<

0.001.

!)

Mean
of
(3a),
(3b)
and
(3c).



Table
21.

Least
squares

analysis
of
the
effects
of
initial
weight,
initial
age,

carcassweight
and
sires

on
the

carcasscharacteristics
of
the

progenyof
test

boars.

Regression
variables

Class
variable

Characteristics

Initial
weight

Initial
age

Carcass
weight

Sires

b

R2%F

b

R2%F

b

R2%F

R2%F

R2%
withers
(1)

-0.01
0.0

0.0

-0.01
0.0

0.1

+0.40
3.0

5.7*

29.0
s.s'**
33.8

midback
(2)

-0.08
0.9

1.3

+O.Ol
0.3

0.4

+0.35
4.6

7.o**

8.8

1.3

16.3

loin
(3b)

-0.03
0.1

0.1

-0.04
0.9

1.5

+0.54
6.4

10.2**

12.4
2.0*

21.2

side
(4)

-0.12
1.0

1.6

+O.Ol
0.1

0.2

+0.55
6.2

9.7**

8.1

1.3

19.1

meanof
1,2,
3

-0.06
0.4

0.8

-0.01
0.1

0.2

+0.41
6.0

10.3***
18.5

3.2**

27.1

meanof
1,2,
3,
4

-0.08
0.7
1.2

-0.00
0.0

0.0

+0.45
6.9

11.8***
15.4

2.6**

26.0

meanof
2,
4

-0.10
1.2

1.9

+O.Ol
0.2

0.4

+0.45
6.8

10.7***
8.3

1.3

19.7

Area
of

longissimus
muscle,

cm2

+0.06
0.5

0.8

-0.00
0.0

0.0

+0.44
6.8

11.5***
14.3

2.4*

25.8

Side
length,
cm

+0.02
0.1

0.2

-0.00
0.1

0.2

+0.06
0.3

0.6

26.3
4.s***
27.2

Fat
thickness,

mm

withers
(1)

-0.02
0.0
0.1

+0.03
0.2

0.7

+0.32
2.8

B.9**

11.8
2.7***
13.2

midback
(2)

+O.Ol
0.0
0.0

+O.OO
0.0

0.0

+0.40
6.1

21.2***
14.8

3.7***
19.3

loin
(3b)

-0.00
0.0

0.0

-0.01
0.0

0.1

+0.45
5.1

17.7***
16.7

4.2***
19.7

side
(4)

+0.02
0.0
0.1

-0.00
0.0

0.0

+0.37
3.0

9.6**

9.1

2.1*

11.3

meanof
1,2,
3

-0.00
0.0

0.0

+O.OO
0.0

0.0

+0.38
5.7

19.8***
16.1

4.o***
19.6

meanof
1,2,
3,
4

+O.OO
0.0

0.0

+O.OO
0.0

0.0

+0.38
5.3

18.2***
14.1

3.4***
17.6

meanof
2,
4

+O.Ol
0.0
0.1

+O.OO
0.0

0.0

+0.39
5.0

16.4***
11.3

2.6**

14.9

Area
of

longissimus
muscle,
cm2

+O.Ol
0.0
0.1

-0.03
0.3

1.1

+0.39
5.0

20
2***

17.8
s.2***
31.4

Side
length,
cm

+0.05
0.5

1.9

-0.02
0.4
1.4

+0.23
3.9

16.1***
25.3
7.4***
32.0

Group
A
data
was

pooled
according
to

feeding
and
that
of
Group
B

according
to

differences
between
boars
and
barrows.

*

P
<

0.05,
**

P
<

0.01.
***

P
<

0.001.

5 tö



Table
22.
Carcass

characteristics
of
boar

progeny:lean
cuts.

Figures
are
in

grams,unless
otherwise
stated.

All
progeny

Progeny
of

Progeny
of
Group
B

Characteristics

Group
A

T-
.

,

T-,

a

Differences
1
)

Total

Boars

Barrows

'

Weight
of
joint

obtained
from
half

n
=

441

n
=

142

n
=

299

n
=

134

n
=

165

boars-

carcass

barrows

mean
s.d.

mean
s.d.

mean
s.d.

mean
s.d.

mean
s.d.

Half
carcass

32543
1295

32095
1145

32763
1363

32911
1533

32638
1195

+273

Shoulder,
meat
+
bone

2760
284

2685
352

2796
242

2920
238

2694
247

+226***

fat
+

skin

998

189

915

226

1042
169

977

154

1094
178

-117***

Foreback,
meat
+

bone

2581
289

3092
318

2328
276

2405
276

2265
278

+l4o***

fat
+

skin

854

171

1171
225

698

139

653

126

733

149

-
80***

Back,
front
half,
meat
+

bone

2477
275

2246
260

2587
280

2691
274

2506
284

+lBs***

fat
+

skin

1301
246

1397
305

1254
213

1153
194

1333
224

-180***

Back,
rear
half,
meat
+

bone

1153
188

1138
188

1154
188

1191
174

1125
198

+

66*

fat
+

skin

601

138

711

158

545

126

487

112

591

135

-104***

Loin,
meat
+

bone

1104
169

1276
231

1019
126

1049
129

994

125

+

55***

fat
+
skin

512

139

631

172

452

118

404
101

491

129

-
87***

Ham,
meat
+

bone

4704
418

4400
440

4853
405

5031
413

4713
397

+3lB***

fat
+
skin

1278
219

1280
215

1279
218

1137
208

1393
226

-256***

Lard

1066
340

962

304

1117
269

956

262

1247
275

-291***

Total
from
ham,
loin,
back,
foreback

and
shoulder

(main
part
of
carcass):

meat
+
bone

14779
1040
14837

922

14737
1085
15287

1086
14297
1089

+99o***

fat+skin

5544
714

6105
783

5270
679

4811
627

5635
710

-824***

meat
+

bone,
%

of
half

carcass
....45.4

2.7

46.2
2.4

45.0
2.8

46.5
3.0

43.8
2.7

+

2.7***

fat
+

skin,
%

of
half

carcass
17.1

2.1

19.0
2.2

16.1

1.9

14.6

1.9

17.3
2.0

-
2.7***

fat
+

skin/meat
+

bone

0.38
0.06

0.41
0.07

0.36
0.06

0.32
0.05

0.40
0.07
+

o.oB***

meat
+

bone,
g/day

91.7
9.5

84.9
9.5

94.9
9.5

99.7
10.0

91.0
9.1

+

B.7***

Carcass
score

520.3
74.1
491.5
70.2

534.0
76.3
579.3
80.2
498.1
72.2
+

81.3***

Group
A
data

was
pooled
according
to
feeding,
that
of

Group
B

according
to

boars/barrows,
and
the

all-progeny
was

summed
over

the
different

testing
stations.

x
)

Least
squares

analysis;
effects
due
to

variation
in

carcass
weight

wereeliminated
by
linear

regression.

*

P
<

0.05,
**

P
<

0.01,
***

P
<

0.001.

(iti



67

where A is the weight in g of the meat + bone of the valuable parts of the
half-carcass, B is the age in days at slaughter minus half the age in days at
a liveweight of 20 kg, and C is the weight of the half carcass in g.

This carcass score is in general use in Finland; the period of growth is divided
by two in order to reduce the effect of growth under uncontrolled conditions.

Total meat+bone as a fraction of the most valuable part of the carcass
as well as the fat+skin component in percent of the carcass was for A progeny
higher than for the B barrows. The fatmieat ratio (fat+skin/meat+bone)
of the two sets of barrows, however, was almost identical (0.40—0.41). The
meat yield per day and the carcass score of the B barrows were higher than for
the A progeny, this being the result of the faster growth of the former. For the
boars, meatiness was significantly greater and fatness significantly less than
for the barrows. The boars’ meat yield (99.7 g/day) and carcass score (579.3)
were likewise significantly higher than were those of their barrow sibs (91.0
g/day and 498.1) (Table 22).

Variation in initial weight and age had no effect on the lean cuts results
(Table 23). However, these factors had a highly significant effect on the daily
meat yield and, in the B progeny, on the carcass score, which is to be expected
because these characteristics depend on the early rate of growth. Variation in
carcass weight had a significant effect on the lean cuts results, i.e. mainly on the
weight results (Table 23).

When weight variation was eliminated by means of linear regression, and
feed variation in the A progeny and boar-barrow differences for B were eliminated
by pooling, the sire differences accounted for 5.8—21.3 % of the total variation
of the characteristics expressing carcass value. The sire effect was very uni-
form (R 2= 11.0—19.3 %) and highly significant for all carcass characteristics
in B progeny, while in A progeny it was highly significant only in yield of
shoulder and back meat (Table 23).

3. Results of phenotype and carcass evaluation and lean cuts analysis ofprogeny,
classified in accordance with grade of sire

The progeny are classified in Tables 24—27 in the same order as the sires
in Tables 13 and 14. The poorest sub-group (III) of the boars of Group A,
ranked according to old test points, proved to be the poorest also in the pro-
geny’s fat thickness and meatiness (Tables 24 and 25). There were statisically
significant differences between sub-groups I and 111 and/or II and 111 in
withers and loin fat, cross-sectional area of the longissimus muscle and yield
of meat+bone. It can be noticed that the sub-groups did not differ, in contrast
to the sires, in rate of growth and feed efficiency characteristics. Neither were
there, because of the uniform rate of growth, significant differences between the
sub-groups in respect of meat+bone yield per day or carcass score. Sub-
groups I and II did not differ significantly in respect of any of the characteristics.
This was the case also for the sires.



68

Table
23.

Least
squares

analysis
of
the
effects
of
initial
weight,
initial
age,
carcass

weight
and
sires

on
carcasscharacteristics

(lean
cuts)
of

progeny.

Regression
variables

Class
variable

Characteristics

Total

,

TT
~

.

~

,

Initial
weight

Initial
age

Carcass
weight

Sires

(Units
are
weight
unless

°

°

°

otherwise
stated)

b

R2%F

b

R2%F

b

R2%F

R2%F

R2%
Group
A,
n
=

142

Shoulder
(S)
meat+bone

-0.33
0.1

0.2

+0.25
0.6

1.0

+2.20
1.5

2.6

21.3
3.6***
25.8

Back
(B)

meat
+

bone

-0.13
0.0

0.0

+0.21
0.6

1.1

+5.38
12.1

23.4***
17.9

3.4***
34.5

Ham
+loin
(H,
L)
meat
+

bone

-0.12
0.0

0.0

-0.02
0.0

0.0

+7.74
11.6

21.1***
13.4

2.4*

30.8

S,
B,
H,
L
and
foreback
combined:

meat
+

bone

-0.21
0.0

0.0

+0.27
0.1

0.2

+18.5
15.9

28.9***
9.0

1.6

30.5

fat
+
skin

-0.65
0.1

0.2

+0.19
0.1

0.1

+18.4
21.8
41.5***
7.9

1.5

33.9

meat
+

bone,
%

of
half

carcass
+O.OO
0.0

0.1

+O.OO
0.1

0.1

-0.00
0.3

0.5

11.2
1.6

11.4

fat+skin,
%

of
half

carcass

-0.00
0.1

0.1

+O.OO
0.1

0.1

+O.OO
7.2

11.2**
10.5
1.6

19.1

fat
+

skin/meat
+

bone

-0.00
0.0

0.1

+O.OO
0.0

0.0

+O.Ol
5.4

B.l**
10.8
1.6

16.7

meat
+

bone,
g/day

+O.OB
9.6

15.9***
-0.03
8.9

14.7***
+0.06
1.3

2.2

5.8

1.0

23.8

Carcass
score

+2.10
1.2

1.7

-0.86
1.9

2.7

+0.94
0.1

0.1

7.6

1.1

11.2

Group
B,
n
=

299

Shoulder
(S)

meat
+

bone

+0.22
0.1

0.3

-0.19
0.2

0.8

+2.90
5.9

22.3***
12.9

3.s***
25.8

Back
(B)
meat+bone

+0.53
0.2

0.9

-0.67
1.0

4.5*

+6.18
10.2

48.1***
19.3

6.s***
40.7

Ham
+
loin
(H,
L)
meat
+
bone

+0.53
0.1

0.6

-0.29
0.1

0.6

+7.26
10.1

43.8***
14.5

4.s***
35.5

S,
B,
H,
L
and
foreback
combined:

meat
+

bone

+l.OO
0.1

0.5

-1.10
0.4

1.8

+19.3
13.1

66.1***
17.4

6.3***
44.7

fat
+

skin

-0.60
0.1

0.3

-0.43
0.1

0.5

+ll.l
11.0

39.3***
13.6

3.s***
21.6

meat+bone,
%

of
half

carcass
+O.OO
0.0

0.1

-0.00
0.2

0.6

-0.00
0.1

0.3

19.3
4.9***
21.4

fat
+

skin,
%

of
halfcarcass

-0.00
0.2

0.7

+O.OO
0.0

0.0

+O.OO
1.7

5.6*

17.7
4.l***
18.1

fat
+

skin/meat+bone
-0.00
0.2

0.6

+O.OO
0.0

0.0

+O.OO
1.1

3.5

18.9
4.4***

18.9

meat
+

bone,
g/day

+0.13
19.0

91.2***
-0.06
13.0

62.5***
+0.07
2.3

11.1***
11.0

3.B***
41.8

Carcass
score

+3.77
2.7

9.7***
-2.28
3.0

10.8***
+2.67
0.5

1.8

15.2'
3.9***
23.2

Data
pooling
as
in
Table
21.
*

P
<

0.05,
**

P
<

0.01,
***

P
<

0.001.



Table
24.
Growth,
feed
efficiency
and

carcassevaluation
of
the
progenyof
Group
A

boars
ranked

according
to
test
points
of
sires,
n
=

142.

c"

Growth

Thickness
of
fat,

mm

u
j>

Age
at.

FU/kg
wt.

g

e,

(20
88
kg)

withers
midback
loin
side

mean
mean
mean

LMA1
)

Sub-group
I.

1.

Nerokas

14

170.6

656.0

3.03

37.3°
20.7

20.9
a

25.0

27.7°
22.9

27.0°
27.2

2.
Nehru

13

167.1

677.7

2.99

35.9°
22.7
22.4
26.3

28.6a

24.5

28.1

27.5

3.
Hiiva

16

182.5

673.8

3.01

36.7C

22.9
23.6
28.0

29.4a

25.5

29.1

27.5

Average

43

173.4

669.2

3.01

36.5

21.9

22.0
26.2

28.4
24.1

27.8

27.2

Sub-group
11.

4.
Haku

13

172.6

654.5

3.00

36.5C

21.5

19.9
a

24.5

27.4C

23.0

26.7
C

29.9"

5.
Hokki

14

174.1

680.0

2.90

35.8°
21.8

21.2
25.8

27.
8
C

23.8

27.3»
26.9

6.
Kantti

9

173.6

670.6

2.98

36.3
C

22.4

21.9
26.1

28.4
a

24.3

27.8

27.0

7.
Fakiiri

8

164.6

702.7

2.86

38.3
a

23.1

22.1
27.0

29.5

25.0

28.9

27.1

Average

44

171.2

677.0

2.94

36.9

22.3

21.7
26.2

28.5

24.3

27.9

28.2

Sub-group
111.

8.
Piiska

9

176.9

661.0

3.05

39.5

22.6
22.2
27.1

29.6

24.9

28.9

24.8
b

9.
Fiksu

15

176.2

657.8

2.91

39.6
a

23.5

22.7
26.1

30.1

24.8

29.1

27.7

10.
Forkki

16

166.2

667.9

3.12

39.7
a

22.4
24.4
29.0

30.0

25.7

29.8

28.1

11.
Suntio

15

175.0

684.0

2.97

44.6
bd

24.7

26.2
b

28.7

33.2
bd

26.7

32.0
bd

25.
l
b

Average

55

173.6

667.7

3.01

41.0

23.3

23.9
27.7

30.8

25.5

30.0

26.3

Sub-group
differences:

The
significance
of
the

differences
was

tested,
and
the
averagesof
the
progeny

sub-groups
of
the
sires
calculated,
by
least

squares
analysis,
in
which

the
growth
per
day

regression
variables
wereinitial

weight
and
initial
age,

and
the

carcassvariables
included
the

carcass
weight.

Effect
of

changes
in
feeding

was
taken
into
account
when
the
data

was
pooled.
The

differences
between
the

averagesweretested
by

meansof
Tukeys
test.
The
significance
of
the

differences
between
the
progenyof
the
sires
is
shown
as
follows:
a
—b:
P
<

0.05,
c
—d:
P
<

0.01.

l
)

LMA
=

longissimus
muscle
area.

69



Table
25.

Lean
cuts

results
of
the
progenyof
Group
A

boars
ranked

according
to
the
test
points
of
the
sires,
n
=

142.

Meat
+

bone,
g

Main
parts
of
half

carcass

>■

meat
+

bone

fat
-f

skin

Carcass

Ö

S)

fat
+

skin/
meat
+

bone

score

■3
p

Back
Ham
+
loin

„.

~..„.
~

~

.

.

.

~

P

fe

%

of
half

%

of
half

meat
+

bone
g/day

S
•£

carcass

carcass

Sub-group
I.

1.

Nerokas

14

3388

5652

14699
45.5

6126
19.0

0.42

83.8

470.6

2.
Nehru

13

3541
a

5675

14795
46.2

6261

19.5

0.43

87.2

500.1

3.

Hiiva

16

3473

5729

14813
45.9

6024
18.7

0.41

80.8

473.5

Average

43

3447

5655

14699
45.9

6050
18.9

0.41

83.9

481.0

Sub-group
11.

4.
Haku

13

3620
abd

6032
d

15436
47.7

5857
18.0

0.38

88.5

525.8

5.
Hokki

14

3490

5842

15185
47.4

5774
18.0

0.38

87.6

522.3

6.
Kantti

9

3216cd

5613

14768
45.8

6372
19.8

0.43

83.8

481.2

7.
Fakiiri

8

3245

5510

14625
45.9

6231

19.5

0.43

87.5

503.1

Average

44

3448

5834

15159
46.9

6092
18.8

0.40

87.4

512.3

Sub-group
111.

8.
Piiska

9

3307

5509

14629
45.8

6119
19.1

0.42

81.7

467.6

9.
Fiksu

15

3335

5747

15066
47.1

5799
18.1

0.39

85.4

500.8

10.
Forkki

16

3334

5647

14693
46.1

6404
20.1

0.44

85.9

493.9

11.
Suntio

15

3183be

5383
e

14402
45.0

6324
19.7

0.44

82.9

468.5

Average

55

3284

5566

14686
46.0

6160
19.3

0.42

83.8

483.1

Sub.
group
differences:

P
<

0.05

I

—
III

P
<

0.05

II—III

II

—
III

II

—
III

Data
analysis

as
in
Table
24.
The
levels
of

significance
of
the

progenydifferences
for
the
11

sires
are
as

follows:
a
—c:

P
<

0.05,
d

—
e:
P
<

0.01.

Td



71

The poorest sub-group (III) of the Group B boars, ranked according to
new test points, also proved to be the poorest in the progeny’s lean cuts results
and cross-sectional area of the longissimns muscle (Tables 26 and 27). The daily
yield of meat-(-bone and the carcass score of sub-group 111 were significantly
(P<0.01) smaller than those of group 11. Sub-group 111, however, did not differ
significantly from the others in respect of fat thickness (Table 26). The only
significant difference in fat thickness was that between the ultrasonically
measured loin fat of I and 11. As for the sires, the rate of growth of the progeny
in sub-group I was lowest and that of II highest (Tables 14 and 26). The sub-
group differences in rate of growth and feed efficiency, however, were not
statistically significant (Table 26).

A least squares analysis was also made of the differences between individual
sires on the basis of the progeny results. The progeny of »Suntio», the poorest
boar of Group A, differed significantly from most of the others in their thicker
layer of fat and lower meatiness (Tables 24 and 25). The half sibs »Haku« and
»Hokki» obtained the best carcass scores on account of their meatiness. The
carcass scores did not, however, differ significantly from those of the other
boars.

For boars of group B, several of the progeny differences were statistically
significant (Tables 26 and 27). Boar »Rata« of sub-group 111 had the best
batch of progeny in respect of rate of growth. »Rata» progeny also had a thin
layer of fat and good meatiness, and attained the second best carcass score.
»Rata» was one of the reserve boars, and on the basis of its test results it was
only slightly below average (Table 14). The progeny of the other boars of sub-
group 111 were significantly poorer than the progeny of the other sub-groups
in respect of a number of characteristics. On the other hand, »Reka», of sub-
group I, had progeny with poor meatiness (Table 27). »Reka» was also one of
the reserve boars and on the basis of phenotype evaluation did not differ much
from those of sub-group II (Table 14). The progeny of the top grade boars
»Rukki», »Riku» and »Nakeri» were meaty but had low rates of growth, and
thus performed according to expectation (Tables 14, 26 and 27).

Sub-group II boars produced progeny of evenly good quality, the average,
in fact, being superior to that of I and 111. The uniformity was apparently
due to the fact that the sires were relatively fast-growing but did not have
thick layers of fat (Table 14). The progeny of full and half sibs too proved
to be very uniform, even when the sires differed rather clearly (Tables 14
and 27).

It is interesting that the mean fat thickness values of groups I, II and 111
obtained by ultrasonic measurement differed significantly only in respect of
loin fat. Apparently the differences were too small to be measured on the live
animal. There were very few significant differences in fat thickness even when
measurements were made on the carcass.



72

Table
26.
Growth

and
fat
thickness
of
the
progenyof
Group
B
boars

ranked
according
to
test

points
of
sires,
n
=

299.
Fat

thickness
was

measured

ultrasonically
on

the
live

animal
(u)
and
on

the
carcasssection

(c).

»

rt
te

i

Thickness
of
fat,

mm

Sires
o,

.5
*■

oo

"

fe

...

,*.

...

,

,„,

,

.

~,

~

~,

,

„

->

mean
mean
,'

of

.e
oo

oo
withers
(1)

midback
(2)

loin
(3)

side
(4)

mean1,2,
3

„

.

~.

1

~

,

...

cm'

ö

8

|

go
S"Si

£

H
c

«1
-o

O
2i-

fe
u

u

c

u

c

u

c

u

c

u

c

u

c

u

c

Sub-group
I.

1.
Rukki
15

102.8
150.6

757.6
2.75
34.5
34.6
16.8
17.
5
e

16.9"
18.4

20.4
21.9
22.7
25.0

efg

19.3
19.7

abc
22.2
24.2

abc
30.0

2.
Riku
19

100.3
159.5

a

696.6
ae

2.86
33.6
32.7

ae

18.3
18.2

e

18.0
18.7

21.5
22.8
23.3
24.7

ae£g
20.5
20.5
22.9
24.2

abc
32.7
e

3.
Naked
22

101.6
156.9

725.9
abc

2.78
34.6
36.1
17.2

20.1
17.4a

19.5
21.4
23.1
23.1
27.0

19.9
21.6
22.6
26.1
30.9abc

4.
Reka
18

100.8
150.9

739.4
2.80
34.3
35.7

17.4
18.2
17.8
18.8

22.1
23.9
23.2
25.6abc
20.4
21.1
22.9
25.1

28.3
fh

5.

Hymyri
20

100.6
153.5

772.6
b

2.55
36.1
36.4
18.0
19.8
18.5

20.0
21.7
23.5
24.2
27.0
20.4
21.6
23.6
26.1
29.1b

Average
94

101.1
154.3

738.3
2.75
34.9
35.1

17.6
18.8
17.8
19.0

21.5
23.0
23.5
25.8
20.1
20.9
22.9
25.1
30.3

Sub-group
11.

6.
Rape
19

97.6
151.9

772.8
b

2.70
36.5
39.1

df

19.3
22.3

bf

20.8
22.0
£

23.6
25.4
25.5

29.2
dfh
22.2
23.8df

25.1
28.3

d£
29.3
b

7.

Kilperi
22

97.1
155.8

741.5
2.72
36.4
35.5
20.0
20.0
21.6
b

20.1
23.4
24.3
26.0
26.9
22.2
22.2
25.3
26.2
32.9

ae

8.
Ripa

18
100.7
149.1

b

796.3
df

2.62
36.1
36.3

18.5
18.4a

19.0
19.3

22.0
22.6
24.5
25.8

abc
20.8
20.5
23.9

25.0
29.6b

9.
Ranu
22

102.7
152.4

719.2
abc

2.85
34.7
35.8
17.6
19.6
17.
2a

19.3
20.2
21.4
a

23.2
26.4
19.4

20.5
22.5
25.2
32.0

efg

10.
Ruuppo
24

100.6
152.6

759.4
2.66
35.4
34.6

abc
17.8
17.
2e

18.0
16.4

ae
21.8
21.8
23.7

24.5e20.5
19.5

ac

23.2
23.8ae
31.3

efg

Average
105
100.2
152.0

758.3
2.71
35.6
36.1

18.5
19.6
19.3
19.5

21.9
23.0
24.4
26.6
20.8
21.3
23.8

25.7
31.4

Sub-group
111.

11.
Rata
21

100.3
150.
l
b

800.7
df

2.55
35.6
34.5

abc
17.4
17.4e

18.9
16.6

ae
22.4
22.1
24.0
24.4e

20.7
19.7

abc
23.6
23.8

ae
31.1

abc

12.
Sapro
18

99.8
151.0

708.6e

2.87
34.3
36.7
18.6
19.9
17.5a

21.
l

b

21.8
24.2
23.4

27.1
20.8
22.0
23.1
26.4
30.7

13.
Rooli
21

99.7
161.
3
a

712.8e

2.92
34.4
37.
3
b

18.8
20.2
17.8

22.1
f

22.0
25.9b

23.6
28.2b£

20.9
23.
l
b

23.2
27.6
b

28.5fh

14.
Sapeli
22

101.1
153.0

735.5
2.77
35.0
36.2

18.1
19.3
17.9

20.9
b

21.3
24.3
23.7
26.8
20.2
21.8
23.0
26.2
31.0

abc

15.
Riski
18

98.7
153.0

754.2
2.69
32.5
36.0
17.8
18.
7
a

18.8
21.7
f

22.9
24.1
23.0
26.8
21.3
21.4
23.0
26.2
27.4d£h

Average
100
100.0
154.0

746.6
2.76
34.7
36.0
18.3
19.0
18.2

20.3
21.9
24.0
23.7
26.5
20.7
21.5
23.3
25.9
29.7

Sub-group
differences

P
<

0.05

I
—II

P
<

0.01

II
—III

')

The
ultrasonic
measurement

was
performed

on
both
sides,
the

carcassmeasurementon
one

only.

2
)

LMA
=

area
of

longissimus
muscle.

The
significance
of
the

differences
wastested
as
described
in
the
footnote
to

Table
24,
except

that
boar
and
barrow
data

were
pooled.
The
significance
of
the

differences
between
the
progenyof
the
15

sires
is
given
as

follows:
a—d:
P
<

0.05,
e—h:
P
<

0.01.



Tabic
27.

Lean
cuts
of

progenyof
boars
of
Group
B

ranked
according
to
the
test
points
of
the
sires,
n
=

299.

Meat
+
bone

Main
parts
of

half
carcass:

o
S

back
ham
+loin

o/ o0
f
half

%

oi
half

meat4-bone
g/day

Sub-group
I.

1.

Rukki

15

3782

5969

14897
45.5

4788
ae

14.6
abce

0.32
ae

97.6

555.3

2.
Riku

19

3995
ef
K

6123
abce

15660
ae

47.4
e

5119

15.4
a

0.33
ae

96.7

568.6
a

3.
Nakeri

22

4018e

5971
a

15201
efs

46.4
aef
«

5209

15.9

0.35

95.8

553.5

4.
Reka

18

3429dfh

5538bf

13947dfh

42.9
bfh

5305

16.4

0.39

90.4
d

485.6
b

5.

Hymyri

20

3572
fh

5639
d

14193th

43.4
bf

5467

16.7

0.39

93.2

507.7

Average

94

3770

5848

14780
45.2

5163

15.8

0.36

95.0

535.9

Sub-group
11.

6.

Rape

19

3684
b

5922

14760
44.8

5563
b

16.8

0.38

96.1

541.5

7.

Kilperi

22

3680f

6066abce
14947

abc

45.5

5220

16.0

0.36

96.5

550.1

8.

Ripa

18

3670
b

5841

14592
b

44.7

4944abc
15.0»b

«

0.34
a

97.1

549.4

9.
Ranu

22

4081
ac

6036
a

15393aef
«

46.7
aef
«

5052

15.3
e

0.33
ac

97.3

563.8
a

10.

Ruuppo

24

3697

6065abce
15059ef

s

45.8
a

5022

15.0
C

0.33
a

98.3
a

564.6
a

Average

105

3793

5986

14950
45.5

5183

15.7

0.35

97.7

557.6

11.
Rata

21

3716

6001
a

14770
45.0

5073

15.4
a

0.34

99.6
abce

566.0a

12.
Sapro

18

3699
b

5741

14363f

44.9

5681
df

17.7
bf

0.40b

89.7
bf

500.6

13.
Rooli

21

3758
ab
<=

5636d

14480f

44.0f
5566
b

16.9
d

0.39

90.4
d

495.2

14.
Sapeli

22

3742
b

5794

14477
bf

44.1
f

5464

16.6

0.38

92.7

508.5

15.
Kiski

18

3489"1

5614
d

14003fh

42.7
bfh

5644
1

'

17.2
d

0.41
bf

90.0d

482.
5
b

Average

100

3692

5758

14419
44.3

5447

16.7

0.38

92.6

514.2

Sub-group
differences

P
<

0.05

II
—III
I
—III
I
—III

I
—III

P
<

0.01

II
—III
II
—III

II
—III
II

111

II
—III

II
—III

Statistical
analyses

as
in
Table
26.

73



74

When Groups A and B were analysed together, the progeny of sub-groups
I and 111 were found to differ significantly (PcO.001) in respect of a number of
carcass characteristics, as well as for daily yield of meat+bone and carcass
score.

These results show that phenotype testing was sufficiently accurate for re-
cognising the boars with the poorest carcass characteristics, but was not able
to improve growth or feed efficiency characteristics.

4. Heritability of characteristics, and phenotypic and genetic correlations between
characteristics

a. Sire-progeny correlations

The daily gain (20—88 kg) of sires and progeny at the testing station corre-
lated significantly, both when Groups A and B were considered together (r =
0.48*) and for Group B alone (r = 0.50*) (Tables 28 and 29). On the other hand
when rate of growth was analysed in terms of age at a weight of 88 kg the
sire-progeny correlation was low. The feed efficiency of the sires correlated
significantly with that of the progeny (r =0.42* and 0.49*), and with the
daily gain of the progeny (r= —0.48* and —0.50*) (Tables 28 and 29).

The rate of growth and feed efficiency of the sires did not correlate signifi-
cantly with the fat thickness of the progeny. In fact, the coefficients were
in different materials of opposite sign (Tables 28 and 29). The daily gain and
feed efficiency of the sires correlated significantly only with the meat + bone
(back) and meat +bone in percent of the most valuable parts of the carcass
of the progeny. Nor was this result, from a breeding point of view advantageous,
as the progeny of the sires that grew rapidly and consumed little feed were less
meaty.

The fat thickness of the sires had the closest relation to the sidefat and cross-
sectional area of the longissimus muscle of the progeny (Tables 28 and 29).
The various fat measurements of the sires were of almost the same value in
predicting the fat thickness of the progeny. The average for all fat thickness
measurements of sires and progeny correlated more closely (r= 0.42*) than did
the other two fat-thickness averages (Table 29). As regards fat thickness values,
the sire (ultrasonic)-progeny (ultrasonic) correlation was not as close as the sire
(ultrasonic)-progeny (carcass) correlation (Tables 28 and 29). The least useful
ultrasonic measurement was that of the progeny’s withers fat, since the sire-
progeny correlation was negative (Table 28). When measured on the carcass,
withers fat was as valuable as the other fat thickness values (Table 29).

The fat thickness measurements of the sires correlated negatively with the
new test points for the progeny, the correlations for withers and loin fat being
statistically significant (Table 28). The sire-progeny correlation for new test
points was positive but not statistically significant, as was the association
between ham measurements. The sires’ ham measurement correlated nega-
tively with the fat thickness measurements of the progeny, except for withers
fat. The correlation between the ham measurement and sol fat was statisti-



Table
28.
Group
B

sire-progeny
correlations
for
growth

characteristics
and
fat
thickness,
n
=

15.

\

Thickness
of
fat,

mm

Sires
Age

Growth
FU/kg
wt.

Test

Ham

at
88

per
day,

increase

mean
points,

mea-

i

,™

do
i

>

,s>n

oo
i

x

withers
midback
loin
sol,

right
mean
mean
•.

0

■>

kg,

(20
—BB
kg)

(20
—BB
kg)

°

1.
2,
J,

new

surement,

days

g

(!)

(2)

(3)

(4)

1,2,
3

2,4,
51
)

i~

system

cm

Age
at
88
kg,

days...
+0.12
-0.22
+0.26
-0.03
+O.lO
+O.lO
-0.01
+0.04
-0.01
+0.02
+O.ll
-0.13

Growth
per
day (20-88kg),

g

....

-0.28
+0.50
-0.50
+0.24
-0.15
+O.lB
+0.03
+0.15
-0.00
+0.12
-0.08
-0.16

FU/kg
wt.
increase

(20-88
kg)

+0.34
-0.55
+0.49
-0.23
+O.lO
-0.20
-0.05
-0.17
-0.02
-0.14
+0.06
+0.26

Fat
thickness,

mm

withers
(1)

-0.17
+0.34
-0.30
-0.21
-0.49
-0.03
-0.23
-0.23
-0.34
-0.25
+0.22
+0.29

midback
(2)

+0.05
+0.12
-0.18
+0.13
-0.04
+0.31

+0.26
+0.16
+0.13
+0.17

-0.24
-0.20

loin
(3)

-0.16
+0.36
-0.40
+0.22
-0.15
+0.26
+O.lB
+0.17
+0.04
+0.15
-0.15
-0.28

sol,
right
(4)

-0.20
+0.35
-0.44
+0.56
+0.16
+0.52
+0.41
+0.49
+0.30
+0.46
-0.35
-0.50

mean1,2,
3

-0.13
+0.34
-0.35
+0.04
-0.28
+0.19
+0.06
+0.02
-0.08
+O.Ol
-0.05

-0.07

mean2,4,
5
1
)

-0.12
+0.30
-0.39

+0.47
+0.12
+0.49

+0.39
+0.42

+0.27
+0.41
-0.36
-0.44

mean1,2,
3,

li-5
-0.15
+0.36
-0.41
+0.17
-0.19
+0.28
+0.14
+0.14
+O.OO
+0.12
-0.13
-0.17

Test
points,

new
system

+0.15
-0.30
+0.38
-0.50
-0.19
-0.51
-0.42
-0.47
-0.32
-0.45
+0.38
+0.47

Ham
measurement,
cm

+0.33
-0.33
+0.45
-0.49
-0.33
-0.36
-0.40
-0.45
-0.39
-0.45

+0.41
+0.39

The
correlations

werecalculated
from
data

corresponding
to
a

liveweight
of
88
kg.
P
<

0.05,
r
>

0.49,
P
<

0.01,
r
>

0.62,
P
<

0.001,
r
>

0.73.

i)

5

=

fat
thickness,
sol,
left.

75



Table
29.

Sire-progeny
correlation
for
growth,

feed
efficiency,
fat
thickness
and

carcass
characteristics,
n
=

26.

\

Thickness
of
fat,

mm

Sires
Age

Growth
FU/

Test

side

at
88

per
day

kg
wt.

mean
points,

,

..

kg,

(20
88
kg)

increase
withers
midback
loin
sol,

right
mean
mean
1,2,
3,

new

days

g

(20-88
kg)

(1)

(2)

(3)

(4)

1

2
_

3

2
,4,

5

i±s

system

Age
at
88
kg,
days
.

+0.07
-0.29
+0.29

+O.OO
-0.01
+0.07
-0.12
+0.03
-0.11
-0.02
+0.17
+0.17

Growth
per
day

(20-88
kg),
g

....

-0.25
+0.48
-0.48
+0.20
-0.05

+0.19
+0.04
+0.15
+0.03
+0.13
-0.09

+0.17

FU/kg
wt.
increase

(20-88
kg)

+0.28
-0.45
+0.42
-0.12
+0.06
-0.20
-0.01
-0.12
+O.Ol
-0.09
+0.02
+0.16

Carcass
evaluation

Fat
thickness,

mm

withers
(1)

+0.20
-0.14
+0.13
+0.34

+0.38
+0.26
+0.45

+0.35
+0.45
+0.39
-0.51
-0.23

midback
(2)

+0.16
-0.18
+0.13
+0.15
+0.22
+0.17
+0.12
+0.19
+0.13
+0.17
-0.23
+0.25

loin
(3)

+0.21
-0.20
+0.16

+0.43
+0.47
+0.29
+0.37
+0.42
+0.41
+0.43
-0.48
+0.13

side
(4)

+0.13
-0.02
+O.OO
+0.51
+0.42
+0.43
+0.41
+0.50
+0.41
+0.50
-0.41

-0.02

mean
1,2,
3

+0.16
-0.16
+0.13
+0.35
+0.40
+0.28
+0.35

+0.36
+0.37
+0.38
-0.44
+0.03

mean2,
4

+0.15
-0.10
+0.07
+0.38
+0.35
+0.34

+0.30
+0.38
+0.30
+0.37
-0.35
+0.12

mean1,2,
3,
4

...

+0.16
-0.13
+O.lO

+0.40
+0.42
+0.32
+0.37
+0.41
+0.39
+0.42
-0.45
+0.04

Longissimus
area,cm2

+0.12
-0.13
+0.20
-0.57
-0.42
-0.32
-0.38
-0.49
-0.41
-0.49
+0.39
+0.07

Length
of
side,
cm

.

+0.20
-0.42
+0.31
-0.21
-0.07
-0.32
-0.34
-0.23
-0.27
-0.27

+0.17
+0.62

Lean
cuts,
half

carcass
Meat
+

bone,
back,
g

+0.28
-0.49
+0.47
-0.62
-0.42
-0.55
-0.63
-0.60
-0.60
-0.63
+0.53
+0.38

»

ham
+

loin,
g

+0.14
-0.21
+0.20
-0.55
-0.57
-0.41
-0.58
-0.55
-0.60
-0.59
+0.54

+0.17

Main
part: meat

+

bone,
g

...

+0.29
-0.36
+0.38
-0.69
-0.56
-0.47
-0.60
-0.63
-0.62
-0.66
+0.58
+0.15

»

»

,

%

•

+0.37
-0.42
+0.43
-0.69
-0.54
-0.48
-0.60
-0.64
-0.60
-0.65
+0.57
+O.lO

fat
+

skin,
g

-0.13
+0.16
-0.18
+0.63
+0.62
+0.49
+0.56
+0.63

+0.58
+0.64
-0.61
+0.20

»

»

,

%

....

-0.12
+0.14
-0.17

+0.61
+0.62
+0.47
+0.56
+0.61
+0.58
+0.62
-0.60
+0.16

fat
+
skin/ meat

+

bone....
-0.20
+0.23
-0.26
+0.68
+0.64
+0.49
+0.62
+0.66

+0.63
+0.68
-0.63
+0.07

meat
+

bone,g/day
+0.07
+0.04
-0.01
-0.51
-0.48
-0.31
-0.44
-0.48
-0.46
-0.49
+0.42
-0.09

Carcass
score

+0.20
-0.13
+0.16
-0.62
-0.57
-0.40

-0.56
-0.58
-0.58
-0.60
+0.54
+O.OO

The
correlations

were
calculated
from
data
for
a

liveweight
of
88
kg.
P
<

0.05,
r
>

0.40,
P
<

0.01,
r
>

0.51,
P
<

0.001,
r
>

0.62.

76



cally significant (r= —0.50*) (Table 28). The side-length of the sires (mea-
sured on the live animal) and that of the progeny (carcass measurement)
correlated significantly (r=o.62***) (Table 29).

The correlation between the fat thickness measurements of the sires and
the lean cuts results of the progeny was generally very close, being highly
significant for most characteristics (Table 29). The progeny's lean cuts results
are a better measure of carcass value than ultrasonically measured fat thick-
ness. Therefore it is encouraging to note that the correlation between the ultra-
sonic measurements on the sires and the carcass measurements on the progeny
was quite high. Boars with thin fat produced better progeny than boars with
thick fat, even in terms of carcass score (P<o.os^—P<o.ool). The 3 fat thick-
ness averages of the sires were almost equally good in predicting the lean cuts
results of the progeny, though the overall average was slightly better (Table
29).

The test points for the sires proved to be very reliable in predicting the
meatiness and fatness of the progeny. The progeny characteristic showing the
weakest correlation with the fat thickness measurements of the sires was the
daily yield of meat-f-bone (P<0.05), a result of the low or negative correlation
between rate of growth and carcass value.

b. Heritability estimates

For estimating the heritability (h 2 ) of the various characteristics of the live
pig and its carcass the amount of data was somewhat limited. But at least it
was representative, in the sense that the progeny were distributed at random
among a large number of farms, this advantageous situation arising from the use
of AI. The sole exception was the three boars mentioned above (p. 44), which
were mated with dams on 36 farms only. In Group A each sire produced
B—l 6 progeny (barrows), and in Group B 6—ll boars and B—ll barrows
(see Tables 24 and 26). Whenever a boar piglet of Group B was replaced with
a barrow, one of the other barrows was excluded in the calculations.

The heritabilities were calculated from half sib correlations of the progeny,
and from sire-progeny regressions (see p. 48); comparison of the two h2-values
obtained shows how reliable the phenotype evaluation was in measuring the
breeding value of the sires. The degree of relationship between the half sibs
could be taken as 0.25, since, owing to the large number of farms involved, the
dams were not closely related. The sire-progeny regressions for fat thickness
were calculated from (i) ultrasonic measurements on sires and progeny and (ii)
ultrasonic measurements on sires and carcass measurements on progeny (Tables
30 and 31). Similarly, the half-sib correlations were calculated from both
ultrasonic and carcass measurements on the progeny.

The differences between the progeny boars and the progeny barrows of
Group B were eliminated by pooling when h 2 was calculated from the sire-
progeny regression. When h 2 was calculated from half sib correlations the boars
and the barrows were assigned a level of their own in the hierarchic variance
analysis. When h 2 was calculated from the entire data the differences between

77



78

the two Groups of boars were eliminated either by pooling or by assigning them
a level of their own in the hierarchic variance analysis. As to the hierarchic
variance analysis, Tables 30 and 31 show the variance of the different levels as
percentages of the total variance, and the significance levels of the between-
sire differences.

The estimates from the half sib correlations for the daily gain (20—88 kg)
in Group B and Group A + Group B were reliable and
respectively); the values calculated from the sire-progeny regressions were
lower (0.35 and 0.32). The age at 88 kg proved to be an unsatisfactory measure
of rate of growth, according to the h 2-value. The h2-values for feed efficiency
could be calculated from sire-progeny regressions (0.18 and 0.25) but. because
the animals were group fed, not from half sib correlations (Tables 30 and 31).

The heritabilities of the ultrasonically measured fat thickness measurements
were low, only the loin fat value being of any importance (Table 30). This result
is due primarily to the progeny rather than to the sire measurements, for the
h2-values calculated from regressions of ultrasonic measurements (sires) on
carcass measurements (progeny) were fairly close to the reliable h 2-estimates
obtained from the half sib correlations for carcass measurements. With regard
to measurements on carcass fat, the h2-values for side and midback fat were the
lowest, and those for loin fat in Group B and for withers fat in Groups A+ B
were the highest. Calculated from the sire-progeny regressions, the h 2-values
for side fat were higher than from half sib correlations, while in Group B the
h2-value for midback fat was zero. The differences in the h2-values for the fat
thickness averages were relatively small, for example 0.42 from half sib correlation
and 0.44 from sire-progeny regression (Tables 30 and 31).

The h 2-values from the half sib correlations for side length and ham mea-
surement were quite high, while those from sire-progeny regressions were smaller
(Tables 30 and 31). Sizeable h 2-estimates were obtained from half sib correlations
for leg points (Table 30).

The h 2 estimates for the lean cuts results, carcass score and the cross-sectional
area of the longissimus muscle were all reliable. The highest heritabilities got
were for the meat+bone components (Table 31).

c. Phenotypic and genetic relations between characteristics

The phenotypic and genetic correlations between fat thickness (carcass)
and lean cuts results are summarised in Table 32. Withers fat has its closest
correlation with meat+bone, whereas loin and side fat are more closely corr-
elated genetically with fat+skin. The genetic correlations between midback
fat thickness and lean cuts results were not significant. As regards the fat
thickness averages, the genetic correlations of back fat x (1,2,3) and total fat x
(1,2,3,4) with lean cuts results were almost equally high. The phenotypic
correlation coefficients for carcass fat vs. lean cuts results were almost
without exception highly significant.



79

Table
30.

Heritability
estimates
(h
J
)

for
groupB

determined
by
(i)
half
sib

correlation
with
progenycharacteristics

and
(ii)

sire-progeny

regression,
using
sire

phenotype
data,
n
=

285.

Characteristics
of

progeny

Variance,
%

regression

Boars/Barrows
Sires

Error

Phenotype
evaluation

Age
at
88
kg,
days

4.62

1.85

93.53

1.19

0.08
±

0.09

0.04

Growth
per
day

(20-88
kg),
g

4.23

19.34

76.43

3.4o***
0.67
±

0.27

0.35

FU/kg
wt.

increase
(20-88
kg)

----
0.25

withers
(1)

36.52

2.39

61.10

1.36

0.15
±

0.13

-0.01

midback
(2)

49.10

0.48

50.42

1.09

0.04
±

0.07

-0.02

loin
(3)

41.30

3.85

54.84

1.65*

0.25
±

0.17

0.21

sol,
right
(4)

42.66

0.89

56.45

1.14

0.06
±

0.08

0.24

mean1,
2.
3

50.67

1.70

47.63

1.33

0.13
±

0.12

0.01

mean
2,4,
5

47.96

1.14

50.89

1.21

0.09
±

0.10

0.16

Test
points,

new
system

48.45

1.41

50.14

1.28*

0.11
±

0.11

0.21

Ham
measurement,
cm

1.25

13.27

85.48

2.4B***
0.47
±

0.23

0.20

Forelegs,
points

0.00

5.09

94.91

1.51*

0.19
±

0.15

-
Rear
legs,
points

1.14

5.73

93.13

1.59*

0.22
±

0.16

-
withers
(1)

36.49

5.01

58.49

I.Bl**
0.29
±

0.18

0.22

midback
(2)

46.53

4.40

49.07

1.85**
0.30
±

0.18

0.00

loin
(3b)

42.35

9.86

47.80

2.96***
0.58
±

0.26

0.23

side
(4)

39.42

3.40

57.18

1.56*

0.21
±

0.15

0.32

mean
1,2,
3

50.12

6.83

43.05

2.sl***
0.48
±

0.23

0.22

mean
1,2,
3,
4

49.96

5.72

44.38

2.22***
0.41
±

0.21

0.26

Side
length,

cm

11.52

20.56

67.92

3.BB***
0.75
+

0.29

0.37

Area
of

longissimus
muscle,

cm2

14.26

16.78

68.95

3.45***
0.65
±

0.27

The
half
sib

correlations
werecalculated

from
the
original
data,

and
the
sire-progeny

regressions
from
data
for
a

liveweight
of
88
kg.

*)

s.e.
=

standard
error.



Table
31.

Heritability
estimates
(h
2
)

determined
by
(i)
half
sib

correlation
with
growth,
carcassand

lean
cut

characteristics
of

progeny,

and
(ii)

sire-progeny
regression,

using
the
sire

phenotype
data,
n
=

427.
Half
sib

correlation

Sire-

Variance,
%

progeny

Characteristics
of

progeny„.
r
°

J

Sires

regression
h2+s

e

Group
A/

Boars/

F

h

2

„„„Sires

Error

Group
B

Barrows

Age
at
20
kg,
days

22.92

0.00

5.61

71.48

I.Bl**
0.27
±

0.14

-
Age
at
88
kg,
days

48.37

1.57

0.36

49.70

1.08

0.03
±

0.05

0.02

Growth
per
day

(20-88
kg),
g

32.53

2.96

8.98

55.53

2.6B***
0.49
±

0.19

0.32

FU/kg
wt.

increase
(20-88
kg)

______
0.18

Carcass
evaluation

Fat
thickness,

mm

withers
(1)

0.00

32.88

11.72

55.40

3.l9***
0.59
±

0.21

0.34

midback
(2)

0.00

46.82

3.63

49.55

1.76*

0.26
±

0.14

0.25

loin
(3b)

0.00

41.95

7.63

50.42

2.57***
0.46
±

0.19

0.36

side
(4)

0.00

39.34

2.28

58.38

1.41

0.15
±

0.10

0.39

mean
1,2,
3

0.00

49.05

7.64

43.32

2.B3***
0.52
±

0.20

0.39

mean
1,2,
3,
4

0.00

49.40

5.93

44.67

2.3B***
0.42
±

0.18

0.44

Side
length,

cm

0.00

13.23

20.69

66.07

4.l9***
0.77
±

0.24

0.30

Lean
cuts,

half
carcass

Meat
+

bone,
back,
g

21.40

12.79

15.57

50.24

4.22***
0.77
±

0.24

>

»

,

loin,
g

0.00

24.35

12.70

62.95

3.o9***
0.58
±

0.21

Main
part: meat

+bone,
g

0.00

29.51

15.64

54.85

3.96***
0.73
±

0.23

»

»

,

%

0.00

30.78

10.89

58.33

2.94***
0.54
±

0.20

fat
+

skin,
g

8.94

37.19

2.53

51.34

1.51*

0.18
±

0.12

»

»

,

%

16.83

37.59

3.57

42.01

I.BB**
0.29
±

0.15

fat
+

skin/meat
+bone

0.00

46.10

4.98

48.91

2.06**
0.34
±

0.16

meat
+

bone,
g/day

15.51

23.94

3.87

56.68

1.71*

0.24
±

0.13

Carcass
score

0.00

36.73

6.23

57.04

2.l3***
0.36
±

0.16

The
half
sib

correlations
werecalculated

from
the
original
data,

and
the

sire-progeny
regressions

from
data
for
a

liveweight
of
88
kg.

80



Table 32. Genetic (r g ) and phenotypic (rp ) correlations between fat thickness, measured ultra-
sonically on live animal and on carcass section, and lean cuts data.

Fat, carcass/lean cuts Fat, ultrasonic/
n = 427 lean cuts,

n = 285Characteristics

rg ± s.e. r P rP

Thickness of withers fat (1)
Meat + bone of back -0.31 ± 0.21 -0.22**» -0.12*
Meat ± bone of ham + loin —0.39 ± 0.21 —o.36*** —0.09
Main part of carcass:

meat + bone —0.42 ± 0.20 —o.36*** —O.ll
meat + bone, % of carcass —0.50 ± 0.19 —o.4s*** —0.14*
fat + skin 0.31 ± 0.31 o.49*** o.3o***
fat + skin. % of carcass 0.27 ± 0.28 o.49*** o.3o***
fat + skin/meat -f bone 0.44 ± 0.23 o.s4*** o.27***
meat + bone (g/day) —0.49 ± 0.25 -o.32*** -0.08

Carcass score —0.49 ± 0.22 —o.4l*** —0.12*
Thickness of midback fat (2)
Meat + bone of back 0.22 ± 0.28 -o.ll* —o.23***
Meat + bone of ham + loin 0.03 ± 0.31 —o.2s*** —o.l9***
Main part of carcass:

meat + bone 0.09 ± 0.29 -o.23*** -o.23***
meat + bone, % of carcass 0.18 ± 0.31 —o.34*** —o.3o***
fat + skin 0.29 ± 0.38 o.s3*** o.s2***
fat + skin, % of carcass 0.22 ± 0.35 o.so*** o,sl***
fat + skin/meat -f- bone 0.08 ± 0.35 o.so*** o.sl***
meat + bone (g/day) -0.30 ± 0.38 -o.24*** -o.27***

Carcass score -0.05 ± 0.36 -0.33»** -o.33***
Thickness of loin fat (3b)
Meat + bone of back -0.07 ± 0.25 -o.l9*** -o.23***
Meat + bone of ham + loin —0.28 ± 0,25 —o.3s*** —0.12*
Main part of carcass:

meat + bone -0.27 ± 0.24 -o.34*** -o.lB***
meat -f- bone, % of carcass —0.27 ± 0.25 —o.46*** —o.24***
fat + skin 0.78 ± 0.14 o.66*** o.36***
fat + skin, % of carcass 0.65 ± 0.18 o.64*** o.3s***
fat + skin/meat+ bone 0.58 ± 0.20 o.66*** o,34***
meat 4- bone (g/day) —0.71 ± 0.17 —o.34*** —o.lB***

Carcass score —0.61 ± 0.19 —o.46*** —o.22***
Thickness of side fat (4)
Meat + bone of back -0.17 ± 0.33 -o.27*** -o.3s***
Meat + bone of ham + loin —0.21 ± 0.34 —o.4l*** —o.36***
Main part of carcass:

meat + bone -0.23 ± 0.32 -o.39*** -o.37***
meat + bone. % of carcass —0.27 ± 0.34 —o.s3*** —o.44***
fat + skin 0.92 ± 0.07 o.7l*** o.s6***
fat + skin, % of carcass 0.71 ± 0.21 o.69*** o,sB***
fat -f- skin/meat + bone 0.63 ± 0.25 o.7l*** o.6o***
meat + bone (g/day) —0.57 ± 0.31 —o.4l*** —o.37***

Carcass score —0.44 ± 0.33 —o.s2*** —o.4s***
Average thickness of back fat (1, 2,3)
Meat + bone back —O.OB ± 0.24 -o.l9*** —o.22***
Meat + bone of ham + loin —0,25 ± 0.25 —o.36*** —o.ls**
Main part of carcass:

meat + bone —0.24 ± 0.23 —o.3s*** —o.2o***
meat + bone, % of carcass —0.26 ± 0.25 —o.4B*** —o.26***
fat + skin 0.48 ± 0.27 o.64*** o.46***
fat + skin, % of carcass 0.39 ± 0.26 o.62*** o.4s***
fat + skin/meat -f- bone 0.41 ± 0.25 o,64*** o.43***
meat + bone (g/day) —0.60 ± 0.22 —o.3s*** —o.2o***

Carcass score —0.46 ± 0.24 —o.46*** —o.26***
Average fat thickness (I, 2,3, 4)
Meat + bone of back —O.lO ± 0.25 —o.22*** —o,33***
Meat + bone of ham + loin —0.25 ± 0.26 —o.4o*** —o.33***
Main part of carcass

meat + bonemeat + bone —0.24 ± 0.24 —o.3B*** —o.3s***
meat + bone, % of carcass —0.27 ± 0.26 —o.s2*** —o.42***
fat + skin 0.57 ± 0.25 o.7o*** o.sB***
fat -f skin, % of carcass 0.46 ± 0.26 o.6B*** o.s9***
fat + skin/meat -f- bone 0.46 ± 0.25 o.7l*** o.6o***
meat + bone (g/day) —0.61 ± 0.23 —o.39*** —o.36***

Carcass score —0.47 ± 0.25 —o.sl*** —o.43***
Correlations calculated from the original data.

* P < 0.05, ** P < 0.01, *** P < 0.001



82

For calculating genetic correlations, the amount of data provided by the
progeny of group B (n = 285) is rather small. The rg for ultrasonically measured
fat thickness vs. lean cuts results is in fact very low. The highest rg values were
for side and loin fat vs. meat+bone of back ( 0.48+ 0.39 and —0.31 +0.34
respectively). The rp ’s for ultrasonically measured fat vs. lean cuts were highly
significant in almost every case (Table 32), the withers fat vs. mean-f-bone rp

being the lowest. Evidently the thickness of the withers fat is a very good
measure of the meatiness of the carcass, but the ultrasonic measurement of
this fat is inaccurate.

The genetic and phenotypic correlations between the cross-sectional area of
the longissimus muscle and characters reflecting the meatiness or meat yield
of the carcass were very high (Table 33). Side length (carcass) correlated well
with meat-)-bone of back, and meat+bone (weight and %) of the main part of
the carcass. No genetic correlation was found between ham measurement and
lean cuts results, nor between rate of growth and carcass characteristics. The
r p for feed efficiency vs. growth per day (20—88 kg) of all progeny (Groups
A+ B) was —o.3s***. It was not possible to determine the genetic correlation
between these characteristics because the progeny were group-fed.

Table 33. Genetic (rg ) and phenotypic (rp ) correlations of cross-sectional area of longissimus
muscle and side length of carcass with lean cut data, n = 427.

Correlations
Characteristics

r g ± s.e. r P

Longissimus area
Meat + bone of back 0.79 ± 0.09 o.62***
Meat + bone of ham + loin 1.00 ± 0.00 o.69***
Main part of carcass:

meat -f- bone 0.89 i 0.05 o.6B***
meat -f- bone, % of carcass 0.81 ± 0.09 o.sB***
meat -f- bone (g/day) 1.00 ± 0.00 o.s2***

Carcass score 0.91 ± 0.05 o.s9***

Side length
Meat + bone of back 0.69 + 0.12 o.49***
Meat -)- bone of ham + loin 0.20 ± 0.23 0.12*
Main part of carcass:

meat -f- bone 0.41 ± 0.19 o.29***
meat + bone, % of carcass 0.53 ± 018 o.2s***
meat ■+• bone (g/day)

, 0.15 ± 0.30 0.13**
Carcass score 0.23 + 0-26 0.14**

Correlations calculated from original data
* P < 0.05. »� P < 0.01, �** P < 0.001



83

5. Investigations on the improvement of station testing of hoars

a. Phenotype and carcass characteristics as indicators of meatiness of carcass

The effectiveness of phenotype evaluation in predicting carcass characteri-
sties was examined by comparing the evaluation results for the Group B progeny
with measurements made on the carcass and with the lean cuts results.

The correlations between ultrasonic and carcass fat measurements were

as close as those reported elsewhere (Table 34, see Table 1). The closest carcass-
ultrasonic correlation was for side fat (boars r = o.B3***, barrows o.77***);
almost as close were the correlations for average side+midback fat x (2,4),
and for the overall average x (1,2,3,4) (Table 34).

The loin fat measurements on the carcass correlated more closely with
ultrasonic measurements other than of loin fat. The carcass-ultrasonic corre-
relation for withers fat (r = o.69*** and o.s3***) was higher than was to be
expected from the correlation between the sires’ fat thickness measurements
and the progeny withers fat measured ultrasonically or on the carcass (see
Tables 28, 29 and 34). If the results of the phenotype evaluation are converted
to test points, variation in lean cuts results can be accounted for as follows:

Independent variables
Old test New test

Dependent variables points points
Ä2o/o Ii 2%

Meat-|-t>one of back, g 9.9 17.2
Meat 4-bone of ham-|-loin, g 9.9 21.1
Most valuable parts of the carcass

meatbone, g 12.1 24.4
» * %of carcass 10.2 20.9

fat-fskin, g 24.8 33.8
» * %of carcass 26.4 36.1

fat+skin/meat+ bone 25.139.2
meat-f bone, g/day 24,1 26.9

Carcass score 20.7 30.4

The new test points accounted better for the variation in lean cuts results
than did the old points. The following variables were significant in accounting
for the test points of the progeny:

Old test points
Step Independent variable Boars Barrows

T value 7f 2% T value /i 2%

1 Average fat thickness x (1.2,3) —3o.32*** 76.2 —63.3l*** 84.3
2 Age at 88 kg -11.85*»* 11.6 -20.07*** 8.5

Total 89.4 96.7

New test points
1 Average fat thickness x (2,4,5) —120.56*** 91.7 —5B 42*** 88.1
2 Age at 88 kg —2s.Bs*** 4.2 —lo.6l*** 2.9
3 Daily gain 2.09* 0.1

Total 99.2 95.9



84

Table
34.

Progeny
(<J
boars
n
=

134,
&

barrows
n
=

165)
of

group
B
boars:
correlation
of
fat

thickness
as
determined

by
ultrasonic

measurement
with
fat
thickness
and

longissimus
area

(LMA)
as

measured
on

carcasssection.

\
Carcass

Loin
(3)

Averages

measurements
Withers
Midback

.

Side

front
middle
back

LMA

(!)

(2)

(a)

(b)

(c)

(4)

1,2,3
2,4

1,2,3,4

Withers
(1)

<J

+0.69
+0.50
+0.46
+0.47
+0.43
+0.50
+0.63
+0.54
+0.62
-0.21

&

+0.53
+0.34
+0.26
+0.20
+0.27
+0.30
+0.45
+0.35
+0.43
-0.14

Midback
(2)

$

+0.49
+0.69
+0.54
+0.58
+0.56
+0.60
+0.66
+0.68
+0.67

-0.04

J-

+0.49
+0.68
+0.53
+0.53
+0.47
+0.62
+0.67
+0.71
+0.70
-0.27

Loin
(3)

$

+0.42
+0.52
+0.45
+0.51
+0.45
+0.53
+0.54
+0.56
+0.56
-0.15

&

+0.29
+0.42
+0.31
+0.42
+0.36
+0.39
+0.43
+0.44
+0.45
-0.17

Sol,
right
(4)

$

+0.57
+0.63
+0.60
+0.60
+0.57
+0.83
+0.68
+O.BO
+0.76

-0.31

&

+0.48
+0.57
+0.54
+0.59
+0.53
+0.77
+0.64
+0.75
+0.73
-0.44

Mean
(1,2,3)

$

+0.63
+0.66
+0.56
+0.60
+0.55
+0.63
+0.71
+0.69
+0.72
-0.16

&

+0.53
+0.57
+0.43
+0.45
+0.43
+0.51
+0.61
+0.59
+0.62
-0.23

Mean
(2,
4,
5)
1
)

$

+0.57
+0.70
+0.63
+0.63
+0.61
+O.Bl
+0.72
+O.Bl
+0.78
-0.23

&

+0.52
+0.64
+0.56
+0.61
+0.55
+0.77
+0.70
+0.78
+0.77
-0.42

4
+

5

Mean
(1,
2,3,
—j-)

tj

+0.65
+0.70
+0.61
+0.64
+0.60
+0.73
+0.75
+0.77
+0.78
-0.21

&

+0.55
+0.60
+0.48
+0.52
+0.49
+0.62
+0.66
+0.67
+0.70
-0.31

P
<

0.05,
n
=

134
r
>

0.18,
n
=

165
r
>

0.12

P
<

0.01,

»>

r
>

0.24

•

r
>

0.15

P
<

0.001,
l

r
>

0.30

»

r
>

0.19

The
basis
for
the

adjustment
of
all

measurements
was

the
averageof

the
weight
at
ultrasonic

measurementand
the

carcass
weight.

J
)

5

=

fat
thickness,
sol,
left.



85

The new test points depended more closely on the thickness of the fat than did
the old test points. Thus the importance of the new test points in accounting
for variation in carcass value is similar to that of the fat thickness values.

When all the phenotypic characteristics evaluated, i.e. ultrasonic measure-
ments of fat and their averages, daily gain, age at 88 kg and ham measurement,
were used as independent variables in stepwise multiple regression analysis, the
phenotype evaluation accounted for 25.5—45.5 % of the total variation in lean
cuts results, 72.1 % in meat+bone produced per day and 51.9 % in carcass
score (Table 35).

Either the sol fat or the average of the sol measurements and the fat of the
midback proved to be the most important of the independent variables. The
other fat thickness values correlated so closely with these fat measurements
that they were unable to give any additional information on the variation
in carcass quality. Ham measurement and, for most traits, daily growth values
gave the most additional information. The age at 88 kg was an important in-
dependent variable in explaining the variations in daily yield of meat+bone
and carcass score. This is natural, as this growth craracteristic was used in
the calculation of yield and carcass scores (see p. 67).

Of the carcass characteristics the cross-sectional area of the longissimus
muscle provided the best estimate of the weight of meat+bone. The second
best estimator was fat thickness or length of side. The usefulness of the various
fat thickness measurements depended on the carcass characteristic in question.
Loin and side fat usually gave additional independent information on the
variation in the meat+bone component. The fat thickness averages, on the
other hand, accounted significantly for the variation in the fat + skin component.
Carcass characteristics and rate of growth together accounted for 47.7 65.6 %

of the overall variation in the lean cuts results, for 78.6 % of that in the daily
yield of meat+bone, and for 65.0 % of that in the carcass score. The results
are some 10—20 % better than those obtained by phenotype evaluation
(Table 35).

The daily gain gave negative T values for the yield of meat + bone and fat +
skin, except for meat+bone of ham+loin, where the T value was positive
(Table 35). There were apparently two kinds of slow-growing progeny: those
low in fat and with good meatiness, and those high in fat and with little meat.
A high rate of growth resulted in an increase in the carcass score. Thus daily
gain has, independently of the value of the carcass, been selected for.

Using data for the Group B progeny, it was checked if the live animal mea-
surements used in predicting carcass traits were confounded with breed diffe-
rences; no systematic differences were found. The Landrace minus Yorkshire
figure for the ham measurement was + 1.4 cm*, but the meat+bone of ham
—25 gNS and the fat+skin amount +93 g***. Yet the ham measurement
correlated positively with the meat+bone and negatively with the fat+skin
components (Table 35). Side and loin fat were thicker in the Yorkshire progeny
than in the Landrace progeny when measured ultrasonically, but in carcass
evaluation, where loin fat was measured in three locations, the result was the
reverse. The differences between the breeds were not significant in respect
of total weight of fat+skin or cross-sectional area of the longissimus muscle.



86

Table
35.

Reliability
of

phenotype
testing
and

carcassevaluation
in

predicting
lean
cut

results:
statistical

analysis
of
data,
n
=

292.

Phenotype
evaluation

Carcass
evaluation

Dependent
Step

Independent
variables
r
p

T
value

R2%Step
Independent

variables
r„

T
value

R2%
Meat
+
bone
of

1.
Mean
fat
(2,
4,5)

-0.44
-B.73***
20.1

1.
LMA

+0.54
+lo.sB***

19.5

back,

2.

Ham
measurement
+0.22
+

3.Bl***
3.8

2.
Side
length

+0.43
+

9.75***
16.5

3.

Growth/day
(20-88
kg).
-0.11
-2.62**
1.8

3.

Fat,
loin
(3b)

-0.36
-3.22**
1.8

Total

25.5

4.

Growth/day
(20-88
kg).
-0.11
-2.35*
1.0

Total

50.6

Meat
+

bone
of

1.

Fat,
sol
(4)

-0.48
-9.76***
20.5

1.

LMA

+0.59
+7.lo***
9.3

ham+loin
2.

Ham
measurement
+0.40
+

B.46***
15.4

2.

Fat,
side
(4)

-0.59
5.54***
5.6

3.

Growth/day
(20-88
kg).
+0.12
+2.30*
1.1

3.

Fat,
withers

-0.47
—3.27**
2.0

Total

39.1

Total

47.7

Meat
+bone
of

1.

Fat,
sol
(4)

-0.53
-11.06***
26.7

1.

LMA

+0.60
+7.66***
9.9

main
part
of

2.
Ham

measurement
+0.34
+

6.9l***
10.4

2.

Fat,
side
(4)

-0.61
4.l4***
2.9

carcass

Total

38.1

3.

Fat,
loin
(3b)

-0.53

-
3.17**
1.7

4.
Side
length

+0.20
+

2.88**
1.4

5.

Growth/day
(20-88
kg).
-0.02
-2.39*

1.0

Total

52.0

Meat+bone
of

1.

Fat,
sol
(4)

-0.48-9.66***
22.7

1.

LMA

+0.59
+

7.43***
9.9

main
part
of

2.

Ham
measurement

+0.29
+

5.55***
7.5

2.

Fat,
side
(4)

-0.60

-
4.o4***
2.9

carcass,%

of

Total

30.9

3.

Fat,
loin

-0.51
-2.77**
1.4



87

Fat
+

skin
of

1.

Mean
fat,
(2,
4,
5)

+0.58
+ll.9o***
30.8

1.

Mean
fat
(1,
2,3,
4)

+0.72
+l3.s4***
27.6

main
part
of

2.

Growth/day
(20-88
kg).
-0.25
-4.23***
3.9

2.

LMA

-0.49
-4.B3***
3.5

carcass

3.

Ham
measurement
-0.08
2.10*
1.0

3.

Growth/day
(20-
88
kg).

-0.24
3.30**
1.6

Fat
+skin
of

1.

Mean
fat,
(2,
4,
5)

+0.60
+l2.s9***
32.9

1.
Mean
fat,
(1,
2,
3,4)

....

+0.73
+l3.93***
28.4

main
part
of

2.

Growth/day
(20-88
kg).
-0.25
-4.37***
4.0

2.

LMA

-0.50
-4.BB***
3.5

carcass,%

of

3.

Ham
measurement
-0.10-2.70**
1.5

3.

Growth/day
(20-88
kg).
-0.25

-
3.35***
1.6

Fat
+

skin/

1.

Mean
fat
(2,
4,5)

+0.63
+l3.9l***
37.3

1.
Mean
fat
(2,
3,4)

+0.76
+ls.ss***
29.1

meat
+bone
of

2.
Ham
measurement
-0.19
-4.66***
4.2

2.
LMA

-0.60
-B.2l***
8.1

main
part
of

3.

Growth/day
(20-88
kg).
-0.19
-3.13**
1.9

Total

65.6

Meat
+

bone
of

1.

Age
at
88
kg

-0.75
-20.08***
39.9

1.

Age
at
88
kg

-0.75
-22.56***
38.3

main
part
of

2.

Fat,
sol
(4)

-0.40
-10.35***

10.6

2.
Mean
fat
(2,
3,4)

-0.48
-B.B9***
5.9

carcass(g/day)

3.

Ham
measurement

+0.19
+6.44***
4.1

3.
LMA

+0.47
+B.66***
5.6

14.
Growth/day

(20-88
kg).
+0.38
-2.00*
0.4

4.

Growth/day
(20

-
88
kg).
+0.39
-3.52***
0.9

s
score
1.

Fat,
sol
(4)

-0.50
-10.36***
18.3

1.

Age
at
88
kg

-0.49

-
B.27***
8.5

3.
Ham

measurement
+0.23
+

6.ls***
6.5

3.

Fat,
side
(4)

-0.62
4.B7***
2.9

4.

Growth/day
(20-88
kg).
+0.39
+2.9l**
1.4

4.

Fat,
loin
(3b)

-0.55
-3.26**
1.3

*

P
<

0.05,
**

P
<

0.01,
***

P
<

0.001.

Data
for

regression
analysis

was
corrected
for
weight
at

ultrasonic
measurementand

carcass
weight;
in
pooling
the
data
the

differences

between
boars
and
barrows

weretaken
into
account.

LMA
=

longissimus
muscle

area.



88

The side length was significantly greater (+ 2.0 cm ***) in the Landrace progeny
than in the Yorkshire progeny; here it should be recalled that there was no
significant difference between the breeds in this respect in the test boars
(p. 52).

b. Potential for early selection of boars

Early testing of the progeny was made in order to investigate the advantages
of early culling based on rate of growth. Breeding for rate of growth and ob-
viously also for feed efficiency could be improved in this way. The final
evaluation, at the end of the test period, could then be based on carcass cha-
racteristics alone.

The early testing was made 8 weeks after the commencement of the test
period. Ultrasonic measurement of the fat, leg assessment and measurement
of the ham were made on the progeny of Group B. The average weight of the
progeny at this stage was 58.7 kg (boars 59.6), and their average age was 122
days (Table 18). For each animal the rate of growth corresponding to this
average weight was calculated from its age at 60 kg and its rate of growth
between 20 and 60 kg liveweight.

The pre-test environment of the piglet had a marked effect on the rate
of growth during the period up to the early test (Table 17). However the results
of the early test were not influenced by the size and the weight of the litter at
3 weeks. The weight of the pig at three weeks correlated significantly with the
weight at the beginning of the test period, but not at the time of the early test.
Variations in initial weight and age did not affect the fat thickness of the early
test (Table 19).

The weight at ultrasonic measurement was significantly related to the fat
thickness at the early test; age at ultrasonic measurement was less closely
related (Table 19). Differences in rate of growth and thickness of fat in the early
test were highly dependent on differences between the sires (Tables 17 and 19).
The influence of sire on fat thickness was more pronounced in the early test than
in the regular test (Table 19). The h2-estimates for daily gain of the boars’ in
the liveweight ranges 20—60 and 20—70 kg were reliable and quite high (Table
36). The h 2-estimate for the daily gain of boar-progeny for the entire test period
(20—88 kg) was unexpectedly high (Table 36). The h2-estimates for the rate of
growth of the barrows were considerably lower than those for the boars.
Variations in growing conditions during the piglet stage had such a great effect
upon the age at 60 kg that h 2 was negligible.

The h2-estimates for the fat thickness were also higher for the boars than
for the barrows, and actually higher than the h2-estimates for fat thickness of
all progeny (A +B) in the regular test (Tables 30 and 36). It was surprising,
however, that the h 2-estimate for midback fat was higher and more reliable
in the preliminary test than in the regular test, though the fat thickness in
question (average 11.5 mm) was the smallest measured in the whole material
(Table 18). The h 2-estimates for leg points in the early test were also higher



89

Table 36. Heritability estimates of growth characteristics at various weight intervals and
early-stage phenotype characteristics (measured at a live-weight of about 60 kg) determined
by half sib correlations for the progeny of Group B boars.

Boars, n = 133 Barrows, n = 165
Characteristics h 2 ± s.e. h 2 ± s.e.

Age/60 kg, days 0.06 ± 0.12 —0.05 ± 0.00
Growth 20—60 kg (g/day) 0.67 ± 0.403) 0.12 ± 0.16

20-70 » * 0.73 ± 0.423) 0.20 ± 0,21
20-88 » » 0.93 ± 0.473) 0.34 ± 0.271 )

60-88 » » 0.59 ± 0.382) 0.18 ± 0.19
Fat thickness,

withers (1) 0.20 ± 0.22 0.06 ± 0.11
midback (2) 0.37 ± 0.30 1) 0.18 ± 0.19
loin (3) 0.31 ± 0.27 1) 0.09 ± 0.13
sol (4) 0.06 ± 0.12 0.19 ± 0.19
mean of 1,2, 3 , 0.26 ± 0.25 0.13 ± 0.16
mean of 2,4, 5 0.29 ± 0.26 0.17 ± 0.18

Ham measurement, cm —0.04 ± 0.00 0.27 ± 0.23 1)
Forelegs 0.58 ± 0.382) 0.44 ± 0.30 2)

Rearlegs 0.80 ± 0.45 3) 0.58 ± 0.34 3 )

Uncorrected data were used in the calculations
Levels of significance of between-sire differences;
‘) P< 0.05, 2) P < 0.01, 3) P < 0.001.

and more realible than those obtained in the final test. The heritability for
the ham measurement was reliable only for the barrows.

The genetic and phenotypic correlations between the early rate of growth
and that during the entire test period were very close (Table 37), due partly
of course to autocorrelation. The early growth per day (20—60 kg), however,
correlated very closely with the rate of growth during the later period (60
88 kg).

With regard to fat thickness, the results of the early test correlated genetically
very closely with those of the final test for withers, loin and average backfat
but very weakly for midback and sol (Table 38). This order was reversed in
the phenotypic correlations, but the differences were smaller. In almost every
case the rp was smaller than the rg . Naturally, there is autocorrelation between
the various fat thicknesses.

Sol fat thickness and ham measurement in the early test were important
in predicting a number of lean cuts results, though the latter was significant only
in respect of meat+bone (Table 39). It should be noted that sol fat and ham
measurement were greatly influenced at this early stage by environmental
factors, as theirh2

- estimates in the boars were extremely low (Table 36). Also
the age at 60 kg, which accounted significantly for variation in daily yield
of meat+bone and carcass score, had a low h2-value (Table 36).

According to the results described above, early selection of boars at a weight
of 60—70 kg, based on their daily growth at a testing station, is reliable, provided



90

Table
37.

Phenotypic
(r

p)

and
genetic
(r

g
)

coefficients
of

correlation
between
early
rate
of
growth

and
that
(a)
for
the
whole
test

period
(b)
during
the
later
stages
of
the
test

period,
n
=

297.

\Age
at
88
kg

Growth
per
day

Growth,
88
kg

,

20-88
kg

60-88
kg

r
p

r
g

±

s.e.

r
p

r
g

±

s.e.

rp

r
g

±

s.e.

Age
at
60
kg

0.89

0.37
±

1.30

-0.30
-0.40
±

0.73

-0.10
-0.57
±

0.69

Daily
wt.
increase

(20-60
kg)

-0.41
-1.01
±

0.01

0.82

1.00
±

0.00

0.22

0.96
±

0.03

Daily
wt.

increase
(20-70
kg)

-0.48
-1.08
±

0.09

0.91

0.99
±

0.01

0.45

0.91
±

0.06

Significance
of

correlation
coefficients:
P
<

0.05,
r
p

>

0.12;
P
<

0.01,
r
p

>

0.15;
P
<

0.001,
r
p

>

0.19.

Table
38.

Phenotypic
(r

p)
and

genetic
(r_)

correlations
between
fat
thickness

measured
ultrasonically
at

liveweights
of
about
60
kg
and

88
kg;
Group
B

progeny,n
=

297.

\
Normal ultrasonic

Withers
fat
(1)

Midback
fat
(2)

Loin
fat
(3)

Sol
fat
(4)

Mean
fat
(1,
2,3)

measure- ment

,

,

,

,

,

rD

re
±

se-
rD

r<r
±

s
-

e
-

rD

rz
±

se-
rn

Tz
±

s.e.
r
D

r.
±

s.e.

(88
kg)

p

g

p

g

p

g

p

g

p

g

Fat
thickness withers

(1)

0.29
1.47
±

0.76
0.28
0.18
±

0.89
0.29
0.90
±

0.11
0.22
-0.11
±

0.84
0.34

1.17
±

0.25

midback
(2)

0.26
1.17
±

0.21
0.34
0.87
±

0.19
0.28
0.64
±

0.31
0.30
0.50
±

0.55
0.35
1.03
±

0.03

loin
(3)

0.22
1.12
±

0.15
0.29
0.24
±

0.82
0.25
0.26
±

0.54
0.26
0.16
±

0.79
0.30
0.70
±

0.33

sol
(4)

0.25
1.23
±

0.34
0.41
0.49
±

0.69
0.33
0.99
±

0.01
0.40
0.80
±

0.30
0.38

1.16
±

0.23

mean
(1,
2,3)

0.31
1.53
±

0.82
0.35
0.59
±

0.55
0.33

0.77
±

0.23
0.30

0.39
±

0.67
0.39
1.21
±

0.29

Significance
of
correlation
coefficients:
P
<

0.05,
r
p

>

0.12;
P
<

0.01,
r
p

>

0.15;
P
<

0.001,
rp

>

0.19.

Analysis
for
on

uncorrected
data.



91

Table 39. Reliability of early-stage (60 kg liveweight) phenotype testing in predicting lean cut
results, n = 287.

Dependent variables Independent variables r p T value R 2 %
cn

Meat + bone of back 1. Fat, sol —0.36 7.oo*** 14.5
2. Ham measurement +O.lB + 4.ll*** 5.0

Total 17.6

Meat + bone of ham + 1. Fat, sol —0.28 5,53*** 9.5
loin 2. Ham measurement +0.19 + 4.o2*** 5.0

Total 13.0

Meat + bone of main 1. Fat, sol —0.40 B.os*** 18.1
part of carcass 2. Ham measurement +0.21 + 4.9o*** 6.7

Total 22.3

Meat + bone of main 1. Fat, sol —0.35 7.lB*** 14.8
part, %of carcass 2. Ham measurement +0.23 + s.ls*** 7.6

Total 20.1

Fat + skin of main 1. Fat, sol +0.43 + B.lo*** 18,3
part of carcass 2. Growth/day (20 —6O kg) —0.20 3.52*** 3.4

Total 22.4

Fat + skin of main 1. Fat, sol +0.46 + B.s9*** 20.1
part, % of carcass 2. Growth/day (20 —6O kg) —0.19 3.36*** 3.1

Total 23.9

Fat + skin/meat+bone 1. Fat, sol +0.48 + 9.67*** 24.7
2. Ham measurement —0.12 3.55*** 3.3

Total 26.2

Meat + bone of main 1. Age at 60 kg —0.71 —lB.oB*** 49.4
part of carcass (g/day) 2. Fat, sol —0.25 6.6B*** 6.7

3. Ham measurement +0.12 + 3.66*** 2.0
Total 58.0

Carcass score 1. Age at 60 kg —0.41 B.22*** 16.4
2. Fat, sol -0.34 - 7.22*** 12.6
3. Ham measurement +O.lB + 4.52*** 4.9

Total 32.7

Data for regression analysis was corrected for weight at ultrasonic measurement and weight
of carcass; differences between boars and barrows were corrected for in pooling the data

* P < 0.05, •* P < 0,01, *** P < 0.001



92

that the effects of initial weight and age are eliminated. Animals with weak
legs could be removed at the same time. Possibly, carcass value could be
predicted closely enough by ultrasonic measurements of fat thickness so that
supplementary information for use in borderline cases would be got.

c. Leg strength of boars on abundant feeding

The feeding standards of the progeny of Group B (Table 9, 1969 71) were
approximately according to appetite, especially up to a liveweight of 60 kg.
They exceeded the feeding standards of the sires by 0.3—0.4 FU/pig/day
(Table 7, 1968—69). For this reason and because of the improved feed quality
the growth of the progeny was a great deal faster than that of the sires. The
average daily gain of the B sires was 604 g whereas that of the B progeny boars
was 761 g and that of the progeny barrows 735 g (Tables 14 and 16).

The dung channel in the stalls was covered with a grid, which caused bruises
and swelling in the joints of the legs. None of the progeny had to be slaughtered
prior to schedule on account of leg faults. Only five of the pigs (1 boar and 4
barrows) out of a total of 298 scored 2 for leg points; i.e. they had to be assisted,
for instance at weighing. At the early test about 20 % of the pigs had slight
walking impediments (3 leg points), the figure being about 38 % at the time of
the regular test. Most of the impediments were hoof defects or severe joint
swelling. The faults were more severe in the rear legs than in the forelegs. Most
of the pigs scored 4 leg points; there were visible defects (slight swelling of rear
leg joint, or non-straightening of foreleg) but there was no difficulty in walking.
At the time of the regular test 32 % of all recorded defects were swollen joints,
22 % were hoof defects and 16 % were crookedness. About 20 % had faultless
legs (5 points) (Table 40).

The points for rear legs obtained by the progeny-boars and progeny-barrows
in the regular test, grouped according to sire, are given in Table 40, which is a
sire-progeny comparison for leg points. It can be seen that the average leg points
of the sires in the three sub-groups were identical, while the legs of the progeny
of group 111 were poorer than those in I and II; the differences, however, were
not significant. 11l contained very few progeny with 5 leg points. In I and II
the pigs were distributed very evenly into groups with 3, 4 and 5 leg points.
The legs of the progeny boars scored higher than those of the barrows, but the
difference was not significant.

Boar »Kilperi» had progeny with exceptionally good legs; 70 % of its progeny
boars had faultless rear legs (Table 40), and the »Kilperi» progeny differed sig-
nificantly from three of the other lots of progeny. It was seen earlier (Tables
30 and 36) that the h2-estimates for leg points were reliable. The error variance
for leg points (regular test) was greater than that in the preliminary test, due
possibly to the greater number of leg injuries occuring over the longer period.

Leg defects caused by rapid growth and associated with bone weakness
could not be ascertained with statistical reliability, but the correlation between
daily gain and points for rear legs was negative for the boars (r p

= —0.16).
The corresponding rp for all progeny was almost zero, as were the correlations
between meatiness and leg points. The r p ’s for the boars’ leg points vs. the



93

various yields of meat found by lean cuts analysis varied from —0.03 to -j-0.17;
for example, the r p for points for rear legs vs. meat +bone of the ham was 0.00.

It can be seen from the above that one-third of the progeny boars were
retained for the duration of the testing period despite their impediments to
walking (Table 40). One-fifth of the pigs had no leg defects. Hereditary leg
weakness and injury were obviously the main factors responsible for the low leg
points, rather than the rapid growth or heaviness of the animals.

Table 40. Leg points 1) of sires and rear-leg points of the progeny in the regular testing (88 kg).
Sires, n = 15. Progeny; boars (S) n = 133, barrows (S') n = 165.

Leg points Frequency distribution of progeny leg points, %

Averages
, 2 points 3 points 4 points 5 pointsfor progeny r r r r

Sires2)

Sires

S S S S' S S' S S S S

Sub-group I.
1. Rukki 4.0 3.7 3.9 - - 28.6 50.0 71.4 12.5 - 37.5
2. Riku 3.5 4.0 3.4 - - 33.3 60.0 33.3 40.0 33.3
3. Naked 3.5 4.3 3.7 - - 27.3 45.5 18.2 36.3 54.5 18.2
4. Reka 4.0 4.4 4.1 - 57.1 20.0 14.3 60.0 28.6 20.0
5. Hymyri 4.0 4.2 4.3 - - 10.0 11.1 60.0 55.6 30.0 33.3

Average 3.8 4.1 3.9 29.5 37.5 38.6 41.7 31.8 20.8

Sub-group 11.
6. Rape 4.0 3.4 3.6 - 12.5 45.5 37.5 54.5 37.5 - 12.5
7. Kilperi 4.0 4.5a 4.0 - 20.0 33.3 10.0 33.3 70.0 33.3
8. Ripa 4.0 3.8 b 3.4 - - 55.6 60.0 44.4 40.0
9. Ranu 3.0 4.0 3.8 - 40.0 50.0 20.0 25.0 40.0 25.0

10. Ruuppo 4.0 4.3 3.6 - 5.9 14.3 41,2 42.8 41,2 42.9 11.7
Average 3.8 4.0 3.7 - 3.4 36.2 44.1 34.0 35.6 29.8 16.9

Sub-group 111.
11. Rata 4.0 3.9 3.9 - - 22.2 25.0 66.7 58.2 11.1 16.7
12. Sapro 3.5 3.8 3.6 - - 20.0 37.5 80.0 62.5
13. Rooli 4.0 3.4 b 3.6 - - 66.7 50.0 22.2 41.6 11.1 8.3
14. Sapeli 3.5 3.6 3.8 - 8.3 40.0 33.3 60.0 25,0 - 33.3
15. Riski 4.0 3,5b 3.4 14.3 - 14.3 63.6 71.4 36.3

Average 3.8 3.6 3.7 2.2 1.8 33.3 41.8 60.0 43.6 4.4 12.7

Overall average .... 3.8 3.9 3.7 0.7 1.9 33.1 41.4 44.1 40.1 22.1 16.7

*) Details of scoring:
1 = incapable of walking
2 = incapable of walking unaided
3 = visible defects impeding walking
4 = visible defects not impeding walking
5 = no visible defects

a) Forelegs and rearlegs were judged together
a-b; P < 0.05



94

Table
41.

Prediction
of
boars'
lean
cut

results
from
a)

phenotype
testing
of
the
boars
and,
b)
the

results
of
the

phenotype
testing
of
the

boars
and

the
lean
cut

results
from
sibling

barrows
(<£).

D

,

a)
Boars
1
)

(n
=

122)

b)
Boars
and
barrows
1
)

(n
=

122
and

122)

variables
Step

Independent
variables
r
p

T
value

R2%Step

Independent
variables
r
p

T
value

R2%
Meat
+

bone,
back
1.

Fat,
sol

-0.40-s.o7***
16.5

1.

Fat,
loin

-0.28

-
3.3B***
6.6

2.

Growth/day
(20-88
kg).
-0.15-3.12**
6.2

2.

&

meat+borie,
%

of

3.

Age
at
88
kg

-0.08
-2.33*

3.5

carcass

+0.47
+

2.28*

3.0

4.
Ham

measurement
+0.19
+

2.24*
3.2

3.

£

meat
+

bone,
midback
+0.47
+

2.26*

3.0

Total

26.3

Total

31.6

Meat
+

bone,

1.

Fat,
sol

-0.50-6.Bo***
23.5

1.

Fat,
sol

-0.43

-
6.lo***

18.4

ham
+
loin

2.
Ham
measurement
+0.35
+

s.4o***
14.8

2.
Ham
measurement
+0.37
+

4.22***
8.8

3.

Growth/day
(20-88
kg).
+0.17
+

2.34*

2.8

3.

+

bone.ham
+

loin
+0.40
+

3.75***
7.0

Total

41.2

Total

41.5

Meat
+

bone,
main
1.

Fat,
sol

-0.54-7.sl***
29.6

1.

£

meat
+

bone,
shoulder
+0.43
+4.47***
9.8

part
of

carcass
2.
Ham
measurement
+0.30
+

4.29***
9.7

2.
Ham
measurement

+0.34
+

3.99***
7.8

Total

38.6

3.

Fat,
sol

-0.43

-
3.4l***
5.7

4.

Fat,
loin

-0.32-
2.23*
2.4

Total

42.8

yj
=

meat
+

bone,
2
)

1.
xt
«■

fat.
sol

-0.44
-s.sl***
19.9

1.

Fat,
sol

-0.44
-3.09**
5.1

%

of
carcass
2.

x2=ham
measurement.
+O.lB
+

2.37*
3.7

2.

#

meat
+

bone,
shoulder
+0.42
+

2.31*
2.8

Total

23.2

3.

£

meat
+

bone,%
carcass
+0.43
+2.10*
2.3

4.

Fat,
loin

-0.33-
2.01*

2.2

Total

37.5

Fat
+

skin

1.
Mean
fat
(1,
2,4)

+0.68
+lo.o7***
42.8

1.
Mean
fat
(1,
2,4)

+0.68
+9.7B***
36.3

2.

Growth/day
(20-88
kg).
-0.28
-3.4l***
4.9

2.

Growth/day
(20-
88
kg).
-0.27
-3.Bs***
5.6

Total

50.6

3.

£

fat+skin,
shoulder...
+0.32
+3.27**
4.0



y 2=

fat
+

skin,
3
)

1.

x 3=

meanfat
(1,
2,4).+0.66

+9.Bo***
41.2

1.

Mean
fat
(1,
2,4)

+0.66
+9.29***

35.8

%

of
carcass
2.

x4=growth/day

2.

Growth/day
(20-88
kg).

—0.27
3.08**
3.9

1(20-88
kg)

-0.28
-3.67***
5.8

3.

&

fat
+

skin,
%

of

3.
x2=ham

measurement.

—0.07
2.07*

1.8

carcass

+0.32
+

2.76**
3.2

Fat+skin/
1.

Fat,
sol

+0.70
+

5.96***
13.5

1.

Fat,
sol

+0.70
+

7.54***
20.2

meat+bone
2.

Mean
fat
(1,
2,3)

+0.60
+

2.81**
3.0

2.

&

meat
+

bone,
ham

....

—0.23
2.63**
2.5

3.

AgeatBBkg

+0.23
+2.79**
3.0

3.

Growth/day
(20-
88
kg).
-0.21
-2.62**
2.4

4.
Ham

measurement
—O.ll
2.53*

2.4

4.

Fat,
withers

+0.51
+

2.50**
2.2

y,
=

meat
+

bone*)
1.

x5=age
at
88
kg

-0.80
-17.32***
60.6

1.

Age
at
88
kg

-0.79
-18.02***
57.8

g/day

2.

Xj
=

fat,
sol

—0.38-6.49***
8.5

2.

&

meat
+

bone,
shoulder
+0.32
+

s.2B***
5.0

3.
x2=ham

measurement
.

+O.ll
+4.l3***
3.5

3.

Fat,
sol

-0.36
-3.59***
2.3

y4=Carcass
score»)
1.

x5=age
at
88
kg

-0.56
-B.l6***
27.4

1.

Age
at
88
kg

-0.55
—9.oo***
26.7

2.
xx
=

fat,
sol

-0.47
-6.45***
17.1

2.

&

meat+bone,
shoulder
+0.40
+s.26***
9.1

3.
x2=ham

measurement.

+0.15
+

3.17**
4.1

3.

Fat,
sol

—0.46
3.76***
4.7

*)

a)

Regression
analyses
of
data

corrected
for
weight
at
ultrasonic
measurementand

weight
of

carcass.

95



96

d. Additional information from barrow sibs in assessment of breeding value of
boars

Data on the barrow sibs’ lean cuts results were analysed for the purpose of
seeing if additional information concerning the breeding value of the boars
could be obtained. The data used were: boars’ phenotype evaluation results,
singly and in combination, and barrow sibs’ lean cuts results, singly and in
combination. These data were used as independent variables in stepwise
multiple regression analyses.

The phenotype evaluation results of boars accounted for 23.2 56.4 % of
the overall variation in the lean cuts results, 76.6 % of that of the daily yield
of meat-(-bone, and 52.3 % of that of the carcass score (Table 41 a). Those
variables which had already been found to provide the best prediction of carcass
value, i.e. sol fat, ham measurement and rate of growth (Tables 35 and 39),
also gave the best account of the variation in the lean cuts results of the boars.
Apart from these, the average withers, midback and sol fat (1,2,4) accounted
significantly for the variation in the yield of fat-j-skin.

The lean cuts results of the barrow sibs accounted significantly for the
variation in the carcass characteristics of the boars (Table 41, a and b), the
contribution being 3.2—9.8 %. When the lean cuts results of the barrows
were combined with those of the phenotype evaluation of the boars as inde-
pendent variables, the value of the boar variables changed to some extent,
loin and withers fat, for example, becoming significant (Table 41b). By combin-
ing phenotype evaluation data (boars) and lean cuts data (barrow sibs) it was
possible to account for 31.6 58.9 % of the overall variation in the carcass
of the boars, 79.3 % for the daily yield of meat -(-bone, and 61.7 % for the carcass
score.

If the lean cuts results for the barrow sibs were taken into consideration in
selecting boars for breeding, the compilation of a carcass score from a single lean
cuts characteristic would be much simpler than from many such characteristics.
Consequently, stepwise multiple regression analyses were carried out including,
as independent variables, in addition to the boar’s phenotype evaluation results,
only the combined lean cuts results from the barrows. Of the barrow variables
the most useful were the meat+bone and fat-f skin percentages. The weight
of meat+bone of the ham and loin provided a better account of the variation
in the carcass score than did the other variables (Table 42); however, percentage
figures are more suitable for the calculation of carcass score than are weight
figures, which are more closely related to the weight of the animal. Relationships
containing the variables providing significant accounts of the variation in the
meatiness, fat content, daily yield of meat-(-bone and carcass score of the boar
are presented in the form of regression equations. In one series of analyses,
the only independent variables were the boars’ phenotype evaluation results
(Table 41), while in the other these results were combined with the lean cuts
results of the barrows (Table 42).

Although the barrow sibs’ lean cuts results provided little additional account
of the variation in the carcass value of the boars, the results are nevertheless
valuable because the barrows meat+bone figures proved to be considerably



Table
42.
Assessment
of
value
of
boars
for

breeding
on

the
basis
of

phenotype
testing

results
of
the

boars
and
the
lean
cut
criteria
of
their

barrow

siblings.
122

boars
(g),

122
barrows
(^-).

Dependent
variables

Independent
variables

T
value

R2%Regression
equation

Main
part
of

carcass:
xx
=

(J,
fat,
sol

s.ls***
14.9

y 1=
meat
+

bone,
%

x2=&,

meat+bone,
%

of
carcass...
+

4.97***
13
-

8
»;,

Yi
=

°-
345-0.0031
x,
+

0.41
x

2

y2=fat
4-skin,
%

x3=<J,
meanfat

(1,
2,4)

+9.29***
35.8

of
carcass

x4=$,

growth/day
(20-88
kg)

....-3.08**

3.9

y2=0.066
4-
0.0039

x3-0.000050
x4+0.18

x

5

x
5=£,

fat
4-skin,
%

of
carcass....
4-
2.76**

3.2

y
3=meat

4-bone,
x6=$,

age
at
88
kg

-17.52***
57.9

g/day

Xl
=

(J,
fat,
sol

-5.79***
6.3

y3=140.49-0.718
xx
4-

82.079
x2-0.639
x6+0.332

x

7

score

xx
=

<J,
fat,
sol

-6.4l***
15.1

y4=927.81-8.384
X]

-
3.194
x,
+

0.0527
x

8

7 '.17



98

Table 43. Heritability values (h2) calculated (using uncorrected data) from the half sib corre-
lations of the carcass characteristics of progeny, measured ultrasonically and after slaughter
at about 88 kg liveweight.

Boars, n = 122 Barrows, n = 122
Characteristics h 2 4: s.e. h 2 4; s.e.

Ultrasonically measured fat thickness:
withers (1) -0.03 ± 0.00 0.24 ± 0.23
midback (2) 0.01 ± 0.05 0.06 ± 0.12
loin (3) -0.05 ± 0.00 0.40 ± 0.302 )

sol (4) -0.02 ± 0.00 0.12 ± 0.16
mean 1,2, 3 -0.05 ± 0.00 0.25 ± 0.23
mean 2,4, 5 -0.02 ± 0.00 0.17 ± 0.19

Fat thickness measured on carcass:
withers (1) 0.51 ± 0.352) 0.19 ± 0.19
midback (2) 0.58 ± 0.382) 0.17 ± 0.18
loin (3) 0.75 ± 0.433) 0.49 ± 0.312 )

side (4) 0.29 ± 0.27 0.09 ± 0.13
mean 1,2, 3 0.75 ± 0.423) 0.25 ± 0.221 )

mean 1. 2,3, 4 0.67 ± 0.403) 0.15 ± 0.17

Lean cuts, half carcass �
meat 4- bone, back 0.98 4; 0.493) 0.63 ± 0.353 )

» , ham 4- loin 0.53 ± 0.362) 0.68 ± 0.365 )

Main part
meat 4- bone 0.82 ± 0.443) 0.83 ± 0.403 )

meat 4- bone, % 0.29 ± 0.26 0.73 ± 0.383 )

fat + skin 0.47 ± 0.34 2 ) 0.07 ± 0.11
fat + skin. % 0.60 ± 0.382) 0.20 ± 0.20
fat + skin/meat + bone 0.63 ± 0.393) 0.40 ± 0.292)

meat + bone (g/day) 0.31 ± 0.28 1) 0.30 ± 0.251)

carcass score 0.33 ± 0.28 1) 0.54 ± 0.34 s
Other carcass characteristics:

longissimus muscle area 0.69 ± 0.41 3) 0.68 ± 0.363 )

side length 0.86 ± 0.463 ) 0.65 ± 0.363 )

ham measurement 0.54 ± 0.37 2 ) 0,40 0.282)

forelegs (points) 0.26 ± 0.25 0.13 ± 0.16
rearlegs » 0.40 ± 0.32 1) 0.06 ± 0.11

Significance levels of between-sire differences: ') P < 0.05 2) P < 0.01 3 ) P < 0.001.



99

more reliable in terms of degree of heritability than did the ultrasonically
measured fat thickness measurements of the boars (Table 43). No h2-estimates
were obtained for these particular fat figures, though the heritabilities of the
corresponding carcass fat thickness, side fat excluded, were very high and reli-
able (Table 43). For the barrows both the fat thickness measured ultrasonically
and on the carcass gave low h2-estimates, of almost the same size. These low
h2’s can be explained, because the fat thickness of barrows markedly affected
by the metabolic change resulting from castration, which evidently disrupts
the expression of inherited characteristics. This effect can be seen also in the
low h 2-estimates of the fat-(-skin components of the barrows (Table 43). The
low h2-estimates for the ultrasonically measured boars’ fat, however, cannot be
the result of anything but the inaccuracy of the measurements. The ultrasonic
measurement of fat thickness of the boars was less accurate than that of the
barrows for the obvious reason that the relative error is greater with thin fat
than with thick. The h2-estimates for ham measurement were relatively high
(Table 43).

If the breeding value of the boars had been predicted from their barrow
sibs’ lean cuts results and rate of growth alone, the following figures would have
been obtained:

Carcass characteristic of boars /f 2%

Meat+ bone of back, g 21.8
Meat +bone of ham + loin, g 15.5
Meat +bone of main part of carcass, g 18.5

» » » » % of carcass 18.3
» » » » g/day 21.7

Carcass score 20.0

Thus, a phenotype evaluation of boars provides a much more effective account
of the variation in their carcass value than a full barrow sib evaluation alone
(see. Table 41 a). A combined boar phenotype- and full barrow sib evaluation
will provide the best result.



100

IV. Discussion

A. Efficiency of phenotype evaluation of boars

7. Influence of animal material on results

Only a small number of animals can usually be housed at a testing station,
so that it is necessary that the preselection, according to the breeding value of
the parents, be as strict as possible. Thus the differences between the boar
progeny will diminish. It is consequently important to ascertain whether such
slight differences can be measured reliably by station testing, or whether the
many environmental factors dominate to such an extent that the improvement
of boar stock in this way is impossible.

In earlier studies boars were tested on farms, the numbers of boars were
large and the differences in fat thickness were considerable (Table 5). There
have been very few studies of the type reported here, in which boars were
selected according to grading at a testing station, and the efficiency of the
selection was checked by evaluation of the boars’ progeny, reared also at a
testing station. Fewson et al. (1962) and Minkema et al. (1964) carried out
similar investigations, using four progeny-tested boars. The former team
graded the boars (two high-grade and two low-grade) on the basis of fat thick-
ness (determined by »leanmeter»), whereas the latter selected by means of an
index calculated from feed efficiency, rate of growth, and ultrasonically mea-
sured fat thickness and the cross-sectional area of the longissimus muscle.
Minkema et al. assessed the one high-grade and 3 low-grade boars (the second
high-grade boar had to be replaced by a low-grade) by progeny testing,
and compared the results with those of an evaluation of progeny of full sibs.
Persson and Lindhe (1972) are using a larger and steadily growing number
of pigs, but this work is still in progress.

The study reported here comprises 26 sires and 441 progeny; even numbers
of this order are relatively small for heritability studies. In order to increase
the differences between the sub-groups it would have been desirable to rear
all the 60 test boars at the same time, but owing to the small size of the testing
station they had to be reared in two groups (30, 30), one after the other. In
addition, selection in the later group could not be made simultaneously (see p.
38). It would have been an advantage if all the animals used in the study had
been of one breed. However, in order to accelerate the process of evaluating
the progeny the use of the two breeds (which to all intents and purposes are the
only ones in Finland) was necessary.



101

The five best boars, five boars of average ranking, and five poor boars were
selected for progeny evaluation from each group of 30 preselected boars. Natu-
rally, the differences between these 3 sub-groups were slight. As an example,
for average fat thickness the sub-group I vs. sub-group 111 difference was only
3.0 mm in Group A and 5.9 mm in Group B, the I —II differences being 1.1 mm
and 1.2 mm respectively (Tables 13 and 14). Consequently, considerable demands
are made on the methods of testing if such slight differences in fat values are to
be measured reliably by ultrasonic techniques, so that boars with the meatiest
progeny may be selected. Ultrasonic measurement has generally been regarded
as satisfactory already when it deviates less than 2 mm from that obtained by
measurement of the carcass (seep. 25). Accordingly, the results of this study, in
which the sub-group with the poorest carcass quality could be distinguished
from the other sub-groups with a reliability of some 95—99.9 per cent, can be
regarded as good (Tables 13—14, 24—27, p. 74), and are comparable with the
results obtained by Fewson et al. (1962) and Minkema et al. (1964) (see.
p. 35), as there was no average-grade sub-group in their studies.

In analysing the data, one object was to determine the effect of breed on
the results. Breed differences in effect were principally between-boar differences,
since selection was based on the number of test points and was quite independent
of breed, apart from trying to balance the number of boars of the two breeds
(see p. 40). The breed differences found were inconsistent, particularly in
carcass characteristics (see p. 85); differences in rate of growth were due mainly
to a few exceptionally slow-growing B Group boars (Table 14). Usually the
differences in rate of growth between the two Finnish breeds of pig are slight.
For example, in a 4-year period immediately preceding this study (1964 67),
the daily growth of Landrace pigs on the pig experimental stations was 10 g
less than that of Yorkshires, whereas earlier it had been 10 g more (Lounamaa
1968). In other Finnish tests too, breed differences have been slight or even
contradictory (Ettala 1971 a, Maijala and Vainikainen 1962, Uusisalmi
1971 d, 1972). This similarity is actually quite understandable, as the breeding
programme has been identical for the two breeds.

Differences between the farms from which the progeny of the two breeds
were taken can hardly have influenced the results, for the number of farms in
question was very high one of the advantages of the artificial insemination
programme. In Group A the Yorkshire progeny were obtained from 26 farms
(average 2.8 pigs per farm) and the Landrace progeny from 44 farms (1.8 pigs
per farm). The corresponding figures for Group B were 37 farms (2.3 litters per
farm) and 53 farms (1.5 litters per farm). Farms from which the progeny of
any one boar were taken were situated fairly uniformly all over the collecting
area. Exceptions were the three boars lent to farms where they were mated
with dams (see p. 44); each of these boars had progeny at 3—6 farms only.
Even this ratio, i.e. 36 farms per boar, may be regarded as satisfactory for
testing purposes (Mennerich 1967), for as yet AI material is scarce in studies
of this kind. Another consequence of the large number of farms was that any
close relationship between the dams was quite unlikely (see. p. 44), so that the
relationship between the half sibs could be taken as 0.25.



102

2. Effects of environment

Environment of the piglet was reflected in the variable
weights and ages of the piglets at the commencement of the testing period.
The weight of each piglet and the whole litter at the age of three weeks was
significantly related to piglet weight at the beginning of the testing and, in
turn, initial weight had a significant effect upon the subsequent rate of growth,
especially in the range 20 to 60 kg, and to some extent also on fat thickness
(Tables 12, 17, 19, 21,23, see p. 33). The effect of initial age on age at 60 and
88 kg was also significant. In analysing the results, corrections of initial weight
were made by means of linear regressions except when hierarchic variance
analysis and covariance analysis were performed. Also in Sweden in evaluating
growth rate and feed efficiency it was considered necessary to make such
corrections (Persson & Lindhe 1972). An alternative would be farm rearing
of piglets under closely controlled conditions, as has been planned in Sweden
in order to minimise variation in initial weight (Persson and Lindhe 1972).

The effect of weight at ultrasonic measurement on the
fat content of the carcass was highly significant when the weight at ultrasonic
measurement varied considerably, as was the case with the test boars (Tables
11, 12). A correction in respect of weight was then necessary (see pp. 50, 32).
This weight-fat relationship was only slight, however, when the ultrasonic
measurement was made once a week, as was the case with the progeny of Group
B (Tables 18, 19). A correction for weight was done in all cases except when
hierarchic variance analysis and covariance analysis were carried out. Generally,
it is quite practicable at boar testing stations to carry out the ultrasonic
measurement at a liveweight close enough to the specified one so that no
corrections for weight will be necessary. The age at ultrasonic mea-
surement was significant in respect of some fat thickness measurements
(Tables 12, 19); corrections for age were not made.

The effect of amount and composition of feed (Tables 7 —10)
on rate of growth and fat thickness was pronounced in the present study, as it
has been in many others (see p. 32); there were large differences in rate of growth
and fat thickness between the sires and the boar progeny of Group B (sires
604 g/day, 16.9 mm, boar progeny 761 g/day, 20.9 mm). The feeding of the
progeny of Group A was altered in the middle of the testing period, and signifi-
cant differences in rate of growth and fat content, but not in meatiness, as-
sociated with the three feeding regimes (old, new or both standards) were found.
Naturally the change in feeding was a disadvantage, and an attempt was made
to eliminate its effects by means of pooling (see p. 47). However, the pooling
may have diminished the differences between Group A boars to some extent,
for although every boar had progeny on each of the feed regimes, the progeny
were not evenly distributed in this respect.

The selection of boars at Finnish boar testing stations has from
the start been based on a test-points system (Ettala 1971 a). The intention
was to reduce fat thickness and increase rate of growth simultaneously, as well
as to improve the related characteristics meatiness and feed conversion. In
order to ascertain the efficacy of the points system, the boars of Group A were



103

graded according to »old» test points, and those of Group B according to »new»
test points. The results showed that the use of test points promoted neither
rate of growth nor feed efficiency, for boar selection was based almost entirely
upon the thickness of the fat (Table 14, pp. 53, 83), This is consistent with
the finding of a previous study (Ettala 1971 b) and with that of work in Sweden
(Persson and Lindhe 1972). The test-points system thus is inappropriate as
a criterion in the station testing of boars, for in most cases station testing is
used for the specific purpose of quantifying growth rate and feed efficiency,
while the test-points system was designed to eliminate differences between
farms in field testing (Spandal 1972). A specific testing station index for
boars has been adopted in Norway and Sweden (Lindhe and Persson 1972,
Moen 1968).

Of course, ejaculate volume, semen quality and leg strength too are im-
portant factors which determine the value of a stud boar, and these
characteristics have consequently been taken into consideration in boar selection
in the present study, as they are normally in station testing (see pp. 32, 52).
The number of pigs remaining for the progeny evaluation of Group A was smaller
than planned, and medium quality reserve boars had to be brought into Group
B. In a recent Swedish study on breeding boars for artificial insemination, only
31.2 % of the selected boars could in fact be used in breeding, the remainder
having to be rejected because of leg weakness, poor semen or inverted nipples
(Persson and Lindhe 1972).

3. Heritability estimates of, and phenotypic correlations between, characteristics

As a test of the reliability of the ultrasonic measurements made on the sires,
the h 2-estimates obtained from sire-progeny regressions were compared with
those from half sib correlations. The two methods gave almost identical h -

estimates (0.44 and 0.424:0.18) for average thickness of fat, but this similarity
was not always found for individual fat thickness values (Table 31). The h -

value for side fat was the only non significant estimate as determined
from carcass measurement and from half sib correlation (Table 31). The h -

estimates were usualy similar to those reported by other investigators (Table 5,
p. 35), although the animal population in question and the environmental
conditions have a substantial effect on the magnitude of the heritabilities.

Ultrasonically measured fat thickness of the progeny did not account for
the hereditary level of their carcass characteristics as well as ultrasonically
measured fat thickness of the sires. Only the h2-estimate for loin fat was sta-
tistically reliable when the calculations were made from half sib correlations
(Table 30). This is seen also in the correlation between the fat thickness mea-
surements of the sires and those of their progeny. The correlations between
ultrasonically measured fat of the sires and the carcass fat measurements of the
progeny were much closer than those between sire fat thickness (ultrasonics)
and progeny fat thickness (ultrasonics) (Tables 28, 29). The difference was
surprisingly large considering the close correlation, found in this and other
studies, between progeny fat thicknesses measured on the live animal and on
the carcass (Tables 34, 1).



104

The h2-estimate for daily gain (20 —BB kg) was considerably smaller when
determined from the sire-progeny regression (0.32) than when calculated from
the half sib correlation (0. (Table 31), due possibly to the fact that
the growth rate of the sires was not expressed fully because of the restricted
feeding. The h2’s determined from the half sib correlations for the boars of
Group B are also obvious overestimates (Table 36) that have affected the
result. This may partly be due to the group feeding, in which the heavier
boars dominated the trough, so that the barrows obtained relatively small
amounts of feed. Also, the number of animal in question was small and the
range in rates of growth was considerable (Tables 14 and 16).

Environmental conditions during the piglet period had such a great effect
on the age at 88 kg that the h2-estimate was zero (Tables 30 and 31). This
expression of rate of growth should actually be dropped as far as station testing
is concerned. Because the animals were group fed the h2-estimates for the feed
efficiency were calculated only from the sire-progeny regressions; the values
obtained, 0.25 for Group B and 0.18 for Group A -f- Group B (Tables 30 and 31),
can be regarded as satisfactory.

Correlations between sires and progeny for daily gain and feed efficiency
were statistically reliable (Tables 28 and 29). Rate of growth and feed efficiency
were closely inter-correlated (Table 15, p. 31), so that the improvement of daily
gain and feed efficiency by breeding is possible simultaneously. However,
correlations between daily gain and meatiness of carcass and fat content were
negative, with few exceptions (Table 35). Probably there have been two types
of slow-growing pigs, those with high meatiness + thin fat and those with
low meatiness -f- thick fat. These opposing trends of rate of growth and meati-
ness or fatness of carcass are evident also in results reported elsewhere (see p.
31). Thus the recommendation that in the selection of stud boars these
characteristics should be considered separately seems to be well founded. With
this in mind, the potential for early selection by means of early phenotype
testing of progeny was studied here.

B. Appraisal of means for the improvement of station testing of boars

1. Potential for early selection of hoars

The h2-estimate, calculated from the half sib correlations of the boar pro-
geny, for the early daily gain (20 —6O kg) was (Table 36). The early
daily gain correlated very closely with the daily gain during the whole test
period (20—88 kg) and with that during the last part of the period (60—88
kg) (Table 37); Admittedly, autocorrelation may have been considerable in the
former. The genetic correlation between the daily gain in the latter, that is
20 —6O vs. 60 BB, was rg = and the phenotypic correlation rp =

o.22***. It is evident from these values that boars’ rate of growth can be
determined quite reliably even at a live weight of 60 kg only. Many other
researchers (Buchenauer 1970, Hofmann 1965, Schmid 1970, Scholz 1965,
Wohlbier et al. 1961, Zagozen and Flock 1970) have also found that rate of



growth of pigs (as well as feed efficiency) can be determined quite accurately on
the basis of a brief period of growth. The test periods which they used were
equivalent to the liveweight range 50 —BO kg. In Germany an early selection
of boars has been made at 70 kg, using the characteristics rate of growth, feed
efficiency and ultrasonically measured fat thickness (Blend 1970); the boars
were on an ad libitum diet.

In the present study, feed efficiency could not be determined at intermediate
stages because the animals were fed in groups. However, growth rate and
feed efficiency were closely interrelated, and good utilisation of feed obviously
accompanied rapid growth, even during early growth. According to many
studies, there are very close (negative) phenotypic and genetic correlations
between rate of growth and feed efficiency (see p. 31).

The ham measurement and the ultrasonically measured sol fat found in
the early test gave, when combined together, a very reliable prediction of lean
cuts results obtained after slaughter at the usual 88 kg liveweight (Table 39).
The genetic correlation between these early-test results and carcass charac-
teristics, however, was only slight (see p. 89). Extreme values in carcass cha-
racteristics could perhaps be predicted by early ultrasonic measurement, so
that it might be useful in the selection of borderline cases. Many researchers
have found a relatively close correlation between early-measured fat thickness
and carcass fatness, but most have concluded that individual differences in
carcass characteristics increase with advancing growth (Otto 1967 a and b,
Rittler et al. 1964, Schmid 1970, Scholz 1965, Skjervold et al. 1960,
Vockert 1969). Inspection of the legs should also be included in early testing.

An early selection made on the basis of daily gain, feed efficiency and leg
strength of boars would have the advantage, among others, that these economi-
cally important characteristics are expressed at their fullest at this stage.
Another advantage would be the more effective use of the testing facilities.

2. Leg strength of hoars on various levels of feeding

On restricted feeding 13 % of the test boars had to be eliminated on
account of leg faults, and a further quarter of them had slightly less serious
faults, there being some impediment to walking (3 leg points). Among the
progeny boars reared on the higher level of feeding, one-third had difficulty in
walking but in only one animal was this greatly restricted (Table 40); leg defects
associated with bone weakness caused by rapid growth could not be subs-
tantiated with statistical reliability. Yet there was a weak negative correla-
tion (rp =—0.16) between daily gain and rear leg points. It appears thus that
the role of feeding in causing leg weakness was a minor one (see p. 32). Carcass
meatiness was not correlated at all with leg points (see p. 92), which result
contrasts with the negative correlation found by Kangasniemi (1971 a) in
progeny tests.

Some of the leg faults were inherited. This was already suggested by the
fact that 5 of the 8 boars rejected because of leg faults came from two litters
only, and was confirmed later by the statistically reliable heritability estimates

8 105



106

calculated from the half sib correlations of progeny (Tables 30 and 36). The
h2-estimates for leg points scored at early testing were unexpectedly high (Table
36). One reason may have been over-estimation; another, that the number
of leg injuries increased with time, thus decreasing the h 2-estimates. In any
event, leg faults proved to have a higher heritability in the present study than
in that of Smith (1966). From a breeding point of view, it is important that
heritable leg defects be revealed by the performance tests.

In the present study leg faults arising through injury were rather numerous.
The grid over the waste channel was obviously one cause. It may well be that
leg weakness and accidentally cansed leg faults are to some extent associated
with clumsiness of movement. Roomy, well-designed stalls provided with
bedding should reduce considerably the incidence of leg defects (Schulman
1971).

3. Predicting carcass quality from results of phenotype testing

For the phenotype testing of boars to reflect as reliably as possible the
breeding value with regard to desirable carcass traits the characteristic should
fulfill the following requirements: 1) be accurately measurable, 2) express
reliably the meatiness of the carcass, and 3) possess a high heritability and a
close genetic correlation to the meatiness of the carcass.

When the measurements of fat thickness made during this study are
examined in the light of these requirements, it can be seen that sol fat meets
requirements (1) and (2) very well, that loin and withers fat measured on the
carcass meet (2) and (3) very well, and that midback fat falls in between these
two extremes (Tables 30—32, 34, 35). As the remaining fat values also meet the
requirements to a satisfactory degree, it follows that the fat values supplement
one another in this respect. The present study shows that the best estimate
of the heritability of carcass characteristics is obtained when the fat thickness,
measured ultrasonically, for withers, midback, loin and side are combined.
Sol fat need be measured on one side only, since measurement on both sides
will not improve the reliability of the estimate. This conclusion is supported
by the fact that it was the average of all the sires’ fat thickness values, measured
ultrasonically, that correlated most closely with their progeny’s lean cuts
results (Table 29). Ultrasonic measurements at all locations would also ensure
an even reduction of fat thickness over the various parts of the carcass as
breeding progresses. A number of authors have actually recommended that
ultrasonic measurement be made at 3 5 locations (see p. 26).

The ham measurement met the above-mentioned requirements, except that
its genetic correlation with carcass meatiness was slight (Tables 30, 35, p. 82).
The phenotypic correlation between ham measurement and lean cuts results
was surprisingly high (Table 35): considerably higher than that obtained by
Uusisalmi (1971 c) when he introduced the use of this characteristic. The
use of ham measurement in the prediction of carcass value should be continued,
at least for reserach purposes.

The side length and cross-sectional area of the longissimus muscle as measured



107

on the carcass of the progeny satisfied all three requirements completely, but
were not measured phenotypically. It is unfortunate that progeny side length
was not measured on the live animals, for the sires’ side length (live animal)
is significantly correlated with that of the progeny carcass (r = o.62***).
Many positive correlations between carcass length and meatiness, or at least
negative correlations between length and fat content, have been reported
(Jonsson 1963, Kline and Gall 1964, Krippl et al. 1965,Muller-haye 1965,
Pearson et al. 1956 a, b. 1959, Uusisalmi 1971a, Varo 1962). Admittedly,
others have found no correlation between these characteristics (Kliesh et al.
1960, Pedersen 1968, Rittler et al. 1965).

It can be seen from Table 35 that the cross-sectional area of the longissimus
muscle was a very good indicator of the meatiness of the carcass. Its heritability
and genetic correlation with meat +bone were also high (Tables 30, 33, 43).
This finding agrees with those of several other studies (Tables 4,6, p. 35).
Ultrasonic measurements of the longissimus muscle have, however, been taken
rarely, because of the difficulty in making accurate determinations. New
ultrasonic equipment nevertheless is expected to bring about an improvement
in the situation (Table 2). In Finland results obtained with such equipment
are encouraging (Varo 1971). As fat thickness steadily decreases with each
generation of stud boars, all other measureable characteristics, in particular the
cross-sectional area of the longissimus muscle, become increasingly important.

4. Additional information about variation in hoars’ carcass quality from harrow
sibs’ lean cuts analysis

Lean cuts results should provide a good indication of the breeding value of
the animal, because the yield of meat+bone in particular showed high and
reliable h2-estimates (Tables 31 and 43, cf. Uusisalmi 1969 b, 1971 d).

The barrow sibs’ results provided statistically significant but rather low
coefficients of determination (R 2 % = 3.2—9.8) (Tables 41, 42), in accordance
with figures obtained by Schmid (1970). The additional information provided
by barrows sibs is valuable not only because the independent variables were
highly transmittable but also because the barrow data gave the highest in-
crease in information of variation in the meatiness (meat+bone in % of carcass)
of the boar. This carcass characteristic is an important one, and its variability
was accounted for less well than was that of the other carcass characteristics
by the phenotypic characteristics of the boars themselves (Table 41a). Schmid
(1970) found that a full sib evaluation supplements a phenotype evaluation of
boars, provided that the animals are reared at the same time.

Combining early testing with evaluation of barrow sibs to provide auxiliary
data in phenotype testing deserves further study, for it would allow the selection
of boars from a large population, and more animals could be tested as boars
rather than as barrows. As the rate of growth does not seem to cause any syste-
matic difference in carcass value, animals selected on the basis of rate of growth
and feed efficiency could very well be used for sib evaluation. If late-castrated
pigs, in the weight range 60 to 88 kg, grow at the same rate as the barrows in



108

the present study, i.e. 861 g/day, this final stage of growth would last five
weeks, long enough for the »boar-taint» to disappear. Kruger et al. (1970)
investigated late castration (at about 80 kg) and found no undesirable effects
provided the animals are kept in clean stalls for one week; the pigs were
slaughtered four weeks after castration (at about 100 kg) and no »boar-taint»
was observed.

For the most reliable results, sib evaluation should be carried out with
uncastrated boars. Risks are incurred with castration, and the boar carcass
is meatier and has less fat than the barrow (Table 22). Both boars could be
selected for breeding, the sib evaluation being based on the phenotype of the
brother.

The optimum number of progeny-boars selected from any one litter for
phenotype tests is two (Jonsson 1971 b). In the present study, boars from the
same litter proved to be of equal value in such tests (Tables 13, 14, 25 and
27).

To summarise, the present study showed that the phenotype testing of
boars provides a very good measure of the meatiness of their progeny, though
the differences in fat thickness were small and the number of boars available
for selection at any one time was low. The value of the testing results could
obviously be further improved by increasing the number of boars and enhancing
the level of feeding. The differences between boars would then be relatively
large even in a population of highly selected boars. The reliability of boar
carcass evaluation can be improved by increasing the number of locations
at which measurements are made, and by testing a second boar from the same
litter. Greater importance should be attached to the rate of growth and feed
efficiency of boars, by altering the evaluation method, e.g. by employing early
testing as described above. As a short generation cycle is another advantage
of phenotype testing, its potential for rapid improvement of boar stock is
obvious.



109

V. Summary

The efficiency of phenotype testing highly selected boars was studied. Two
groups, each of 30 boars, were tested centrally at stations for growth rate and
ultrasonically measured fat thickness. According to test points, made up of a
combination of these two traits, the 5 best boars, 5 average boars and
the 5 poorest boars where selected for progeny evaluation. In all 26 boars
with 441 progeny were tested.

1. The progeny evaluation showed that ultrasonic measurement of the fat
thickness of the boars gave a very reliable estimate of the meatiness of their
progeny. This conclusion is based on the following observations:

a. Those boars, as a group, giving the poorest carcass quality could be
distinguished with statistical reliability (P<0.05 or P<0.001) from the other
groups, even though the differences in fat thickness of the boars were quite small.

b. The average fat thickness by ultrasonic measurements made on the
boars correlated very closely with averages obtained from lean cuts analysis
of the progeny carcasses.

c. The heritability estimate for fat thickness from sire-progeny regression
(average fat thickness determined by ultrasonic measurements on the boars
and direct measurement on the carcasses of the progeny) was as good as that
from half sib correlation of the progeny (0.44 and 0.42 iO. 18 respectively).
This similarity was not found for individual fat thickness measurements (e.g.
midback fat), however.

2. The correlations between sires and progeny were significant (P<0.05)
for both daily gain (period 20 —88 kg) and feed efficiency. The heritability est-
imate for daily gain from sire-progeny regression was slightly lower (0.32) than
that found from half sib correlations of the progeny (0.49^:0.18).

3. The importance of rate of growth and feed efficiency has been neglected
in selection for breeding as the test points used for the selection of the boars
depended almost entirely on fat thickness.

4. The daily gain of the progeny boars could be determined with sta-
tistical reliability (P<0.001) as early as two months after the start of the testing
period; their liveweight at this stage was about 60 kg. Thus it is possible to
select boars early on the basis of their initial growth (from 20 to 60 or 20 to
70 kg).

5. More than 30 % of both boars and progeny boars suffered from some
form of difficulty in walking. 13 % of the boars were eliminated because of
leg faults. Leg faults in progeny were mainly inherited or caused through injury.
The effect of level of feeding on leg strength appeared to be slight.



110

6. The results of the phenotype tests on the boars were affected by the
following factors:

a. Changes in feeding had significant effects upon the rate of growth and
fat thickness.

b. Variations in initial weight and initial age had a significant effect upon
growth rate and feed efficiency. Their greatest effect was on growth rate ex-
pressed as age in days at a certain weight.

c. The effect of the weight at ultrasonic measurement on fat thickness
was highly significant whenever the variation in the former was considerable.
This effect was only slight when the ultrasonic measurement was taken at a
liveweight as close as possible to 88 kg (the animals were weighed weekly).
The effect of age at ultrasonic measurement was smaller than that of weight.

8. Among the measurements made on the live progeny and on their car-
casses the most useful, in estimating the breeding value of the boar as regards
meatiness, were the ultrasonic measurements of fat thickness at withers, mid-
back, loin and side (sol). The ham measurement gave an additional significant
account of variation in meatiness. Of the measurements made on the half-
carcass, the cross-sectional area of the longissimus dorsii muscle was the best
index of meatiness. Length of side also was a useful measurement.

9. The barrow sibs’ lean cuts results gave a highly significant account
(R 2% = 3.2 9.8) of variations in the quality of the progeny boar carcass.



111

REFERENCES

Andrews, F. N. & Whaley, R. M. 1955. Measure fat and muscle in live animal and carcass.
Rep. Ind. Agric. Exp. Sta. 68:27 29. (Ref. Horst, P. 1964 b.)

Anon. 1958. Uusi luotausmenetelmä [A new method of ultrasonic measurement]. Sika,
195: 36-37.

Assadi, J. 1967. Die Messung der Querschnittfläche des M. longissimus dorsi am lebenden Tier
unter Verwendung des Ultraschall-Schnittbildgerätes und die Beziehung dieser Mess-
werte zu einigen Schlachtmerkmalen beim Schwein. Diss., 50 p. Giessen.

Berruecos, J. M., Dillard, E. U. & Robison, O. W. 1970. Selection for low backfat thickness
in swine. J. Anim. Sci. 30: 844—848.

Biedermann, G. 1971. Neue genetische Parameter wichtiger Mast- und Schlachtleistungs-
merkmale bei der Deutschen Landrasse. Z. Tierz. Ziicht.biol. 88: 291 307.

Blendl, H. M. 1970. Ad-libitum-Fiitterung in der Eigenleistungsprlifung beim Schwein. Tier-
ziichter 22: 576 577.

Brunstad, G. E. & Fowler, S, H. 1959. Some carcass characteristics of swine in the eighth
generation of production under four combinations of full and limited feeding. J. Anim.
Sci. 18:211-220.

Buchenauer, D. 1970. Untersuchungen iiber die Aufzuchtleistungen eigenleistungsgepriifter
und ungepriifter Sauen, sowie die Mastleistung und den Schlachtkorperwert ihrer Nach-
kommen. Diss., 105 p. Giessen.

Buck, S. F., Harrington, G. & Johnson, R. F. 1962. The prediction of lean percentage of
pigs of bacon weight from carcass measurements. Anim. Prod. 4: 25 36.

Clausen, H. 1959. Haling af rygflaeskets tykkelse pä levende grise. Landokon. Fors.lab.
Efterärsmode 1959; 264—266.

Cole, D. J. A. & Hardy, B. 1971. Effect of dietary digestible energy and crude protein on per-
formance and carcass of growing pigs. 10th Intern. Congr. Anim. Prod., Theme VII.
Mimeogr. 7 p. Versailles.

Cooke, R., Lodge, G. A. & Lewis, D. 1972. Influence of energy and protein concentration in
the diet on the performance of growing pigs. 1. Response to protein intake on a high-
energy diet. Anim. Prod. 14: 35 46.

Cross, H. R., Carpenter, J. W. & Palmer, A. Z. 1970. Pork carcass muscling; fat, lean and
bone ratios. J. Anim. Sci. 30; 866 871.

Diekmann, A. 1960. Erkennung von Fettansatz und Fleischbildung am lebenden Schwein mit
Ultraschall. Diss., 69 p. Gottingen.

Draper, N. R. & Smith, H. 1966. Applied Regression Analysis p. 407. New York. London.
Sydney.

Dumont, B. L. 1957. Nouvelles mdthodes pour I’estimation de la quality des carcasses sur les
pores vivants. FAO/FEZ Tagung iiber Mastleistungspriifungen an Schweinen, Kopen-
hagen. (Ref. Lauprecht, E. 1960).

Enfield, F. D. & Whatley, J. A., Jr. 1961. Heritability of carcass lenght carcass backfat
thickness and loin lean area in swine. J. Anim. Sci. 20: 631 634.

Englisch, H-G. 1969. Die Beeinflussung der mittels Ultraschall-Echolot-Messungen festges-
tellten Schlachtleistungsmerkmale durch Masse und tägliche Zunahme der getesteten
Zuchttiere. Arch. Tierzucht 12:413—425.

Ettala, E. 1971 a. The performance testing of boars. I. Performance testing on Finnish experi-
mental stations and genetic and environmental influences on test results. J. Scient.
Agric. Soc. Finl. 43: 53 62.

» 1971 b. 11. Phenotypic and genetic correlations. J. Scient. Agric. Soc. Finl. 43: 103 110.
Fender, M. 1962. Untersuchungen fiber die Zuverlässigkeit der Abschätzung des Fleisch- und

Fettanteiles bei Schweinen aufgrund von Ausschlachtungsbefunden. Z. Tierz. Ziicht.
biol. 77: 273-284.



112

Fewson, D., Fender, M. & Kirsch, W. 1962. Priifungsanstalten als Hilfsmittel fiir die Selek-
tion in der Schweinezucht. 2. Selektionsversuch zur Priifung der Effektivität der Eigen-
leistungsprufung von Ebern. Z. Tierz. Zucht.biol. 77: 265 272.

Flock, D. K. 1970. Genetic parameters of German Landrace pigs estimated from different re-
lationships. J. Anim. Sci. 30: 839 843.

» Schutzbar, W. von & Lauprecht, E. 1970. Beziehungen zwischen der Eigenleistung
von Jungebern und den Priifungsergebnissen ihrer Vollgeschwister und Nachkommen.
Z. Tierz. Zucht.biol. 86: 316-324.

Fredeen, H. T. & Jonsson, P. 1957. Genic variance and covariance in Danish Landrace swine
as evaluated under a system of individual feeding of progeny test groups. Z. Tierz.
Ziicht.biol. 70; 348-363.

Gerlach, J. 1967. Untersuchungen iiber die Selektion von Zuchtsauen unter besonderer Be-
rucksichtigung einer Eigenleistungspriifung im Ziichterstall. Diss., 66 p. Hohenheim.

—» 1970. Echolotmessungen als Hilfsmittel fur die Zuchtwertschatzung beim Schwein.
4. Mitteilung. Untersuchungen iiber die Selektion von Zuchtsauen unter besonderer
Beriicksichtigung einer Eigenleistungspriifung im Ziichterstall. Z. Tierz. Ziicht.biol.
86: 325-348.

Gerwig, C. 1965 a. Die Anwendung des Ultraschallverfahrens för die Beurteilung der Schlacht-
qualität bei Ebern. Schw. Landw. Mon.hefte 43:421—423.

» 1965 b. Haltung und Fattening von Ebern in Eigenleistungspriifungen Schw. Landw.
Mon.hefte. 43:423-426.

Glodek, P. 1964. l)ber ein unter praktischen Verhältnissen anwendbares Verfahren zur Abschät-
zung des Fleischbildungsvermögens lebender Schweine mit Ultraschall. Ziichtungskunde
36: 356-370.

—» Hosman, L. & Haring, F. 1971. Mastleistung und Schlachtkorperwert europäischer
Landrasse-Schweine bei unterschiedlicher Fiitterungsintensität. 10th Intern. Congr.
Anim. Prod., Theme 11. Mimeogr. 7 p. Versailles.

Gray, R. C., Tribble, L. F., Day, B. N. & Lasley, J. F. 1968. Results of five generations of
selection for low backfat thickness in swine. J. Anim. Sci. 27: 331—335.

Hald, A. 1952. Statistical theory with engineering applications, 765 p. New York, London.
-Hartmann, W. & Fewson, D. 1967. Zuchtplanung beim Schwein. Ziichtungskunde 39: I—B-
Harvey, W. R. 1966. Least-squares analysis of data with unequal subclass numbers. ARS

20—8. Agric. Res. Service, U.S. Dept, of Agric. 157 p.
Hazel, L. N. & Kline, E. A. 1952. Mechanical measurement of fatness and carcass value on

live hogs. J. Anim. Sci. 11: 313 318.
—» & Kline, E. A. 1959. Ultrasonic measurement of fatness in swine. J. Anim. Sci. 18:

815-819.
Hetzer, H. O. & Harvey, W. R. 1967. Selection for high and low fatness in swine. J. Anim.

Sci. 26: 1244-1251.
Hofmann, E. 1965. Untersuchungen zur Frage der frtihzeitigen Abschätzung der Mastleistung

der Schweine auf Grund von Teilleistungen sowie die Beurteilung der erblichen Veran-
lagung von Vatertieren. Diss., 83 p. Berlin.

Hofmann, F. & Peter, W. 1964. Die Anwendungsmoglichkeiten der Speckdickenmessung mit
Ultraschall bei der Selektion in der Fleischschweinezucht. Arch. Tierzucht 7: 255 287.

» Peter, W. & Boettcher, H. 1965. Eigenleistungspriifungmit Ultraschallbei Jungsauen.
Arch. Tierzucht 8: 331—336.

Holland, L. A. & Hazel, L. N. 1958. Relationship of live measurements and carcass cha-
racteristics of swine. J. Anim. Sci. 17: 825 833.

Horst, P. 1964 a. Entwicklung eines Verfahrens zur Durchfiihrung von Ultraschallmessungen
beim lebenden Schwein. Z. Tierz. Ziicht.biol. 80: 341 364.

» 1964b. Kritische Bewertung der Methoden zur Bestimmung des Schlachtkorperwertes
beim lebenden Schwein 11. Teil. Z. Tierz. Ziicht.biol. 80: 365—401.

» 1969. Untersuchungen iiber die zilchterische Bedeutung des Ultraschall-Schnittbildver-
fahrens bei der Eigenleistungs-Priifung von Jungebern. Z. Tierz. Ziicht.biol. 86: 58 88.

» 1971. Erste Untersuchungsergebnisse iiber den Einsatz des »Vidosom-Schnittbildgerätes
beim Schwein. Ziichtungskunde 43: 208 218.



113

Jensen, P., Craig, H. B. & Robison, O. W. 1967. Phenotypic and genetic associations among
carcass traits of swine. J. Anim. Sci. 26: 1252 1260.

Johansson, I. & Korkman, N. 1950. A study of the variation in production traits of bacon
pigs. Acta Agric. Scand. 1: 62—96.

J onsson, P. 1959. Investigations on group feeding versus individual feeding and the interaction
between genotype and environment in pigs. Acta Agric. Scand. 9: 204—228.

» 1963. Danish pig progeny testing results. Z. Tierz. Zucht.biol. 78: 205 252.
* 1971 a. Population parameter estimates of the Danish Landrace pig. Acta Agric. Scand.

21: 11-16.
» 1971 b. Model computation on future expectation of genetic gain. Genetic covariation

betveen the two sexes in Danish pig progeny testing. Acta Agric. Scand. 21: 17 25.
Kalm, E. 1972. Kriterien zeitgemässer Leistungsprufung beim Schwein. Tierzuchter 24: 94,

96-97.
Kangasniemi, R. 1971 a. Jalkavikojen, lihan laadun ja lihakkuuden välisistä vuorosuhteista

[On the inter-relationships between leg defects, meat quality and meatiness]. Sika
1971, 5: 25, 26, 29, 32. 33.

» 1971 b. Tuloksia sikakantakokeissa mitattavista ominaisuuksista, niiden keskinäisistä
vuorosuhteista ja periytyvyydestä. [Results of characteristics measurable in progeny
tests on pigs, and of their inter-relationships and heritability]. Mimeogr. Bp. Tikkurila.

Kirsch. W., Fewson, D. & Fender, M. 1962. Priifungsanstalten als Hilfsmittel fiir die Selek-
tion in der Schweinezucht. 1. Modellrechnungen iiber die Effektivität verschiedener
Priifungsmethoden und Priifungssysteme. Z. Tierz. Zucht.biol. 76: 388 407.

Kliesch, J. & Horst, P. 1961. Weitere Untersuchungen zur Frage der Anwendung der Ultra-
schallmesstechnik fiir die Beurteilung der Fett- und Fleischentwicklung am lebenden
Schwein. Zuchtungskunde 33:447—461.

—» & Horst, P. 1962. Untersuchungen zur Methodik der Ultraschallmessungen am le-
benden Schwein unter Verwendung eines Schnittbildgerätes. Zuchtungskunde 34: 350
358.

» Neuhaus, U., Silber, E. & Kostzewske, H. 1957. Versuche zur Messung der Speck-
dicke am lebenden Tier mit Hilfe des Ultraschalls. Z. Tierz. Zucht.biol. 70: 29 32.

» Silber, E. & Horst, P. 1960. Weiterer Beitrag zur Frage der Bedeutung der Körper-
länge beim Schwein. Zuchtungskunde 32: 263 268.

Kline, E. A. & Goll, D. E. 1964. Relative value of longissimus dorsi muscle area at eight
different locations in pork carcass evaluation. J. Anim. Sci. 22: 414—417.

—* & Hazel, L. N. 1955. Loin area at tenth and last rib as related to leanness of pork
carcasses. J. Anim. Sci. 14; 659 663.

Krippl, J., Patz, H. & Simon, D. 1965. Untersuchungen iiber die Erblichkeitsverhältnisse
wichtiger Leistungseigenschaften beim deutschen veredelten Landschwein. Bayer.
Landw. J.buch. 42:917 940.

Kruger, L,, Dzapo, V. & Jesswein, H. 1970. Der Einfluss einer Spätkastration auf die Mast-
und Schlachtleistung bei Schweinen. Z. Tierz. Zucht.biol. 86: 356 371.

Langholz, H.-J. 1965. Das ziichterische Hilfsmittel der stationären Nachkommenpriifung
beim Schwein. I. Systematische Einfliisse auf die Ergebnisse aus der Mastleistungspru-
fung fiir Schweine. Acta Agric. Scand. 15: 115 144.

—» 1966.111. Heritabilitäten und genetische Korrelationen beim nonvegischen Land-
schwein. Acta Agric. Scand. 16:97 114.

Langlet, J. F., Ernst, E. & Glaner, H. D. 1968. Die Möglichkeit einer Selektion bei Schwei-
nen mit Hilfe von Echolotmessungen unter den Verhältnissen der Praxis. Z. Tierz.
Zucht.biol. 85: 27 35.

Lauprecht, E. 1960. Bericht iiber Ergebnisse von Messungen mit dem Ultraschall-Echolot
an Schweinen. Zuchtungskunde 32:441—449.

—» Kirsch, W., Rittler, A., Mönzer, W. & Fewson, D. 1965. Untersuchungen iiber
die Abschätzung des Anteiles der fleisch-und fettreichen Teilstiicke aufgrund von
Echolotmessungen am lebcnden Schwein. Z. Tierz. Zucht.biol. 81: 89 99.

—* Scheper, J. & Schröder, J. 1957. Messungen der Speckdicke lebender Schweine
nach dem Echolotverfahren. Mitt. Deut. Landw. Ges. 72: 881 882.



114

Lauprecht, E., Schutzbar, W. von & Flock, D. 1967. Untersuchungen iiber die Heritabilität
der mit dem Ultraschall-Echolot and Jungebern ermittelten Speck- und Muskeltiefe.
Z. Tierz. Ziicht.biol. 84: 80 95.

» Walter, E., Saathoff, T. 1960. Beitrag zur Messung der Querschnittsfläche des langen
Riickenmuskels (musculus longissimus dorsi) und der dariiberliegenden Fettschicht
von lebenden Schweinen mit dem Ultraschall-Echolot. Ziichtungskunde 32: 450—464.

Lindhls, B. 1971. Tulesborapport nr. 1. Svinskötsel 61, 1: 14 16.
—» & Persson, J. 1972. Stationsindex för galtar. SHS., Medd. 54, 25 p.
-» - & Sundgren, P.-E. 1969. Ekolodindex. SHS., Medd. 29, 32 p.
Lodge, G. A., Cundy, M. E., Cooke, R. & Lewis, D. 1972. Influence of energy and protein

concentration in the diet on the performance of growing pigs. 2. Differing nutrient
density at a constant energy: protein ratio. Anim. Prod. 14: 47 55.

Lounamaa, P. 1968. Katsaus Suomen Sianjalostusyhdistyksen toimintaan vuosina 1949 1968.
[A survey of the programme of the Finnish Pig Breeding Society, 1949—1968]. Sika
196: 19-45.

Lush, J. L. 1947. Family merit and individual merit as bases for selection. Part I. Amer. Nat.
81: 241-261.

Maijala, K & Vainikainen, V. 1962. Über die Ursachen der Variation in der Dicke des Riic-
kenspeckes an der Versuchsstation fiir Schweinezucht. J. Scient. Agric. Soc. Finl. 34:
66-82.

Melrose, D. R. 1957. Beinschwäche bei Schweinen. Tierziichter 19: 322.
Mennerich, A. 1967. Untersuchungen zur Methode und Anwendung des Echolotmessver-

fahrens in der Leistungpriifung beim Schwein. Diss., 96 p. Göttingen.
Ming-Pi Mi, 1962. Genetic and environmental variation analysis program. Dept, of Genetics

& Dairy Sci., Univ. of Wisconsin. Mimeogr.
Minkema, D., Bekedam, M., Kroeske, D. & Stegenga, Th. 1964. A selection experiment

with pigs: comparison of the effectiveness of individual selection and sib selection. Neth.
J. Agric. Sci. 12: 318-329.

Moen, R. A. 1968. Seleksjonsindeks for ungräner. Husdyrforsoksmotet Norg. Landbr. Hogsk.,
1968: 226-230.

» 1972. Erfaringer fra norsk fenotypetest av ungräner. Scand. Agric. Res. Soc. Mimeogr-
-10 p. Kävlinge.

Muller-Haye, B. 1965. Die Anwendung von Ultraschallmessungen zur Abschätzung der
grobgeweblichen Zusammensetzung von Schlachthälften beim Schwein. Diss., 88 p.
Göttingen.

Nebe, H. 1969. Feno- og beinkvalitet hos griser. Norsk Landbruk. 14 —l5: B—9.

Nenonen, E. 1971. Parhaan regressioyhtälön valitsemisesta [On the selection of the best reg-
ression equation]. Thesis, 24 p. Helsinki.

» 1972. Maatalouden tutkimuskeskuksen tilastopakkaus [The statistics package of the
Agricultural Research Centre]. Tikkurila.

Otto, E. 1967 a. Verlauf der Zunahme des Riickenmuskels bei Schweinen im Gewicht von
40 bis 115 kg. Tierzucht 21: 402.

* 1967 b. Riickenspeckzunahme bei Schweinen im Gewicht von 40 bis 115 kg. Tierzucht
21: 602.

—» & Sieg, G. 1963. Ultraschallmessungen an lebenden Schweinen. Arch. Tierzucht 6:
470-480.

» & Sieg, G. 1965. Ultraschallmessung der Riickenspeckstärke und Ruckcnmuskelflache
bei Jungebern. Tierzucht 19: 26 30.

» Sieg, G. 1966. Ultraschallmessungen bei weiblichen und männlichen Tieren in Zucht-
betriebe. Tierzucht 20: 589 591.

Partanen, J. 1965. Sikakantakokeiden tulokset vuodelta 1964. Summary: Results of pig
progeny testing in the year 1964. Sika, 181; 25 51.



115

Pearson, A, M., Bratzler, L. J., Deans, R. J., Price, J, F., Hoffer, J. A., Reineke,
E. P. & Luecke, R. W. 1965 a. The use of specific gravity of certain untrimmed pork
cuts as a measure of carcass value. J. Anim. Sci. 15: 86—92.

» Bratzler, L. J., Hoffer, J. A., Price, J. F., Magee, W. T. & Deans, R. J. 1956 b.
The fat-lean ratio in the rough loin as a tool in evaluation of pork carcasses. J. Anim.
Sci. 15:896-901.

* Bratzler, L. J. & Magee, W. T. 1958 a. Some simple cut indices for predicting carcass
traits of swine. I. Cut-out and loin lean area. J. Anim. Sci. 17: 20 26.

—» Bratzler, L. J. & Magee, W. T. 1958 b. 11. Supplementary measures of leanness. J-
Anim. Sci. 17:27 33.

» Deans, R. J. & Bratzler, L. J. 1959. Some lumbar lean measures as related to swine
carcass cut-outs and loin eye area. J. Anim. Sci. 18: 1087 1093.

Pedersen, O. K. 1968. Sikojen teuraslaatu [The carcass quality of pigs]. Sika 196: 5 18.
Persson, J. 1972. Stationsprövning framtidsväg vid urval av elitgaltar. Husdjur 1972,

3: 26-27.
» & Lindhä, B. 1972. Individprövning av galtar pä station i Sverige. Scand. Agric.

Res. Soc. Mimeogr. 14 p. Kävlinge.
Peter, W. 1962. Untersuchungen fiber die Anwendung des Ultraschall-Echolotgerätes in der

Fleischschweinztichtung. Arch. Tierzucht 5: 385 390.
Plonka, S., Duniec, H. & Kostyra, T. 1969. Wykorzystanie przyzyciowych pomiarow

grubosci sloniny dla selekcji knurkow hodowlanych. Summary: The use of ultrasonic
equipment for measuring the thickness of back fat of live animals in breeding boars
selection. Acta Agr. & Silv. Ser. Zoot. 9; 31—40.

Price, J. F., Pfost, H. 8., Pearson, A. M. & Hall, C. W. 1958. Some observations on the
use of ultrasonic measurements for determining fatness and leanness in live animals.

J. Anim. Sci. 17: 1156.
» Pearson, A. M., Pfost, H. B. & Deans, R. J. 1960. Application of ultrasonic reflection

techniques in evaluating fatness and leannes in pigs. J. Anim. Sci. 19: 381—387.
Quijandria, 8., Jr. & Robison, O. W. 1971. Body weight and backfat deposition in swine:

curves and correction factors. J. Anim. Sci. 33:911—918.
Rittler, A. 1964. Untersuchungen fiber Echolotmessungen als Hilfsmittel ftir die Schätzung

des Zuchtwertes junger Eber. Diss., 48 p. Hohenheim.
» 1968 a. Echolotmessungen als Hilfsmittel ftir die Zuchtwertschatzung beim Schwein.

2. Mitteilung. Weitere Untersuchungen fiber die genetische Fundierung der Echolot-
messungen an Jungebern. Z. Tierz. Zficht.biol. 84: 179 187.

—* 1968 b, 3. Mitteilung. Abschätzung des Fleisch- und Fettanteiles auf Grund von Echo-
lotmessungen am lebenden Schwein. Z. Tierz. Zficht.biol. 85: 78 89.

* Schoen, P., Schelper, E. & Fewson, D. 1964. Zur Frage der Genauigkeit von Echolot-
Messungen am lebenden Schwein in verschiedenen Gewichtsabschnitten. ZUchtungs-
kunde 36: 159-168.

» Werkmeister, F., Fender, M. & Fewson, D. 1955. Untersuchungen fiber die Bezie-
hungen zwischen Hassen an der Schlachthälfte und dem Anted der fleisch- und fettreichen
Teilstficke beim Schwein. ZUchtungskunde 37: 216 223.

Robison, O. W. 1962. Growth and backfat deposition curves in swine. J. Anim. Sci. 21; 975.
» Cooksey, J. H., Chapman, A. B. & Self, H. L. 1960. Estimation of carcass merit of

swine from live animal measurements. J. Anim. Sci. 19: 1013 1023.
Scheffe, H. 1961. The analysis of variance. 477 p. New York.
Schierbaum, E. 1961. Untersuchungen fiber die Beziehungen zwischen Röckenspeckdicke und

Gesamtkörperfettgehalt. Diss., 71 p. Hannover.
Schmid, P. 1970. Ein Beitrag zur Eigen- und Geschwisterleistungsprfifung beim Schwein.

Diss., 106 p. Zfirich.
Schmidt, L. 1964. Untersuchungen fiber die Eigenleistungsprfifung bei Ebern. Bayer. Landw.

J.buch 41: 427-439.
Schmitten, F. 1967. Untersuchungen zur Schätzung phänotypischer und genetischer Parameter

fiir Merkmale der Mastleistung und des Schlachtkorperwertes beim Schwein unter be-



116

sonderer Beriicksichtigung der Fleischbeschaffenheit und ihrer Beziehung zur Schlacht-
körperzusammensetzung. Diss., 274 p. Bonn.

Schoen, P. J. 1962. Untersuchungen iiber die Eignung des Ultra-Schall-Verfahrens fur Mess-
ungen am lebenden Schwein zur Abschätzung des Fleisch- und Fettanteils. Diss., 55 p.
Hohenheim.

Scholz, U. 1965. Untersuchungen iiber die Eignung verschiedener Mast- und Schlachtleistung-
seigenschaften fiir die Zuchtwahl bei Schweinen. Diss., 88 p. Giessen.

Schulman, A. 1971. Jalkaheikkous [Leg weakness]. Sika 1971, 4: 25.

Schumm, H.-R., Otto, E., Sieg, G., Frederick, W. & Schaaf, A. 1966. Beitrag zur Eigenleis-
tungspriifung bei Schweinen. Arch. Tierzucht 9: 37 51,

Siers, D. G. & Thomson, G. M. 1972. Heritabilities and genetic correlations of carcass and
growth traits in swine. J. Anim. Sci. 35: 311—316.

Skjervold. H. 1962. Nytten av unggrismilingene i avlsarbeidet. Buskap og Avdrätt 14: 177
180.

* Indrebo, T. & odegArd, A. K. 1960. Registrering av ryggspekktykkelsen pä levende
svin. Meld. Norg. Landbr. Hagsk, 39, 4: 1 18.

SkArman, S. 1960. Erfarenheter vid användning av ultraljud för mätning av späcktjockleken
hos slaktsvin. Sv. Svinavelsför. Tidskr. 3: 38—43.

Smith, C. 1966. A note on the heritability of leg weakness scores in pigs. Anim. Prod. 8: 345
348,

Sokal, R. R. & Rohlf, F. J. 1969. Biometry, p. 253 298. San Francisco.
Standal, N. 1962. A scoring-system for pig selection based on performance testing for backfat

thickness and rate of gain. Meld. Norg. Landbr. Hagsk. 41, 3, 9 p.
* 1968. Muligheter till ä pävirke slaktekvaliteten hos svin ved avlsarbeide, Scand, Agric.

Res. Soc. Mimeogr. 9 p. Stockholm.
Steel, R. G. D. & Torrie, J. H. 1960. Principles and procedures of statistics. 481 p. New York.
Stouffer, J. R., Wallentine, M, V., Wellington, G. H. & Diekmann, A. 1961. Development

and application of ultrasonic methods for measuring fat thickness and rib-eye area in
cattle and hogs. J. Anim. Sci. 20: 759—767.

Sundgren, P.-E. 1964. Ultraljudtekniken och den praktiska svinaveln. Akt. Lantbr.högsk.
47: 40-44.

» 1965. Ekolodets använding vid urval av suggor och galtar. Akt. Lantbr.högsk. 75;
73-76.

» 1967. Individprövning av svin för tillväxt och fettansättning. Scand. Agric. Res. Soc.
XIII Congr., Mimeogr. 5 p. Copenhagen.

» 1969. Individpröving av svin med utnyttjande av ultraljudteknik. Mimeogr. 24 p.
Uituna.

» 1972. Erfarenheter av individpröving av galtar uppföddaunder fältförhällanden. Scand.
Agric. Res. Soc., Mimeogr. 6. p. Kävlinge.

Uusisalmi, U. 1969 a. Vorläufige Ergebnisse iiber das Messen der Schlachteigenschaften beim
Schwein. J. Scient. Agric. Soc. Finl. 41:50 59.

* 1969b. Carcass analysis as an aid in pig breeding. Acta Agr. Scand. 19; 214—220.
» 1971 a. The conventional carcass evaluation and the carcass dissection analysis of pigs.

J. Scient. Agric. Soc. Finl. 43:40 52.
» 1971 b. Evaluation of carcass quality of live pigs at weights of 50 kg and 90 kg. J. Scient.

Agric. Soc. Finl. 43: 111-127.
» 1971 c. Ham quality and its relationship to carcass quality. I. Ultrasonic and other

measurements. J. Scient. Agric. Soc. Finl, 43: 129—139.
* 1971 d. Inheritance in ham and its components in finnish Landrase and Yorkshire

breeds. Acta Agric. Scand. 21: 73 83.
* 1972. Simplified dissection as an aid in carcass evaluation on the Landrace and York-

shire breeds. J. Scient. Agric. Soc. Finl. 44: 1 11.
Vanschoubroek, F., Wilde, R. de & Lämpö, Ph. 1967. The quantitative effects of feed

restriction in fattening pigs on weight gain, efficiency of feed utilization and backfat
thickness. Anim. Prod. 9:67 74.



117

Varo, M. 1962. t)ber die Begrenzung der Beurteilungseigenschaften bei der Eberauslese. Ann.
Agric. Fenn. 1:267 283.

* 1971. Sikojen ultraäänikuvauksen tuloksia [Results of ultrasonography of pigs]. Sika
1971, 22: 5-7.

» Partanen, J. 1965. Bewertung der Heritabilität von einigen Eigenschaften bei Schwei-
nen. Ann. Agric. Fenn, 4: 46—48.

Vockert, P. 1969. Untersuchungen zur Möglichkeit des friihzeitigen Erkennens von Merkmalen
der Schlachtkörperzusammensetzung mit Hilfe des Ultraschallmessverfahrens am le-
benden Schwein. Diss., 100 p. Berlin.

Weiss, F.-K. 1967. Merkmale der Fleischbeschaffenheit beim Schwein, ihre Erblichkeit und
ihre Beziehungen zu Mastleistung und Schlachtkorperwert. Diss., 98 p. Gottingen.

Weniger, J. H., Glodek, P., Mennerich, A. & Schmidt, L. 1967. Untersuchungen zur Met-
hode und Anwendung des Echolotverfahrens in der Leistungspriifung beim Schwein.
Bayer. Landw. J.buch 44: 842 874.

Wright, U. K. 1967. Combined performance and progeny testing in Northern Ireland- a prog-
ress report. Bth Study Meet. Europ. Assoc. Anim. Prod. Mimeogr. 8. p. As.

Wussow, W. & Grosse, F. 1959. Fettgehalt ganzer Schweine und Mastleistungen, ihre gegen-
seitige Beeinflussung und ihre Eignung als Berechnungsgrundlage fiir die physiologische
Futterverwertung. Arch. Tierzucht 2:427 438.

Wohlbier, W., Friesecke, H. & Kirchgessner, M. 1961. t)ber die fruhzeitige Beurteilung
einigcr Masteigenschaften wachsender Schweine. Ziichtungskunde 33:488 492.

Zagozen, F. & Flock, D. 1970. Genetische Aspekte der Mastleistung beim Schwein in ver-
schiedenen Gewichtsabschnitten. Ziichtungskunde 42:43 52.

» Schröder, J. 1970. Leistungsfutteranteil je kg Gewichtszunahme als Hilfsmerkmal
zur Schätzung der Schlachtkörperzusammensetzungbei der Eigenleistungspriifung beim
Schwein. Ziichtungskunde 42: 53 66.

Zobriskv, S. E., Brady, D. E., Easley, J. F. & Weaver, E. A. 1959. Significant relationships
in pork carcass evaluation. IE Measurements and cuts of fat as criteria for live hog value.
J. Anim. Sci. 18:583-588.



118

SELOSTUS

Karjujen koeasematestauksen tuloksellisuudesta

Elsi Ettala

Ketieläinten jalostustieteen laitos, Helsingin Yliopisto ')

Tutkimuksen tarkoituksena on ollut selvittää, voidaanko voimakkaasti valikoidun karju-
aineksen jalostusarvoerotsaada karjujen yksilökokeissa luotettavasti esille. Tutkimusmateriaali-
na on ollut kaksi 30 karjun testausryhmää (A ja B), joista kummastakin on valittu jälkeläis-
arvosteluun testauspisteiden puolesta 5 parasta, 5 keskinkertaista ja 5 heikointa karjua. Karju-
jen poistot jalkavikojen ja hedelmällisyyshäiriöiden vuoksi ovat jossain määrin vähentäneet
sekä jälkeläisarvosteltujen karjujen lukua että valintaryhmien välisiä eroja. Jälkeläisarvostelu
on suoritettu 26 karjulle. Jälkeläisiä on ollut 441.

1. Jälkeläisarvostelu on osoittanut, että karjujen silavaluotaus on antanut varsin luotetta-
van arvion karjujen lihakkuutta periyttävästä kyvystä. Tämä on ilmennyt seuraavasti;

a. Ruho-ominaisuuksiltaan heikoin valintaryhmä on kyetty erottamaan tilastollisella luo-
tettavuudella (P < 0.05 —P < 0.001) muista ryhmistä, vaikka testauskarjujen silavanpaksuus-
erot ovat olleet varsin vähäisiä (äärimmäisten ryhmien keskimääräiset silavanpaksuudet ovat
eronneet toisistaan A-karjuilla 3.0 mm ja B-karjuilla 5.9 mm). Paras ja keskinkertainen valinta-
ryhmä eivät ole eronneet merkitsevästi toisistaan. Näiden silavanpaksuuserot ovat olleet erit-
täin pienet (keskimäärin A-karjuilla 1.1 mm ja B-karjuilla 1.2 mm).

b. Isien silavaluotaustulosten (säkä, keskiselkä, pakara ja kylki) keskiarvo on korreloitunut
erittäin merkitsevästi jälkeläisten keskimääräisiin ruhon paloittelutuloksiin sekä merkitsevästi
tai lähes merkitsevästi jälkeläisten teurasruhosta mitattuihin silava-arvoihin. Sen sijaan
isien ja jälkeläisten luotaustulosten välillä on ollut hyvin heikko yhteys (B-ryhmä). Yhteyden
kiinteys on riippunut siitä luotettavuudesta, millä jälkeläisten ruhot on arvioitu.

c. Isien luotaustulosten ja jälkeläisten teurasruhosta määritettyjen keskimääräisten silava-
mittojen välisestä isä-jälkeläisregressiosta laskettu silavan paksuuden h2 -arvio (0.44) on ollut
yhtä hyvä kuin jälkeläisten puolisisarkorrelaatiosta määritetty (0.42 ± 0.18). Yksityisten silava-
mittojen kohdalla on kuitenkin ollut vaihteluja.

2. Karjujen kasvunopeuden ja rehunkäyttökyvyn periytymisestä ja vaikutuksesta jalostus-
valintaan on todettu seuraavaa;

a. Isien päiväkasvun ja jälkeläisten päiväkasvun sekä kummankin ryhmän rehunkäyttö-
kyvyn välillä on ollut tilastollisesti luotettava yhteys (P < 0.05). Isä-jälkeläisregressiosta
laskettu päiväkasvun h2-arvio on ollut jonkin verran heikompi (0.32) kuin jälkeläisten
puolisisarkorrelaatiosta määritetty (0.49 ± 0.18).

b. Kasvunopeus ja rehunkäyttökyky ovat jääneet jalostusvalinnassa kokonaan huomioon
ottamatta. Tämä on johtunut valintaperusteina käytettyjen testauspisteiden luonteesta. Pis-
teet ovat muodostuneet miltei pelkästään silavan paksuuden mukaan. Kasvunopeus ja re-
hunkäyttökyky ovat vaikuttaneet niihin hyvin vähän. Kyseinen testauspistejärjestelmä ei
sen tähden sovellu käytettäväksi koeasematestauksessa.

3. Yksilökokeen tuloksiin ovat merkitsevästi vaikuttaneet seuraavat ulkoiset tekijät:
a. Eläinten ruokintataso ja rehun laatu ovat vaikuttaneet merkitsevästi kasvunopeuteen ja

silavan paksuuteen.

2) Nykyinen osoite: Maatalouden tutkimuskeskus, Kotieläinhoidon tutkimuslaitos, Tik-
kurila.



119

b. Alkupainon jaalkuiän vaihtelut ovat vaikuttaneet merkitsevästi kasvunopeuteen jarehun-
käyttökykyyn. Voimakkaimmin ne ovat vaikuttaneet ikäpäivien luvulla tietynpainoisena il-
maistuun kasvunopeuteen.

c. Luotauspainon vaikutus silavan paksuuteen on ollut erittäin merkitsevä silloin, kun
painot ovat vaihdelleet huomattavasti. Sen sijaan painon vaikutus on ollut vähäinen, jos luo-
taus on suoritettu viikkopunnitusten yhteydessä mahdollisimman lähellä 88 kg:n elopainoa.
Luotausiän vaikutus on ollut pienempi kuin painon.

4. Liikuntahäiriöistäkärsineitä eläimiä on ollut sekä testauskarjuista että jälkeläiskarjuista
yli 30 %. Testauskarjuista on jalkavikojen vuoksi poistettu 13 %. Jälkeläisissä ilmenneet
jalkaviat ovat olleet pääasiassa joko perinnöllisiä tai tapaturmaisesti syntyneitä. Ruokinnan
vaikutus jalkojen kestävyyteen on vaikuttanut vähäiseltä.

5. Koeasematestauksen kehittämiseksi tehdyissä selvityksissä on todettu;

a. Karjujen päiväkasvu on tilastollisella luotettavuudella (P < 0.001) voitu määrittää jo
kahden kuukauden kuluttua kokeen alkamisesta karjujen ollessa n. 60 kg:n painoisia. Tämä te-
kee mahdolliseksi karjujen esikarsinnan varhaiskasvun (20 —6O kg tai 20 —7O kg) perusteella.
Esikarsinnassa voitaisiin ottaa huomioon myös jalkojen heikkous, koska varhaistestauksessa
annettujen jaikapisteiden h 2 -arviot on todettu tilastollisesti luotettaviksi.

b. Jälkeläiskarjujen mittausten ja ruhon paloittelutulosten perusteella on todettu, että var-
min arvio karjun lihakkuudesta on saatu, kun silavaluotaus on suoritettu säältä, keskiselältä,
pakaralta ja kyljeltä (soi). Kinkun ympärysmitta on antanut merkitsevän lisäselityksen ruhon
lihakkuuden vaihteluista. Teurasruhomittauksista on longissimus rfomi-lihaksen poikkileik-
kauspinta-ala ollut paras lihakkuusmuuntelun selittäjä. Merkitsevän lisäselityksen on anta-
nut myös kylkipituus.

c. Leikkosisarten ruhon paloittelutulokset ovat antaneet tilastollisesti erittäin merkitsevän
lisäselityksen karjun ruhon arvon vaihteluista (R2 %: 3.2 —9.8 %-yksikköä).