I

’I
II

II
II

I

I

II

I

I

I

I

I

I
II

II

I I

I

III

I

I
I

I

I
I

I
I
I

II

I

II

II

N I

I

°°“ II
(JO—x

I
.0300“)

 

'—I

ma INFLUENCE o: BREED AND OTHER
FACTORS ON ms PROLIFICACY,
LIVABILITY. AND wemmc WEIGHTS

0F SWINE

Thesis for the Degree of M. S.
MICHIGAN STATE COLLEGE
CImr'Ies Laine NickeI

19M

1115515

This is to certify that the

thesis entitled
The Influence of Breed and Other

Factors on the Prolificaqy. Livability
and Weaning Weights of 8 wine

presented by
Charles Nickel

has been accepted towards fulfillment
of the requirements for

ii..— degree in Wabandry

£2. £4ng

Major professor

Date, W 179 1351

 

 

 

 

THE INFLUENCE OF BREED AND OTHER FACTORS ON THE
PROLIFICACY, LIVABILITY, AND WEANIKG WEIGHTS OF SWINE

By
Charles Leslie Eigggl

A Thesis
Submitted to the School of Graduate Studies of
Nflehigan State College of Agriculture and
Applied Science in partial fulfillment

of the requirements for the degree
of

MASTER OF SCIENCE

Department of Animal Husbandry
1951

THESIS

 

ACKNOWLEDGMENT

The writer desires to express his appreciation
for the assistance rendered by Dr. Ronald H. Nelson,
Head of the Department of Animal Husbandry, for his
continued interest and suggestions during the course
of’this work.

He is also greatly indebted to mr. Harry F.
Moxley, Swine Extension Specialist of the Department of
Animal Husbandry for permission to carry on the study
and to use the materials contained therein.

A word of appreciation is extended to Dr. William
D. Baten, Research Professor of Statistics for his help

in statistical procedures.

ilf)i>rifjfi

TABLE OF CONTENTS

PAGE
INTRODUCTION 1
OBJECT OF THE STUDY 4
REVIEW OF LITERATURE 5
NETHODS OF PROCEDURE 1?
RESULTS AND DISCUSSION 22
CONCLUSIONS 56
BIBLIOGRAPHY 58

APPENDIX 62

Table

Table

Table

Table
Table

Table
Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

INDEX TO FIGURES AND TABLES

I.

II.

III.

IV.
IV-A

V.
V-A.

VI.

VI-A.

VII.

Breed Prolificacy Differences
from Review of Literature

Breed mortality Differences
from Review of Literature

Conversion Table of 56 Day Wean-
ing Weights.

Breed Prolificacy of Sows

Analysis of variance of Breed
Prolifieacy of Sows

Breed Prolificacy of Gilts

Analysis of variance of Breed
Prolificacy of Gilts

Breed Influence on Number of
Pigs Reared by Sows

Analysis of Variance of Pigs
Reared by Sows

Breed Influence on Number of Pigs
Reared by Gilts

VII-A. Analysis of Variance of Pigs

VIII.

Reared by Gilts.

Effect of Breed on 56 Day Weaning
Weights (Sows)

VIII-A.Analysis of Variance of 56 Day

IX.

IX-A.

X.

XI.

XI-A.

Weaning Weights (Sows)

Effect of Breed on 56 Day Weaning
Weights (Giltfi)

Analysis of Variance of 56 Day
Weaning Weights (Gilts)

Comparative Performance of Bows
and Gilts

Analysis of Variance of Individual
Weaning Weight Between Four Breeds

Mean and Standard Error of Individual
Weaning Weights

PAGE

17
24

24
25

25

28

28

29

29

31

31

32

32

37

37

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

XII.

XIII.

XIII-A.

XIV.

XIV-A.

XV.

XV-A.

XVI.

XVI-A0

XVII.

XVII‘A.

XVIII.

XVIII-A.

XIX.

XIX'Ae

XX.

Analysis of Covariance of the
Effects of Letter Size on In-
dividual Weaning Weights

MOnth of Farrow and Survival
Rate at 8 Weeks

Season of Farrow and Survival
Rate at 8 Weeks

Spring and Fall Variations In
Litter Weight.

Analysis of Variance of Seasonal
Variation in Litter Weight

Yearly Averages of the Prolif-
icacy of Sows

Analysis of Variance of Yearly
Differences in Prolificacy of
Sows

Yearly Averages of the Prolif-
icacy of Gilts

Analysis of Variance of Yearly
Differences in Prolificacy of
Gilts

Number of Pigs Reared by Sows
in the Different Years

Analysis of Variance of Yearly
Differences in Number of Pigs
Reared by Sows

Number of Pigs Reared by Gilts
in the Different Years

Analysis of Variance of Yearly
Difference in Number of Pigs
Reared by Gilts

Weaning Weights of Pigs Farrowed
by Sows in the Different Years

Analysis of Variance of Yearly
Differences in Weaning Weights
of Pigs Farrowed by Sows

Weaning Weights of Pigs Farrowed
by Gilts in the Different Years

PAGE

38

42

42

43

45

45

45

46

46

47

47

48

48

49

49

50

Table XX-Ae

Table H10

Figure 1.

Figure 2.

Figure 3.

Analysis of Variance of Yearly
Differences in Weaning Weights
of Pigs Farrowed by Gilts

Comparison of Performance by
Various Methods of Crossbreeding

Frequency of Litter Size

A Study of the Number of Pigs
Weaned by the Number Farrowed

Yearly Effects on the Record
of Performance of Sows and
Gilts

PAGE

50

26

55

51

INTRODUCTION

There is nothing inherent in any system of breed-
ing which guarantees success or failure under all cir-
cumstances. The element of uncertainty which is con-
stantly overshadowing the outcome of any breeding program
requires the constant attention of the breeder for signs
of the stimulating rewards of success or the depressing
misery of failure.

Since 1900, considerable strides have been made
in animal breeding. This is probably most particularly
true in dairy cattle where individual variations in pro-
duction can be measured. In more recent years progeny
testing has appeared in other classes of livestock, in-
cluding swine.

The history of the sow testing program began in
this country in Minnesota in 1929, when purebred breeders
were invited to send individual animals to the experiment
station where rate and cost of gain as well as carcass
yield were studied. In 1954, the sow testing program
was expanded with the actual testing carried out at the
farms and was found to stimulate more careful selection
in swine herds. This type of project has since gained
much pepularity throughout the country and was introduced

into Michigan in 1945.

most of the purebred swine associations actively

- 2 -
promote pregeny testing and have agreed upon requirements
governing the tests. Although specific requirements vary
between breed associations, their common goal is to recog-
nize productive sows in order that their offspring may
be used as herd replacements.

This study involves the swine litters entered in
the Michigan Sow Testing Pregram and the relative merits
of the various breeds of swine as well as the crossbreds
entered therein. Inasmuch as the breeding of livestock
is a specialized business with the economic well being of
the Operator often dependent upon its success, it is well
for the participant to know all that he can about the
product which he is producing.

Experienced operators are well aware that a success-
ful swine enterprise can be based on good animals of any
improved breed and that greater variations often exist
between strains of one breed then between two different
breeds. However, essential facts concerning the perform-
ance and improvement of individual breeds cannot easily
be ignored and may well serve to direct the future progress
and comparative status of a breed.

The writer is fully cognizant of the fact that
much competition and rivalry exists between respective
breeds of swine, as in all breeds of livestock. However,
persons who contemplate the breeding of livestock should

endeavor to acquaint themselves with the present attributes

-3-
and future possibilities of a breed. With these facts
in mind it was felt that an Opportunity presented itself
in this study which could not be overlooked.

OBJECT OF THE STUDY

Although swine type has varied considerably in
this country in recent decades, desirable market type
swine have generally been develOped by selection and breed-
ing on the basis of pedigree and external appearance.

The conformation and quality of several leading breeds
have become remarkably similar due to uniform ideals held
by most swine authorities and by various breeders working
with different breeds. Although much agitation for a
leaner hog has been forthcoming in recent years, it may
be said in general, that satisfactory progress has been
made in selection according to external characteristics.

It has been only in recent years however, that
much study has been given to the factors which effect
the prolificacy and the ability Of sows to raise the
pigs farrowed by them. Although most authorities agree
that the commercial swine producer might better devote
his time to providing adequate rations and keeping his
herd healthy, there is great Opportunity in the swine
industry for increasing litter size and especially for
controlling the various factors that might increase
the size and number of pigs reaching weaning age. Some
Of these factors include the breed involved, the relative
merits of sows and gilts and perhaps the season of year
farrowed. It was the purpose of this study to find
the effects of breeds and to study the merits of sows

- 5 -
and gilts within each breed. Also included were the
effect of size of litter farrowed on the size of litter
weaned along with the effect of litter size on the in-
dividual pig weights at weaning time. Other factors
studied included, the effect of seasonal variation on
the number of pigs farrowed, weaned, and the weaning
weights of litters. The effects of various methods of
crossbreeding on the above mentioned factors were also
studied, although the data on this subject is somewhat
limited.

A study of this nature becomes complicated by the
fact that variations in feeding and management securing
between herds may mask or distort breed differences.
Ideally, each breed should be kept under controlled
conditions, but such a course is impossible under com-
mercial production. Consequently it is necessary to
deal with large samples selected at random which was the
method followed in this study.

REVIEW OF LITERATURE

Inasmuch as this paper was concerned with several
aspects of sow performance the review of literature was
divided into various subjects directly concerned with
the objects of this study.

In one of the earliest studies of breed differences
on prolificacy of sows, Bitting (2) in 1898 reviewed the
herd books of three pepular breeds and reported that
Chester Whites excelled both Berkshires and Poland Chinas
by .14 and 1.5 pigs per litter respectively. He reported
on 600 litters of Chester Whites, 400 litters of Berkshires
and 1,086 litters of Poland Chinas.

Christenen, £3 £1. (5) studied the records of
breed prolificacy and mortality of pigs at the North
Dakota Agricultural Experiment Station covering the period
from 1909 to 1926. The average litter size for 393 sows
was 9.6 pigs per litter farrowed. Yorkshires averaged
11.7 pigs per litter, Duroc Jerseys 10.7, Chester Whites
9.6, Berkshires 8.7, and Poland Chinas 8.2. No mention
was made as to the number of sows of each breed farrowing.
In the same study it was reported that the average weaning
percentage for all breeds was 69.8%. In the individual
breeds, Yorkshires weaned 74¢, Duroc Jerseys 67%, Chester
Whites 75¢, Berkshires 69% and Poland Chinas 65%.

Surface, (28) in 1909 reviewed the herd books of

- 7 -
Poland Chinas and Duroc Jerseys. 0n 54,515 litters of
Poland Chinas he found the average litter at birth to
be 7.4 pigs, while on the 21,652 litters of Duroc Jerseys
the average was 9.55 pigs farrowed per litter.

Rommel (25) also working on the herd books of the
Poland China and Duroc Jersey breeds, found that in the
Poland Chinas the average litter size during the preced-
ing 20 years had increased by 0.5, while there was no
change regarding the Durocs. The former breed had an
average litter size of 7.52 and the latter 9.26.

Wentworth and Aubel (52) in a study conducted in
1916 on 5,540 litters, stated that 5 centers of deviation
exist in swine fertility which may possibly correspond
to genetic factors involved in its inheritance. They
supported the contention that small litters were dominant
to large ones.

Lush (15) in a statistical study at Iowa State
College, found in the instructional herd (7 different
breeds maintained), that the breed differences account
for 5% of the variance in live pigs farrowed. Nearly 15%
of the remaining variance was caused by permanent individual
differences between sows within a breed. He further
estimated that it would require 10-20 years to increase
the average fertility by as much as one pig per litter.

Merrie and Johnson (20) analyzed 1,055 litters
taken at random from Poland China records made no mention

of average litter size but reported an increase in the

- a -

average litter size of one pig during the period 1900-1921.
One of the most extensive American studies on

breed differences which has been reported is that made
by Lush and Molln (14) in connection with their study of
inheritance in sows. Information was obtained on 7,296
litters from 2,560 different sows, covering the period
from 1920 to 1957. The data used was collected from ex-
periment station and college herds in eight states and
in herds maintained by the United States Bureau of Animal
Industry. A tabulation of pigs farrowed by breeds and

the number of breeds reported on follows.

Table I

 

Breed P.C. D.J. C.W. Hamp. York. Berk. Tam. Landrace
Average 7.98 9.78 9.55 8.66 10.75. 7.74 7.45 9.74
Number of 1851 5,557 852 267 194 485 218 114
Litters

These workers concluded that breed differences
were statistically significant and that a high degree of
consistency existed regardless of station.

The same workers reported on the number of pigs
weaned per litter and used data including 4,790 litters
farrowed from 1654 different sows. The following table
gives the breed, average number of pigs weaned per litter

and the number of litters of each respective breed studied.

- 9 -

 

Table II
Breed P.C. D.J. C.W. Hamp. Yerk. Berk. Tam. Landrace
Average
Litter 5.18 5.62 6.10 6.25 7.08 5.50 4.02 5.09
Size
Number of 1492 2,104 607 101 145 85 146 114
Igtters

 

Differences were found to be significantly differ-
ent between breeds.

Weaver and Bogart (51) reported on 117 sows that
weaned an average of 5.8 pigs. Only 24% of them raised
as many as eight pigs.

Lush (15) indicates that selection for the number
of pigs weaned would progress at about the same rate as
selection for number of pigs farrowed (one pig increase
every 10-20 years). He further states that about 9% of
the variance in numbers of pigs weaned is due to perman-
ent differences between sows. About 10% of the variance
in weaning weight is similarly caused. This work was
done on Iowa State College's instructional herd.

Many workers have studied the effects of age of
the sow on prolifieacy and mortality of pigs. Snyder (50),
in a study concerning 72 mature sows and 87 first litter
gilts, reported a farrowing average of 10.9 pigs per litter
for sows against 8.2 pigs per litter for gilts. The sows
weaned 6.56 pigs per litter while the gilts weaned 6.25

- 10 -
pigs per litter. Percentages of farrowed pigs weaned was
60% for the former and 76$ for the latter. Johansson,
as cited by Smith and Donald, (26) supports this in his
observation of 1,671 litters of Swedish Large White Pigs.
He indicated that first litter sows farrow fewer pigs
than older sows but that they suffer lower mortality.
Vestal (51) found yearling sows farrowing an average of
about two pigs less than aged sows, which differed little
regardless of age. He also showed that two year old
sows weaned the maximum number of pigs with the yearlings
weaning 1.62 pigs per litter less.

Lush (14) found that gilts averaged g pig smaller
litters at weaning then did the average of four of the
first six litters when consecutive records of the same
individual sows were kept. He also reports that the two
year old sow weaned é pig more than the average of the
first six litters. He concluded his study of effect of
age on prolifieacy with the observation that size of litter
rose slightly more than t pig from the first to the second
litter, then one more pig per litter at two years of age,
then varied only a little until the sows were four and
one half years old at which point a gradual decline was
observed.

Stewart (27) reported on the effects of age from
the records of 749 inbred Poland China and Minnesota.#1
gilts. He found that litter size increased with the age

I‘
‘\

- 11 -
of dam at farrowing. Gilts farrowing at 520 days of
age averaged one pig less and those farrowing at 410
days, % pig more than those at one year. He further re-
ported that on the average, gilts making the greatest
gains during gestation farrow the largest litter but
variations in gain may be an effect rather then a cause.
He concluded that age and weight tOgether account for
four p of the variance in size of first litters and pro-
vide the most reliable criteria for use in fertility
selection.

Young (56) reports the following records on 508
sows and 270 gilts. The sows farrowed an average of 9.0
pigs compared to an average of 7.1 pigs from gilts. Sows
lost 1/5 of their pigs prior to weaning compared to 1/4
lost by gilts. The average size of litter weaned by
sows was 6.0 pigs and by gilts to 5.5 pigs. No mention
was made as to the numbers studied.

McPhee (18) in determining the feasibility of using
litter size as a selection index, examined 589 litters.
He concluded that there is a significant correlation be-
tween a sow's first litter and her later litter but no
such correlation existed between the size of litter a
gilt was farrowed in and the size of litter produced by
her. In contrast Weaver and Bogart (52) concluded that
gilts from the higher producing dams make better sows
than did gilts from poorer producers. Their work with

- 12 -
65 selected dams was rather conclusive although no
statistical evidence was offered.

The review of literature on the effects of re-
spective breeds on weaning weights of litters is rather
limited. Lush and MOlln (l4) recognized that statistical-
ly significant differences did exist between breeds and
that the differences appear consistently in all herds
studied.

Russell and Hutton (24) studied the litter weights
of 40 mature sows and 87 gilts. They found the pigs of
mature sows to average 8 to 9 pounds heavier at 70 days
of age than the pigs of 87 gilts. Menzies-Kitchin, (19)
working with English breeds, found the litters of sows
to be somewhat heavier at six weeks of age than the gilt
litters. Smith and Donald (25) studying the records of
sows and gilts farrowing the same numbers of pigs report-
ed a slight difference in favor of the sows litters at
the end of eight weeks. Hostetler,'gt‘gl. (11) studied
167 individuals and indicated that no great difference
existed between weights of sows and gilts at weaning,
although the sows weaned slightly more pigs. EcMeekan
(17) also showed first litter gilts to wean litters con-
parable in weight to older sows.

The effect of litter size on the weaning weights
of pigs has been a controversial issue among workers in

the past.

- 13 -

Hostetler, 33 El. (11) indicated that the average
weaning weight per pig decreases as the number of pigs
in the litter increase. Menzie's-Kitchen (19) substantiates
this, stating there was no significant difference at 6
weeks between the average weights of pigs from litters
of different size at birth. He indicated that there is,
however, a difference in average weight according to the
number of pigs surviving at six weeks: the larger the
number the lighter the pig. Further, Johansson, as cited
by Smith and Donald (25) found a decrease in weight as
the litter size increases. Murray (21), however studied
the influence of size of litter on total litter weight
at eight weeks and found that the litter weight increases
with increase in litter size up to 12, after which the
weight decreases. Alexsson, as cited by Smith,.gtԤ1.
(26), calculated the correlation between the number of
pigs at birth and the litter weight and obtained
r 3-0.815 4 0.0322. The larger the average fertility,
the stronger was the negative correlation between litter
size and weight of individual animals at 5 weeks. Smith
and Donald (25) concluded that no general relationship
between weaning weight with respect to litter size existed.
Carroll and Krider (4) express the view that the number
of pigs nursing does not have a uniform influence on ‘
the average weaning weight of the pigs. They maintain
that the evidence is not clear that a pig in a small

- 14 -
litter is any better nourished up to weaning than a pig
in a reasonably large litter.

The effect of season or month of the year in which
sows are farrowed has come under the scrutiny of several
workers. Hestetler (11) states that more pigs were weaned
per litter in the fall months than spring although no
indication of the number of litters studies was advanced.
Young (36) in a study of 180 Indiana farms states that
the average death loss was 27% for fall pigs, while for
the spring pigs it was 32%. Davidson (6) reports the
most satisfactory months for farrowing two litters an-
nually in England and Northwest EurOpe are March and
September. He further states that litters farrowed in
June give the best results in numbers and weight at wean-
ing, while November litters were the poorest.

Vestal (31) reports March and September to have
the heaviest losses of suckling pigs. Wilcox, gt‘gl. (34)
report that in early farrowed pigs (before April 1) an
average of 68% were alive at weaning. On an equal number
of farms studied having pigs farrowed after April 1, 70%
survived till weaning. In grouping of pigs farrowed into
spring and fall litters the investigators found that only
66.2% of the spring pigs survived while fall litters
saved an average of 68.1% of pigs farrowed. Hepkins (10)
tabulated the effect of time of farrew on number of pigs

weaned for 3 years. The months studied were March, April

- 15 -

and May. The 3 year average for March was 63.7% saved,
for April 66.0% saved and for May 71.2% survived. Henzies-
Kitchen (19) investigated English records, where sows
are farrowed quite uniformly throughout the year, and
concluded that approximately one more pig survived per
litter during the summer than during the winter months.

The review of literature concerning the effect
of litter size at birth on the number of pigs weaned is
somewhat limited. Menzies—Kitchen states that there ap-
pears to be little advantage in producing more than 12
pigs at birth. In the case of more than 12 pigs the ad-
dition in number was more than offset by an increase in
death rate.

A considerable amount of work has been done on
the effects of crossbreeding on litter size and pig
mortality. Lush,‘gt‘gl. (15) in a study comparing cross-
bred and purebred brood sows and involving 108 litters
concluded that cross breds had a general superiority
over purebreds in the percentage of pigs living till
weaning. In weaning weights the crossbreds showed great-

er ant more consistent gains. The workers further state

that back cross and 3 breed crosses when sired by a pure-
bred bear compare favorably to the first-cross pigs. Mc-
Meekan (16) compared purebreds and crossbreds up to 56
days of age and found the crossbreds superior. In his
data 202 litters of purebred pigs of four breeds had a
death loss of 21.2% up to weaning as compared to 12.8%

- 15 -
mortality for the 65 litters of crossbreds. The average
birth weight of purebreds was slightly larger at birth
but smaller at weaning than crossbreds.

Winters, gt‘al., (35) at Minnesota reports first
cross, 3 breed crosses and back crosses all superior to
comparable purebreds. He further states that first cross
and back cross groups were approximately equal in superiority
to purebreds, but both were excelled by 5-breed crosses.

The cross litters averaged 1/5 pig too 2 pigs larger at
weaning. 0n the average each pig weighed from 5 to 7
pounds more than purebreds.

Carroll and Roberts (3) however, in reviewing
the literature of 50,000 animals concluded that hybrid
vigor cannot be expected in the majority of crosses be-

tween breeds of swine.

- 17 -

METHODS OF PROCEDURE

The data used in this study were obtained from
the records of the Michigan Sow Testing Project carried
on by the Animal Husbandry Extension Service of Michigan
State College.

This project has been in effect for six years,
from 1945 to date. In the six years 2296 litters with
19,825 pigs have been entered.' Rules governing this
project are simple and easily complied with. Each litter
must be ear marked and weighed at birth or shortly after
in the presence of an official witness. These weights
must be recorded with the Extension Office before the
pigs are four weeks of age. A second weighing of the
litter occurs at or near 56 days of age. If pigs are
not weighed exactly at 56 days, the weights are calculated
to the 56 day basis using the conversion table below. ‘

 

 

Table III

Days of Age Factor Days of Age Factor Days of Agg Factor
40 1.64 51 1.14 62 .87
41 1.58 52 1.11 65 .85
42 1.52 55 1.08 64 .84
43 1.46 54 1.05 65 .82
44 1.41 55 1.02 66 .80
45 .57 56 1.00 67 .79
46 1.52 57 .98 68 .77
47 1.28 58 .95 69 .76
48 1.24 59 .95 70 .75
49 1.21 60 .91 71 .75
50 1.17 61 .89 72 .72

 

- 13 -

Awards are given to litters falling into pre-
determined weight groups. The results of the project
were analyzed annually and circulars were distributed
among cooperators in the project to stimulate more ef-
ficient production.

A total of 2,052 litters were available for this
study. They were first grouped into straight breeds,
cross breds, three breed crosses, back crosses, four
breed crosses, and litters which were designated on entry
blanks as grades. The straight breeds were not necessari-
ly registered and should not be considered as such. Those.
litters falling into the straight breed classification
were then regrouped into their specific breeds with the
sows, second litter or above, being computed separately
from the gilts, (first litter). A study was then con-
ducted on breed prolificacy, number of pigs surviving
at 56 days, and the 56 day weaning weights.

The second study made involved the comparison of
performance by sows and gilts entered as straight breeds.
No conversion factor designed to equalize gilt and sow
litter weights was used in this study at any time.

A study of the effect of litter size on weaning
weights was conducted using only a small portion of the
available data. An analysis of covariance and correlation
analysis was calculated in this portion of the study.

Since the farrowing date of the litter was listed

- 19 -
on each entry blank it was thought that a report of
the numbers of litters farrowed by months would be of
interest both from the standpoint of the livability of
pigs in various seasons and to ascertain the papnlarity
of individual months for farrowing pigs by the farmers
of this state.

The number and percentage of pigs weaned accord-
ing to size of litter at birth was also considered in
this report.

The effects of various years reporting was carried
out to reveal if any year was superior and whether the
program was progressing towards its goal of "more pork
from fewer sows".

A brief study of the comparison of various cross-
breeding methods was conducted as a conclusion to this
paper.

A statistical analysis of the data was made using
the formula shown on page 18. An analysis of variance
was calculated between the sows of each breed and the
gilts of each breed for prolificacy, livability and wean-
ing weights. An analysis was also carried out on the
effects of season on the weaning weights of litters. The
last analysis of variance was carried to study the ef-
fect of years on prolificacy, livability, and weaning
weights of both sows and gilts of the straight breeds

only. Other data are presented in table form.

- 20 _

Inasmuch as the analysis of variance was used ex-
tensively in this work a simplified explanation of its
use is included. .

In analyzing data certain results are obtained
which are distinctly different from other results gather-
ed, the obJect being to compare the two groups. The
heterogeneity of the variation is the factor which is
being tested, and the degree of its expression deter-
mines the significance of the findings of the experi-
ment. Therefore in studies of variation it is necessary
to be able to differentiate the variation according to
causes or groups of causes, especially where such dif-
ferentiation is an essential part of the analysis of the
results.

The analysis of variance supplies the mechanism
for this procedure and in addition supplies the results
in a form to which tests of significance can be applied.

- 21 -
Formulae Used In Statistical Analysis

Analysis of Variance:

 

 

8X2 -(§§E = Total Sum of Squares (29)
n
S S 8
(—-—9‘2" " Cfi """ C? " 0' L = Sum of Squares Be-
n1 n2 n tween Breeds (29)

Correlation Analysis:

b =‘§§y I Regression Coefficient (29)
3:2
r = 55! a Correlation Coefficient (29)

 

“(8:53) \|(Sy2)

 

d} = l - r2 = Standard error of Corre-
q n _ 2 lation Coefficient (12)

 

 

 

/ n2 - (ma/$115
e n , 2 ‘__ ' Standard error of Re-
gression Coefficient (51)

3x2

 

Standard deviation

\J szln-l

IAF;—_—

Standard error of mean (1)

RESULTS AND DISCUSSION
Breed as a Factor in Size of Litter Farrowed

The comparison of the various breeds in respect
to prolificacy in this study was, of necessity, broken
down into sows and gilts. It is quite evident that a
breed having a majority of sow records would enjoy a
distinct advantage over other breeds with a majority of
gilt entries. The average litter size farrowed by the
nine breeds studied of both sows and gilts are reported
in Tables IV and V. Statistical treatment of the data
compiled between breeds reveals that there is highly
significant difference as shown in Tables IV-A and VeA.
This is in agreement with Lush and Mblln (14) who found
significant differences existing in the prolificacy of
various breeds in their inheritance study.

The exceptional prolificacy of the Yerkshire breed
reported in this data is substantiated by Christensen,
‘gt'§;., who found the average Yorkshire litter size to
be 11.7. The prolificacy table of Lush and Mblln (14)
also indicates that the Yorkshire and Duroc breeds are
to be highly recommended in respect to prolificacy. A
'partial explanation of the Yorkshire prolificacy might
be found on the basis of work by Zeller and Hetzer (87)
on the effect of type on production efficiency. Work-
ing within a single breed, they concluded that sows

- 23 -
classified as large type were generally superior to
those of intermediate or small type. The Yorkshire
breed as a whole could be classified as "large" type.

This is however, only an Opinion of the writer.

- 24 -
Table IV
Breed Prolificacy of Sows

 

Number or: Number of:

 

 

Breed Pigs Sows Ave. S.E
Daron 3094: 290 10. 66’ c142
Hampshire 2328 229 10.172 .088
Chester White 1143 105 10.881 .223
Spotted P. C. 934 91 10.262 .170
Poland China 582 59 9.861 .275
Yorkshire 617 49 12.593 .406
Berkshire 523 52 10.063 .252
OIC 312 28 11.142 .650
Minnesota #1 35 4 8.752 2.79
Table IV-A

Analysis of Variance of Breed Prolificacy of Scws*

 

 

Source Of Variation Dc Fe Sc Sc Me Sc F
Total soc 4,961 **
Between Breeds 8 403 50.38 9.92
Within Breeds 898 4,558 5.08

 

s:
A pendix A
**H§gh1y Significant

Calculations to determine the percentage of the total
variance between and within breeds were as follows:

 

_ Number of Sows ~ = m of Mean
K " Number arm ”“1 mm” 5,331.. at Between
Breeds and within
breeds.
Between breeds variance 3 A I KB 8 50.38 '?
A I 101B = 50.38 - 5.08/
101B = 45.30
B = .45
45 5 55.46 s 8.1% of the total
variance
100$ - 8.1% = 91.9% of the total
variance

From the above calculations it is apparent that differences
between breeds are responsible for approximately 8.1% of
the total variance.

Differences within breeds or between gilts of the same
breed are responsible for 91.9% of the variance.

Table V

Breed Prolificacy of Gilts

 

Breed

Number of Number of

 

 

 

Pigs Sows Average r 8.3.
Duroc 3,853 402 9.58 z .096
Hampshire 1,820 199 9.15 t .129
Chester White 985 108 9.12 t .233
Spotted P.C. 581 64 9.08 2 .195
Poland China 478 57 8.39 2 .211
Yorkshire 649 59 10.64 t .272
Berkshire 209 24 8.71 ft .252
010 262 29 9.03 t. .384
Minnesota #1 23 3 7.67 t .913

Table V-A

Analysis of Variance of Breed Prolificacy of Gilts*

 

Source of Variation

 

13.1». 3.5. 111.8. E
Total 944 3,660

Between Breeds 8 11.50 c.oz**
Within Breeds 936 3,568 3.81

 

*Appendix B

we
Highly Significant

Calculation of the estimate of variance shows that dif-

ferences between breeds accounted for approximately 7.3fi

of variance.

Differences within breeds or between gilts

of the same breed accounted for approximately 92.7%.

_ 25 -
F I G U R E I

#00 . r'“'

300 -'

Number of Litters

 

100

 

 

 

 

 

 

 

 

 

 

 

 

1 2 3 h ’5 6 7 8 9 10 ll 12 13 14 15 16 17 18

Number in Litter

L I T T E R F R E Q U E N C Y

Average litter size at birth: 9.96

-27...

Breed Influence on Number of Pigs Weaned

The term prolificacy is measured more directly
by the number of pigs born alive than by the number
reared. The former measure more nearly approaches the
true fertility of a breed, inasmuch as the number of
pigs weaned is probably influenced to a greater extent
by management and other environmental factors than by
the capacity of survival inherited from the sow.

Sows and gilts were again studied separately.
Results are shown in Table VI and VI-A. The analysis
of variance on this data, shown in Table VII and VII-A,
revealed that significant differences at the 1% level
existed. The same breeds which excelled in number of
pigs farrowed were superior in number of pigs reared

per litter.

- 23 -
Table VI

Breed Influence on Number of Pigs Reared by Sows

 

 

 

Number or *Number ofI Standardfi
Breed Pigs Sows Average Error
Duroc 2,591 290 8.93 i .089
Hampshire 2,044 229 8.93 i. .168
Chester White 980 105 9.33 .2 .122
SPOtted PeCe 782 9.1 8e59 1 e111
Poland China 496 59 8.41 t .201
Yorkshire 492 49 10. O6 1' . 247
Berkshire 453 52 8. 71 1‘ . 214
010 284 28 10.14 1: .238
Minnesota #1 33 4 8.25 i: .240
Table VI-A

Analysis of Variance of Pigs Reared by Scws’

 

 

Source of Variation D.F. S.S. M.S. F
Total 906 3,585
Between Breeds 8 155 19.38 5.07**
Within Breeds 898 3,430 3.82

 

I"Appendix C
**Hign1y Significant

Breed differences accounted for 4% of the total variance.

Differences within breeds or between sows of the same

breed accounted for 96% of the variance.

- 29 -
Table VII
Breed Influence on Number of Pigs Reared by Gilts

 

 

 

Number of Number of Standard
Breed Pigs Gilts Average Error
Duroc 3346 402 8.33 t .092
Hampshire 1676 199 8.42 t .104
Chester White 902 108 8.35 t .168
SPCtted PeCe 521 64 8.14 1 e243
Poland China 414 57 7.26 i .258
Yorkshire 558 59 9.46 t .258
Berkshire 194 24 8.08 4 .264
010 238 29 8.21 =-.395
Minnesota #1 l9 3 6.33 1:.664

 

Table VII-A
Analysis of Variance of Pigs Reared by Gilts**

 

 

Source of Variation D.F. S.S. M.S. F
Total 944 3284 **
Between Breeds 8 104 13.00 3.82
Within Breeds 936 3180 3.40

 

*Appendix D
**Hign1y Significant

The breakdown of the total variance shows that 3% was
due to breed differences while differences within breeds

or between gilts of the same breed accounted for 97¢.

Effects of Breed on Weaning Weights of Pigs

All litter weights were adjusted to exactly eight
weeks after birth. The sows were again calculated separate-
ly from the gilt averages. Management plays an important
role in the weaning weights of litters reported in this
study because of the doubtless attempts to encourage high
litter weight through supplemental feeding. Hewever
there is no reason to assume that any one breed has been
pushed more than any other.

Table VIII shows the average litter weight by breeds
of sows. Table VIII-A, the analysis of variance of wean-
ing weights, reveals that significant differences to
exist between breeds. This is in accordance with Lush
and MOlln (14) who reported significance differences on
56 day litter weights.

Tables II and IX-A show the effect of breeds on
the litter weight of gilts.

Table VIII
Effect of Breed on 56 day Weaning Weights (Sows)

 

 

Total Standardi
Breed Weight Number Average Error
Duroc 82,829 290 285.53 4 9.05
Hampshire 69,061 229 301.58 I: 6.28
Chester White 32,889 105 313.23 1: 6.09
Spotted P.C. 24.914 91 273.78 1:15.25
Poland China 18,583 59 314.96 -ll.00
Yorkshire 17,348 49 354.04 1:14.64
Berkshire 14,909 52 286.71 1: 9.38
0IC 10,141 28 362.17 1116.54
Minnesota #1 1,158 4 289.50 i=36.67

 

Table VIII-A
Analysis of Variance of 56 day Weaning Weights (Sows)*

 

 

”80.1.99 01’ variat1°l DeFe . SeSe Megs Fe
Total 906 12,999,815 . **
Between Breeds 8 415,605 51,961 3.71
Within Breeds 898 12,584,210 14,013

 

I"Appendix E
"Highly Significant

Effect of breed on the 56 day weaning weights of litters
farrowed by sows accounted for 2.6% of the total variance.
The remaining 97.4% of the variance is found within

breeds or between the sows of the same breed.

Table 11
Effect of Breed on 56 day Weaning Weights (Gilts)

 

 

 

Total Standard‘
Breed Weight Number Average Error
Duroc 109,257 402 271.78 1 3.95
Hampshire 49,943 199 250.97 t 5.66
Chester White 27,101 108 250.94 t 6.93
Spotted P.C. 16,814 64 262.71 t 9.74
Poland China 14,027 57 246.09 i 10.41
Yorkshire 16,389 59 277.77 i: 5.62
Berkshire 5,698 24 237.42 1:13.41
01C 8,481 29 292.45 i=16.18
Minnesota #1 759 3 253.00 ir27.10
Table IX-A

Analysis of Variance of 56 day Weaning Weights (Gilts)III

 

 

Source of Variation D.F. 8.3. M.S. E
Total 944 4,312,428 **
Between Breeds 8 464,361 58,045 14.12
Within Breeds 936 3,848,067 4,111

 

*Appendix F
**Highiy Significant

Breed effect accounted for 12% of the variance en the
weaning weights of gilts. 88% of the variance was caused

by differences within breeds or between gilts of the same

breed.

p?
A
-..~3-

Comparative Performance of Sows and Gilts

The comparative performance of sows and gilts on
the basis of prolificacy, rearing ability, and weaning
weight of litter is summarized in Table I. This data
includes 907 sows and 945 gilts.

Table I

Comparative Performance of Sows and Gilts

 

 

 

Afarrow Weaning Percentage Average
Average Average Raised Litter Wt.
SOWB 10.55 8.99 85.23 311.74

Gilts 9.38 8.33 88.80 264.77

 

The results listed in this table are generally in
accordance with those obtained in other studies. The
sows farrowed 1.17 more pi‘a per litter than did the gilts.
Usually an efficient producer eliminates those gilts
from his herd whose initial litter is below average,
thus the sows remaining in the herd have been retained
on a partial selection basis for prolificacy. Another
theory concerning prolificacy of sows and gilts is that
a larger number of eggs are ovulated by a sow during
the estrus period than in gilts. However, this is compli-
cated somewhat by the plane of nutrition of the individual

animals at breeding.

- 54 -

The percentage of pigs raised of those farrowed
indicates that the advantage distinctly rests with the
gilts. This is in line with other reports as previously
mentioned in the review of literature. There are many
apparent reasons why this is true.

Figure 2 shows that the advantage of larger litters
is partially offset by increased death loss. Usually
extremely large litters are handicapped by the presence
of small, weak pigs at birth which JeOpardize there chances
of survival. Inasmuch as most gilts are smaller than
mature sows, the possibility of pig losses through crush-
ing, pinning and general carelessness on the part of the
mother is somewhat smaller. Wilcox (34).gt'§1., report
that 44% of all losses of suckling pigs was attributed
to crushing. .

The heavier litter weights at weaning recorded by
sows is substantiated by Russel and Hutton (24) and
Menzies-Kitchen (19) who found slightly heavier litter
weights in sows than gilts at siX'weeks. This may be
explained in part of the increased number of pigs weaned
per litter but this point is open to debate. The added
maturity of the sow over the gilt may also allow for
increase milk production by sows, consequently result-

ing in higher litter weights.

Percentage surviving at 8 weeks

94

92
9O
88
86
84

s 82

80
78
76
74
72
7O
68

66
64

- 62

60
58
56

52
50
48

- s5 -

FIGURE 2

A STUDY OF THE NUMBER OF PIGS WEANBD BY THE NUMBER FARHOWED

 

No. of pigs born

alive and number 0
* surviving at

ei ht
3 weeks i

No. of pigs born
alive and percentage
* surviving at eight

. weeks i

 

 

 

A L

2 3 4 5 6 7 8 9 10 ll 12 13 14 .15 l6 l7 18 19

 

No. of pigs born alive per litter

The number of pigs raised to eight weeks of age increased
up to 17 in this study. Hewever the advantage of larger
litters is mostly offset by the increased death losses in
the larger litters.

(Based on 2,114 litters)

«ll

«10

D]

No. of pigs surviving at 8 weeks

- 36 -
Effect of Litter Size on Individual Weaning Weights

A study of the effect of litter size on individual
weaning weight at 56 days of age was carried out for the
purpose of finding the relationship between litter size
and weaning weight of individual pigs.

In order that the environmental factor be kept as
neglible as possible, four herds were selected from the
available data, each herd representing a different breed.
Although these litters were farrowed in different years
it is assumed that the treatment within each herd accord-
ed the sows and their litters by the operator was consist-
ent from year to year.

An analysis of variance was calculated between
the four herds to see if breed alone might induce sig-
nificant differences on the weight of litter weaned.
Results of the analysis of variance are shown in Table

II.

- 57 -
Table.XI
Analysis of Variance of Individual Weaning Weight Between

Four Breeds*

 

 

Source of Variation D.F. 8.3. M.S. E
Total 61 3885
Between Breeds 3 172 57.33 .895

Within Breeds 58 3713 64.02

 

*
Appendix G

Table XI-A
Mean and Standard Error of Individual Weaning Weights

 

 

Number off Méan ‘ Standard’
Breed Litters Weight- Error
Spotted P.C. 15 34.67 2.26
Duroc Jersey 22 35.68 2.35
Chester White 10 32.00 1.12
Hampshire 15 35.00 1.26

 

Since no significance was found, breed apparently
had no effect upon litter weight in this portion of the
study.

To obtain the effect of litter size upon pig
weight using the combined results of the four herds,
and analysis of covariance was calculated. Results are

shown in Table XII.

r :1
CO

Table XII
Analysis of Covariance of the Lfiects of Letter Size on

. . W . , . *
IndiVidual weaning weights

 

 

 

 

.1. . . q - I.)
Source of Variation D.F. SK“ Snx Si“
Total 61 256 54 3885
Between Herds 3 2 -40 172
Within Herds 58 230 94 3713

*

Appendix G

From the above table the correlation coefficient

was calculated as follows:

a. r =' va

{(59:23) (3Y3)

r : 94 94 - 94__ = 0.10

J(230)(571s) 4852990 924

 

 

 

b. Standard error of correlation coefficient LES
calculated from the following formula:

1 - r2 = .99 = .99 a 1.27

m

in - 2 {66 7.74

Inasmuch as the error term exceeds the correlation coef-

 

ficient it was concluded that little correlation existed
between these two variables. It would have taken a sample
approximately seven times as large as the one used in this
study to give a significant value.

These results disagree with those obtained by Alexson,
as cited by Smith and Donald (25) who found a negative cor-

relation of r = -O.815 é 0.0722.

I

C' C!
LO
I

(3
(3‘

Although little ceiielation was tained the re-

F4,
0
i"
t-J
O
CO

gression coefficient was (QlCul"te as

 

8.. b : Slag—f : o 94‘ = 04.].
O
x“ h.0

 

 

 

 

 

 

b. The Stzndard Er1or of Regre ession Coefiicient:
SXZ- bey 3713 — .'.54
I”: "' Ca 6 O .
e a 812 = 2? = 4.27 = 0.519

These results again show that the error term exceeds
the regression coefficient and cannot be considered a
reliable measure of the effect of litter size on weaning
weight of the individual pig.

Tie inte rp1etation of+ :e results obtained in this
studv indie tes the a small CC; elation may exist between

size of litter and the weight of individual pig 8 at 56

Is in the four herds studied. Hammond, as cited by

pl

(4

a
Smith and Donald, (85 he 3 pointed out that sows which are
ry fer ile usuall" have a good milk supply. This may

mean simply that as the litter size increases, the number
of teats used and the total quantity 0f1111k pr oduced rise.

This view is supported by Car1oll and Kiider,

- 4o -
(3) who state that the amount of milk produced by a sow
is closely related to the number of pigs which she suckles.
Smith and Donald (25) advance the theory that both fertility
and milk yield are closely related with the function of
the pituitary gland, and it may well be that large litters
and heavy milking are associated because of the possession
of an active pituitary.

Further interpretation of such studies and this

study in particular, might proceed along the line of thought
that there is a special feeding of sows with large litters
with the production of heavy litters as a specific goal.
The supplemental feeding of the pigs themselves to produce
extraordinarily high 56 day weights would, of course,
give a bias picture of this study. More instructive re-
sults could be obtained by raising large and small litters
on sows of known high fertility and sows of known low
fertility and recording 3-week weights which would eliminate
the effects of supplemental pig feeding to a large extent.

Effect of Season on Size of Litter

Season, or month of farrow also appears to exercise
some influence on pre-weaning mortality. Results given
in Table XIII show little seasonal variattén in the number
of pigs born alive. Not a great deal more variation is
encountered in pigs surviving at eight weeks. Hewever,

this variation would undoubtedly appear larger if equal

- 41 _
numbers were farrowed in both seasons. The survival per-
centage reported here does not agree with studies of
other workers, most of whom report fall litters excelling

spring litters in the number of pigs raised to weaning.

Effect of Season on Litter Weight

The effect of season on weight of litter was
studied and is summarized in Table XIV. Analysis of
variance of the data is given in Table XIV-A. It was
found that a significant difference existed between the
seasons. This can be explained to a large extent by weath-
er eonditions. Management methods might also improve
the weights of fall farrowed litters. Inadequate housing,
dry lot confinement, etc., might effect late fall farrowed
litters. Hewever, it is undeniable that early spring

litters face the same problems in this state.

- 42 -

Table XIII

Month of Farrow and Survival Rate at 8 Weeks

 

 

Jane Feb 0 Mgr. Apr. May June
Pigs farrowed
per litter 11.50 10.18 9.96 10.31 9.95 9.83
Pigs weaned
per litter 9.87 8.23 8.65 9.50 8.67 8.96
Survival rate
(percentage) 85.87 80.84 86.86 92.04 87.13 91.15
No. of litters 8 103 1,028 632 203 46

 

Table XIII-A

Season of Farrow and Survival Rate at 8 Weeks

 

SPRING (m. ,

FALL (Aug. .

 

Apr., may) Sept., Oct.)

Pigs farrowed per

litter 10.08 10.04
Pigs weaned per

litter 8.94 8.86
Survival Rate

(percentage) 88.69 87.34

No. of litters 1863 127

 

 

 

11y Aug. Sept. Oct. Nov. Dec. Total
.85 9.43 10.40 10.64 10.84 10. 00 10.09
.55 8.63 8.95 9.00 8.60 9.33 8.99
.80 91.11 86.05 84.59 79.14 93.33 88.19
20 40 62 25 15 3 2185

 

Table XIV
Spring and Fall Variations in Litter Weight

 

 

 

 

Total weight Number off
per month Litters Average
Spring
Mar. 289,192 1,028 281.32
Apr. 185,859 632 294.08
May 61,482 203 286.09
Total 536,533 1,863 287.99
Fall
Aug. 8,275 40 206.88
Sept. 18,859 62 300.66
Oct. 7,220 25 288.80
Total 34,136 127 268.79
Table XIV—A

Analysis of Variance of Seasonal Variation in Litter Weight*

 

 

Source of Variation D.F. S.S. M.S. F
Total 1989 17,982,532 **
Between Seasons 1 43,858 43,858 4.86
Within Seasons 1988 17,938,674 9,023

 

*Appendix H
"Highly Significant

- 44 -
Effect of Year on Prolificacy, Livability and Weaning
Weight.

One of the purposes of a progeny testing program
is to stimulate efficiency of production. This is
especially true in swine where so many variables enter
into efficient production. In an effort to find out
whether more efficient production was developing, a
study of yearly differences was made.

This study also assists in clarifying breed dif-
ferences which have already been reported. It is mani-
fest that if any one year excelled others and one breed
recorded many entries in that particular year, their
overall average might give a biased picture of the breeds
actual performance.

The yearly performance of sows and gilts was
calculated separately. Analysis of variances were run
on each table to test the statistical significance of
the differences. Table headings on the following pages

are self-explanatory.

- 45 -
Table XV
Yearly Averages of the Prolificacy of Sows

 

1945 1946 1947 1948 1949 1950 Total

 

Number of
pigs
farrowed 1,011 1,796 1,023 1,320 2,385 2,025 9,560
Number of
Sows 102 185 98 121 214 187 907
Average 9.91 9.71 10.43 10.91 11.14 10.82 10.54

 

Table XV-A

Analysis of Variance of Yearly Differences in Prolificacy
a:

 

 

of Sows
Source of Variation D.F. S.S. M.S. F
Total 906 1915
Between Years 5 279 55.80 30.66**
Within Years 901 1636 1.82

 

*
Appendix J
**Hign1y Significant

Table XVI
Yearly Averages of the Prolificacy of Gilts

 

1945 1946 1947 1948 1949 1950 Total

 

Number of
pigs
farrowed 710 1,064 962 1,467 2,249 2,415 8,867
Number of
gilts 79 116 104 158 235 253 945
Average 8.99 9.17 9.25 9.28 9.57 9.55 9.38

 

 

 

Table XVI-A
Analysis of Variance of Yearly Differences in Prolificacy
of Gilts*
Source of Variation D.F. S.S. M.S. F
Total 944 3490
Between Years 5 38 7.60 2.07
Within Years 939 3452 3.68

 

'"Appendix J.

-47..

Table XVII

Number of Pigs Reared by Sows in the Different Years

 

 

 

1945 1946 1947 1948 1949 1950 Total
Number of
pigs
weaned 881 1,553 868 1,076 1,963 1,730 8,071
Number of
Sows 102 185 98 121 214 187 907
Average 8.64 8.39 8.86 8.89 9.17 9.25 8.89
Table XVII-A

Analysis of Variance of Yearly Differences in Number of

Pigs Reared by Sows*

 

 

Source of Variation D.F. S.S. M.S. F
Total 906 4,478
Between Years 5 93 18.60 5.82”
Within Years 901 4,385 4.87

 

*Appendix K
**High1y Significant

- 43 -

Table XVIII
Number of Pigs Reared by Gilts in the Different Years

 

1945 1946 1947 1948 1949 1950 Total

 

Number of
pigs weaned 597 944 804 1,321 1,989 2,129 7,784
Number of

Gilts 79 116 104 158 235 253 945
Average 7.56 8014 7073 8.36 8046 8042 8.24

 

Table XVIII-A
Analysis of Variance of Yearly Differences in Number of
Pigs Reared by Gilts‘

 

 

 

Source of Variation D.F. S.S. M.S. F
Total 944 4,448
Between Years 5 89 17.80 3.84"”‘
Within Years 939 4,359 4.64
I“Appendix K

*‘Highiy Significant

-49...

Table XIX
Weaning Weights of Pigs Farrowed by Sows in the Different

Years

 

1945 1946 1947 1948 1949 1950 Total

 

Total Weight
of Litters 30,404 54,173 29,647 39,195 69,108 60,211 282,748

Numb er of
Litters 102 105 98 121 214 187 907
Average 298.07 292.83 302.53 323.93 322.93 321.98 311.73

 

Table XIXpA
Analysis of Variance of Yearly Differences in Weaning'
Weights of Pigs Farrowed by Sowe*

 

 

 

Source of Variation D.F. 8.8. M.S. F
Total 906 12,117,225
Between Years 5 668,269 133,654 10.52**
Within Years 901 11,448,956 12,707
1:
Appendix I

**Higniy Significant

- 5O -

Table XX
Weaning Weights of Pigs Farrowed by Gilts in the Different

Years

 

1945 1946 1947 1948 1949 1950 Total

 

Total

Weight of

Litters 20,892 30,436 24,271 43,086 64,451 67,064 250,200
Number of

Litters 79 116 104 158 235 253 265

Average 264.46 262.38 233.37 272.70 274.26 265.08 264.76

 

Table XX-A
Analysis of Variance of Yearly Differences in Weaning
Weights of Pigs Farrowed by Gilts*

 

 

Source Of variatiol DeFe SOS. Mes. F
Total 944 9,023,183
Between Years 5 330,861 166,172 714**
Within Years 959 8,701,322 9,267

 

allAppendix L
**Hign1y Significant

- 51 -
Figure 3 -
YEARLY EFFECTS ON THE RECORD OF PERFOREANCE OF SOWS

AND GILTS

  
 
  

 

11 mean 10.84_ __________ _____ ~_
10 --------- pigs farrowed
9_¥ean___§a§9-___ _______,,._
pigs weaned
8

 

 

1945 1946. 1947 1948 1949 1950

YEARLY AVERAGES OF FARROWING AND WEANING NUMBERS (SONS)

 

fl

pigs farrowed

 

pigs weaned

 

 

 

1945 1946 1947 1948 1949 1950

YEARLY AVERAGES OF FARROWING AND WEANING NUMBERS (GILTS)

   
 
   
 

 

 

 

mean Weight 313
wo_____-_______. —————————— (mm)
275 - - ——~
Mean Weight 265 7“““‘
r—_————-——-—-—— ——————————— (giltfl)
250
225

 

1945 1946 1947 1948 1949 1950

Yearly Averages of Weaning Weights of Bows and Gilts

-52..

A study of the graphs in Figure 3 reveals that,
with minor exceptions, the trend covering the six years
has been upward. There has been a tendency, during the
past three years, for the number of pigs farrowed to
level out. This may indicate that prolificacy was reach-
ing a maximum or practical level in this project. This
might in turn lead to a similar level in weaning numbers
and weaning weights. It may be said in general, since
the trend is upward, that the project over the six years
has proved successful.

Since only one sharp drop appeared on the entire
graph, it may be concluded that no particular breed suf-
fered severe yearly effects to their overall average.

The sharp drop of weaning weights of the litters farrowed
by gilts in 1947 may have had a slight effect on the weights

of the more popular breeds.

Effects of Various Crossbreeding in Study

Cressbreds, as interpreted in this study, are the
progeny resulting from the matings of different breeds.
Probably the commoneSt use of crossbreeding, at least in
the animal kingdom, has been with swine. Crossbred animals
usually exhibit an increase in vigor and growth over either

of their purebred parents. The exact cause of this occur-

- 55 -
ence, which is known as heterosis, or hybrid vigor, is
not clearly understood. Some authorities suggest that
the inhibiting genes, which are carried in the germ plan:
with desirable genes, are neutralized by genes from the
other species or breed.

There were several methods of crossbreeding re-
ported in these data. They include (1) first cross pigs,
which were produced by crossing a purebred sow with a
purebred boar of another breed, (2) three-breed cross
pigs, produced by breeding first cross gilts or sows to
a purebred boar of a third breed, (3) backcross pigs,
produced by breeding first-cross gilts or sows to a pure-
bred boar of one of the two original parent stocks and
(4) four-breed cross, by breeding a first cross cow to a
first cross boar.

Although many experiments have shown that cross-
bred pigs are somewhat superior in the feed-let to pure-
bred pigs, it has generally been accepted that they should
not be used for breeding purposes. However, in the recent
Minnesota experiment reported by Winter, 31.21., (35)
the crossbreds excelled the purebreds in prolificacy,
rearing ability and weaning weights.

Although a rather small number of the various
crossbreds were entered in this project the results of
their performance were tabulated and are compared with

the purebreds in Table XXI.

— 54 -

Table XXI

Comparison of Performance by Various iethots of Crossbreeding

 

 

1 first- C—breed bacb- 4:5reed
rurebreds cross cross cross cross
Farrowing
average Cf SOWS 10054 100:2 10067 10075 11.30
Weaning
average of sows 9.00 9.13 9.51 9.50 11.00
Litter weight
of sows 711.74 343.4 301.9 751.0 391.0
Number of sow
litters studied 907 104 T: 16 3
F8 .Hr1ox. ing
8V6TO“G of gilts 9.78 9. 3 9.51 9.50 11.00
Weaning average
of gilts 8."? 8.42 9.40 8.56 9.80
Litter weight
of gilts 2 4.77 276.1 322.5 300.7 255.0
Number of gilt
litters studied 945 56 23 16 3

Interpretation of the above results is complicated
somewhat by the inequ8lity of numbers. Tge 4-breed cross
litters could be discounted ent re:ly because of their ex—
treme minority. Comparing the flD reoreds ard first cross
litters, the advantage in prolificacy distinctly rests
with the purebreds for both sets and gilts, altimi ugh more
first cross pigs were raiScd to weaning. The S-breed cross
and back cross litters had superior records to purebreds
and first crosses. This is substantiated in part b; Winters
33 81., (36) who reports that F—breed crosses aVeraged 1/3

to 2 pigs l8rgor at weaning.

- 55 -

Although crossbreeding in swine is a common prac-
tice it should be remembered that the haphazard crossing
of various breeds may quickly result in a nondescript

swine herd which would be uneconomieal and certainly

without pride to the owner.

l.

2.

3.

4.

5.

6.

SUMMARY AND CONCLUSIONS

A study was made of 1852 litters of purebred swine
representing nine breeds on the prolificacy, livabili—
ty and 56 day weaning weights of each breed.

The study on prolificacy revealed that some differences
did occur between breeds in this respect, and that
these differences were statistically significant.

They varied from a low of 7.67 to a high of 12.59

pigs per litter.

The number of pigs weaned per litter by the various
breeds differed. This difference was statistically
significant. The same breeds which excelled in pro-
lificacy raised the most pigs to weaning. They
varied from a low of 6.23 to a high of 10.14 pigs

per litter.

Breed differences existed in weaning weights which
were statistically significant. The range in this
study was from 362 pounds to 237 pounds per litter.
The comparative performance of sows and gilts showed
that sows farrowed 1.17 pigs more per litter than
gilts. However gilts weaned 88.80 per cent of their
pigs while sows weaned 85.23 per cent. The sow litters
weighed 53 pounds more per litter than did gilt lit-
ters at eight weeks.

The correlation coefficient between litter size and

7.

8.

9.

10.

- 57 -
weight of individual pigs at weaning was found to
be 0.10 f .0165. The regression coefficient was
found to be 0.41 f 0.519. This data was calculated
on 62 litters from four different herds.
Size of litters farrowed in spring and fall were
very similar and 88.69% of the spring pigs survived
till weaning compared with 87.54% of the fall pigs.
Spring farrowed litters were 19.2 pounds heavier at
weaning than were the fall litters. These differ-
ences were statistically significant.
Studies of the effect of the yearly differences on
prolificacy, livability and weaning weights were found
to be statistically significant with one exception.
Improvement in all three factors over a period of
six years was noted.
A study of the effects of crossbreeding revealed that
baokcrosses and 3-brced crosses were superior to
first-crosses and purebreds in prolificacy and number
of pigs weaned. These averages are unreliable because

of the small number of crossbred litters studied.

l.

2.

3.

4.

5.

6.

7.

8.

9.

100'

ll.

BIBLIOGRAPHY

Eaten, We Do
1938 Elementary Mathematical Statistics. First
editibn. John Wiley and Sons, Inc., New
York. 338 pp.

 

Bitting, A. w.
1898 The Fecundit of Swine. Tenth Annual Report
'Ef—tHe Indiana ricuItural Experiment
Station. 42-46.

Carroll, W. E. and Roberts, E.
1942 Crossbreeding In Swine. Illinois Agricultural
‘Experiment Station BuIletin. 489.

 

and. Krider, J. In
I950 Swine Production. First edition. McGraw-
HIII Book Company Inc., New York. 498 pp.

 

Christensen, F. W., Thompson, 0.A., and Jorgenson, L.

1926 Prolificac of Sows and MOrtality of Pi 5.
North Dako¥a71gficulturaI'Experimefit a ion

Bulletin 194. p. 83.
Davidson, H. R.
1948 The Production and Marketin of P1 8.
First edition. Iongsmans, ErEEn and Company,
London. 413 pp.

 

Ferrin, E. F.
1932 Production Tests for the Selection of Breed-
ing Hags. Proceedings of the American Society
25 Animal Production. '1933. I34-IB7.

 

 

Goulden, C. H.
1939 Methods of Statistical Anal'sis. First
edItIon,_70hnFWiley and Sons, nc., New York.

 

Henke ’ L. A.
1935 Is Fecundity in Swine Inherited. Journal
.3; Heredity. 26. 455-56.

Hapkins, J. A., Jr.
1928 An Economic Study of the Hog Enterprise
in Humbolt County. Iowa Agricultural Ex-
periment Station Bulletin. 225.

Hostetler, E. H., Nance, R. E. and Foster, J. E.
1930 Cost of Raising Pigs to Weaning Age. North
Carolina Agricultural Experiment Station
Bulletin. 272.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

- 59 -

Lindquist, E. F. '
1940 Statistical Anal sis in Educational
Research. FIFEtiEdItI5n. Houghton
n Company, Boston. 210 pp.

 

 

Lush, J. L., Anderson, A. L., Culbertson, C. C. and
Hammond, W. E.
1934 The Reliability of Some Measures of Pro-
duction in Brood Sows. Procedures 2; 152
American Society 2; Animal Production:
Igzzo 28 " O

 

, and Molln, A. E.
1942 Litter Size and Weight as Permanent Char-
acteristics of Sows. United States Depart-
ment of Agriculture Technical Bulletin 836.

, Shearer, J. L. and Culbertson, C. C.
Crossbreeding Hugs for Pork Production.
Iowa State College Agricultural Experiment
Station Bulletin. 380.

McMeekan, C. P.
1941 Growth and DeveIOpment of the Pig. Journal
‘3; Agricultural Science: 31. pp. l7-3I.

 

1956 Critical Study of Important Factors in
Successful Swinekeeping. New Zealand Journal
2; Agriculture: 52. 278-289.

 

chhee, H. C.
1931 Size of Litter as a Selection Index in
Swine. Proceedings 2; the American Socict
9; Animal Production: mm - 61'." .__z

Menzies-Kitchen, A. W.
1937 Fertility, Mbrtality, and Growth Rate in
Pigs. Journal‘gf Agricultural Science: 27
611-624.

Morris, H. P. and Johnson, D. W.
1932 Effects of Nutrition and Heredity upon
Litter Size in Swine and Rate. Journal 25
Agricultural Research: 32. 511-521.

 

hurray’ G. N.
1934 A Statistical Analysis of Growth and Carcass
Measurements of Baconers. Underste oort
Journal 2; Veterinagy Science: 2. 301-360.

 

Rice, Vc Ac
1942 Breedin and Im rovement of Farm Animals.
r itifin. e raw- ‘IIBbok‘Company,
Inc., New YOrk. 750 pp.

23.

24.

25.

26.

27.

28.

29.

31.

_ 50 _

Rommel, Go Mo
1906 The Fecundity of Poland-China and Duroc-
Jersey Sows. United States Department of
Agriculture Circular 95. 12 pp.

Russell, E. Z. and Hutton, R. E.
1938 Comparative values of mature Sows and Gilts
for Producing Market Hogs. United States
Department of Agriculture Circular 472.

Smith, A. Do Bo, and Donald, H. P.
1937 Weaning Weights of Pigs and Litter Sampling
of Pigs. Journal 2; Agricultural Science:
21. 485-562.

 

, Robinson, 0. J., and Bryant, D. M.
1935 The Genetic of the Pi . First Reprint.
Reprinted fr3fi'tHE liographia Genetics
VII. The Hague, Martinus Nijhoff. 160.

 

Stewart, H. A.
1945 An Appraisal of Factors Affecting Prolificacy
in Swine. Journal 23 Animal Science: 3.
250-259.

Surface, F. M.
1909 Fecundity In Swine. Biometricka: 9,
433-436.

 

Snedecor, George w.
1946 Statistical Methods. Fourth edition. Iowa
State ColIegeIPress, Ames, Iowa. 485 pp.

 

Snyder, W. P.
1915 Pork Production. Nebraska Agricultural
Experiment Station Bulletin. 147.

Venkatachalam, Ganapathy
1949 Estimates of Heritability of Birth Weight
and Weaning Weight of Lambs. Michigan
State College, Ph. D. Thesis. 69 pp.

veataj- ’ C. M.
1936 Feeding and management Experiments with
Brood Sows and Litters. Indiana Agricultural
Experiment Station Bulletin. 413.

Weaver, L. A. and Bogart, Ralph
1943 Some Factors Influencing Efficient Production
of Sows. Missouri Agricultural Experiment
Station Bulletin. 461.

36.

-61...

Wentworth, E. N. and Aubel, C. E.
1916 Inheritance of Fertility in Swine.
Journal of Agricultural Research: Q.
mm?

Wilcox, R. H., Carroll, W. E., and Harnung, T. G.
1933 Some Important Factors Effecting Costs in
Hog Production. Illinois Agricultural
Experiment Station Bulletin. 390.

Winters, In Me, Kiser, 00 Mo, Jordan. P. So and.
Peters, W- H.
1935 A Six Years' Study of Crossbreeding Swine.
Minnesota Agricultural Experiment Station
Bulletin. 320.

Young, G. E.
1930 The Two—Litter HOg System on Indiana Farms.
Indiana Agricultural Experiment Station
Bulletin. 338.

Zeller, J. H. and Hetzer, H. 0.
1944 Influence of Type of Hog on Production
Efficiency. United States Department of
Agriculture Circular. No. 698.

-62-

Appendix A

Analysis of Variance of Litter

Size of Sows

 

 

 

 

Breed 3x2 N Ex
Duroc 34,663 290 3,094
Hampshire 27,734 229 2,328
Chester White 13,037 105 1,143
Spotted P. C. 9,824 91 934
Poland China 6,000 59 582
Yorkshire 8,183 49 617
Berkshire 5,429 52 523
0.1.0. 3,362 28 312
Minnesota #1 418 4 3

105,650 907 9,569
c. r. = (9,569)3 = 100,954
Total $3 = EX2 - c.r. = 4,961
SS Between Breeds = 100,852 - 100,449 = 403

-63..

Appendix B
Analysis of Variance of Litter Size of Gilts

 

 

 

Breed EXZ N EX
Duroc 38,423 402 3,853
Hampshire 17,306 199 1,820
Chester White 9,614 108 985
Spotted P. C. 5,429 64 581
Poland China 4,148 57 478
Yorkshire 7,328 59 646
Berkshire 1,855 24 209
0.1.0. 2,488 29 262
Minnesota #1 181 3 23

86,672 945 8,857

 

c. T. = {885722 = 83,188

943
Total s§ = 86,672 - 83,188 = 3,484

Between Breeds = 83,104 - 83,102 - 92

-64-

Appendix C

Analysis of Variance of Litter Size at Weaning (Sows)

 

 

 

Breed EXZ N EX
Duroc 23,787 290 2591
Hampshire 19,718 229 2044
Chester White 9,310 105 980
Spotted P. 0. 6,821 91 782
Poland China 4,308 59 496

- Yorkshire 5,084 49 492
Berkshire 4,067 52 453
0.1.0. 2,924 28 284
Minnesota #1 279 4 33

76,298 907 8121

 

- 2 .
Go To - 181212 I 72,713

Total 83 a 3x2 - c.r. = 3,585

Between Breeds SS = 72868 - 72713 8 155

Analysis of Variance of Litter Size at Weaning (Gilts)

-65..

Appendix D

 

2

 

 

 

Breed EX N EX
Duroc 29,203 402 3,346
Hampshire 14,541 199 1,676
Chester White 7,862 108 902
Spotted P. C. 4,479 64 521
Poland China 2,770 57 414
Yorkshire 5,500 59 558
Berkshire 1,608 24 194
0.I.C. 2,080 29 238
Minnesota #1 _f 119 3 19

68,062 945 7,824
c. r. = (782423 = 64,778
945
Total es s Exz - C.T. = 3,284

Between Breeds SS

= 64,882 - 64,778 = 104

-66..

Appendix E
Breed as a Factor on 56 Day Weaning‘Weights (Sows)

 

2

 

 

 

 

Breed EX N EX
Duroc 30,491,295 290 82,829
Hampshire 22,884,911 229 69,061
Chester White 10,707,432 105 32,889
Spotted P. 0. 8,727,522 91 24,914
Poland China 6,267,171 59 18,583
Yorkshire 6,646,378 49 17,348
Berkshire 4,503,282 52 14,909
0.1.0. 3,889,721 28 10,141
Minnesota.#1 351,382 4 1,158

94,469,094 907 271,832
0. T. = 1271,8821? = 81,469,279
907
Total 88 = 2x2 - c.T. = 12,999,815

Between Breed SS = 81,884,884 — 81,469,279 = 415,605

-67...

Appendix F

Breed as a Factor on 56 Day Weaning Weights (Gilts)

 

 

 

 

 

Breed 2x2 N Bx
Duroc 32,207,739 402 109,257
Hampshire 18,794,428 199 49,943
Chester White 7,361,109 108 27,101
Spotted r. 0. 4,799,824 64 16,814
Poland China 8,797,669 57 14,027
Yorkshire 4,660,683 59 16,389
Berkshire 1,452,204 24 5,698
0.1.0. 2,698,891 29 8,481
Minnesota #1 194,949 3 759

70,956,250 945 250,955
0. T. = $250,355)3 = 66,648,822
945
Total 88 s 212 - c.T. = 4,312,428

SS Between Breeds = 464,361

Appendix 0'
Analysis of Variance and Covariance of Effect of Litter

Size on Weaning Weight

 

Spotted P. C. Duroc ‘Hsmpsfiire Chester White

 

 

 

 

 

 

 

 

 

 

'Pigs Ave. pigs Ive. pigs Ave. Pigs Ave.
weaned pig weaned’ pig weaned pig weaned pig
per wt. per wt. per wt. per wt.
litter lbs. litter lbs. litter lbs. litter lbs.
I Y X Y I V X 7
10 ME 6 E7 9 31 9 BR
11 ho 7 27 7 51 10 29
9 A6 3 12 58 12 29
9 2 h? 15 59 ll 32
5 2 7 17 11 5 10 - 25
2 58 9 a: 10 5 9 57
6 2 11 0 10 57 10
8 2 10 ho 3 5h 8 5%
6 29 g 27 25 8 55
11 20 19 8 35 12 51
8 51 10 50 8 h
8 g 8 55 11 5%
9 K 8 25 lo 5
12 ho 6 l9 9 5
9 hl 12 2 9 3
10 3%
ll
10 h5
10 A7
10 h
a El
9 52
Sum 125 520 190 796 1DH 525 99 ‘512
8x2 1155 1720 1h20 999
8y2 19096 50990 18715 ' 9860
Sxy h358 . 6882 , 5075—9 5096
Correction terms: Sums 5f!Sq ares:
For x (556)2 62:.h, 86 Sx 52 2-h986-256
For y (2155) /62= 7:2. 7 3,2 78 roe-7477145885
- FOI' XV (556)(2155)/ 2=19.507 SKY 19555-19507=5Lt

Between Herds:

(1%;13 / .. ........... .(9gi; : C.T.

- 59 -

Appendix H

Computation of Suns of Squares and Products characterizing
the regression of the pig weights on the number of pigs

weaned per litter.

Breed Weights = 552023 x 579622 4 $52522l§312)2-C.T.=172
22 15 lo

 

 

15
_ 2 2 2 2 -
Pig Numbers - 123 1(190) l(144) I 99) - 4986 - 26
"2ET—"—IIT"
Products = 5L520)(123)l(796)(190)/L525)(l44) {13121L991
15 22 15 10

- 19,808 : 4o

_ 70 -

Appendix I
Analysis of Variance of Seasonal Variation in Litter Weight

 

 

 

812 I N 81
Spring 170,799,552 1868 536,533
Fall 10,882,788 127 34,136
Total 181,682,885 1990 570,669

 

c.T. . _L570,669)2 = 163,649,803
1990

Total 88 = EX2 - c.T. = 17,982,582

SS Between Breeds = 163,693,661 - 163,649,803 = 43,858

-71..

Appendix J
Analysis of Variance of Yearly Effects on Prolificacy of

 

 

 

Sows

Year 512 N 81
1945 10,303 102 1011
1946 18,228 185 1796
1947 11,101 98 1028
1948 14,981 121 1820
1949 24,958 214 2385
1950 23,114 187 2025

102,680 907 9560

c.T. = 9 560 2 = 100,765
1‘49571‘

Total 58 e E12 - c.T. = 1,915
SS Between Years = 101,044 - 100,765 I 279
Analysis of Variance of Yearly Effects on Prolificacy of

 

 

 

Gilts

Y3 81' m2 EX
1945 6,672 79 710
1946 10,252 116 1064
1947 9,862 104 962
1948 14,801 158 1467
1949 22,271 235 2249
1950 23,829 253 2415

88,687 945 8867

 

c.T. . (886722 = 88,197 88 Between Years =
945
83,235-83,197 a 88

Total $8 = 8x3 - c.T. a 8,490

-72-

Appendix K
Analysis of Variance of Yearly Effect on Number of Pigs
Reared by Sows

 

 

 

 

Year E12 N EX

1945 7,865 102 881

1946 14,396 185 1558

1947 7,964 98 868

1948 10,046 121 1076

1949 19,481 214 1968

1950 16,546 187 1780
76,298 907 8071

c.T. = (8071)2 = 71,820 88 Between Years =

Total SS = EX2-0.T. = 4,478 71,918 - 71,820 = 98

Analysis of Variance of Yearly Effect on Number of Pigs
Reared by Gilts

 

 

 

Year EXZ n 81
1945 5,427 79 597
1946 8,058 116 944
1947 6,692 104 804
1948 11,712 158 1821
1949 17,961 285 1989
1950 18,715 258 2129

‘68,565—7 945 :7784
c.T. = {778422 . 64,117 88 Between Years =

64,206-64,117 89

Total 88 2 E12 - C.T. = 4,448

- 73 -

Appendix L
Analysis of Variance of Yearly Effect on Weaning Weights
of Pigs Farrowed by Sows

 

 

 

 

Year EXz N EX
1945 9,641,016 102 80,404
1946 16,957,114 185 54,178
1947 9,671,141 98 29,647
1948 18,678,112 121 89,108
1949 26,475,522 214 69,108
1950 20,810,259 187 60,211
97,228,164 907 282,748

c.T. = (282E748)2 - 85,110,989 Between years 88 =

85,779,208 -

Total 88 = EX2 - C.T. = 12,117,225 85,110,989-668,269

Analysis of Variance of Yearly Effects on Weaning Weights
of Pigs Farrowed by_Gilts

 

 

Year EX2 N EX
1945 5,893,071 79 20,892.30
1946 8,570,928 116 30,435.82
1947 6,571,497 104 24,271.00
1948 12,162,545 158 43,086.00
1949 18,771,227 235 64,451.00
1950 18,939,211 253 67,064.18
70,908,419 945 250,200.50

 

C.T. = {250120022 = 61,876,286
Between Years SS =

Total 58 = 812 - C.T. = 9,082,188 62,207,097-61,876,236 =
880,861

 

.1. any

. uh»

...a.v..c«....

 

M IIIIIIIIIII E UNIVE IIIIIII BRARI

 

Iw11.11,le111111Ilmlmll mu“