W. J. (Wilbur Jerome) Carmichael.

Variations in farrow : with special reference to the birth weight of pigs online

. (page 1 of 3)
Online LibraryW. J. (Wilbur Jerome) CarmichaelVariations in farrow : with special reference to the birth weight of pigs → online text (page 1 of 3)
Font size
QR-code for this ebook









Agricultural Experiment Station
















TABLES . . 80




The object of this research has been to study the variations in
farrow among the several breeds of swine, together with some of
the factors other than breed that may cause noticeable differences
among the litters at the time of birth.

The breeding and farrowing records which have been kept in
connection with the herd of swine under the management of the Ani-
mal Husbandry Department of the University of Illinois, during the
years 1903 to 1916 inclusive, have furnished the data for this study.
These records include 720 litters containing 5,840 pigs of seven dis-
tinct breeds and a number of different crosses, as well as a few
litters the exact breeding of which is not definitely known. Com-
plete information was not obtained in connection with every litter,
but an attempt was made to secure: (a) the breeding of sire and
dam; (b) age of dam; (c) length of gestation period; (d) number
of pigs in the litter; (e) sex of the pigs; (f) order of farrowing;
(g) individual weight of pigs; and (h) the number of pigs farrowed
dead or immature.

For the most part the records were kept by the herdsman actively
in charge at the barns. The information contained on his records
was copied into the permanent records of the office, from which the
data for this study were taken. Where there was any doubt as to
the accuracy of the data, it was assumed that there was no in-
formation on that particular phase in connection with the litter in
question. However, there has been no attempt to eliminate any
litters produced by sows on any specific ration or under any par-
ticular conditions, nor has there been any elimination other than
that necessitated by lack of data or obvious error in the information
collected. All litters farrowed in the herd within the given years
have therefore been considered in this study so far as the data would

Individual pig weights were taken before the pigs were allowed
to suckle, tested spring balance scales being used for this purpose*
and the weights were recorded by pounds and tenths. In case it

*Now Secretary of the National Swine Growers' Association.


68 BULLETIN No. 226 [May,

was not possible to get the birth weights in the order of farrowing,
the pigs were weighed individually as soon thereafter as possible,
and the fact that the order of farrowing was not obtained was noted
on the record. In some cases individual weights were not obtained
because a few hours' time had elapsed before there was an opportu-
nity to make such weighings. In those cases the litter weights, even
tho taken, were not used for this study.

The rations on which the sows were fed and the method of feed-
ing and handling doubtless caused some of the variations herein-
after noted. However, the rations have not been considered since
the feed records are not complete for the entire number of sows nor
for a sufficient number of them to make such a study of much value.
The herd sows have in general been handled similar to those in many
herds of pure-bred swine.

A very systematic study of the field is difficult since so many
factors may influence the birth weight of pigs and cause other varia-
tions in connection with farrowing. Influences may so completely
overlap one another that in some cases it is almost impossible to
account for the variations which occur. However, the possible in-
fluences considered in this study, and which will be taken up in the
order named, are as follows: (1) length of gestation period; (2)
size of litter; (3) age of dam; (4) breed; (5) time of year; (6)
order of farrowing; (7) succeeding litters from the -same sow.

The gestation period for sows is generally considered to be from
112 to 114 days, with a popular belief that the older sows will carry
their litters from one to three days longer than will the gilts. Table
1 shows the relation of the age of the dam to the length of gestation
period for 549 litters.

Judging from these data, the general belief that older sows have
a gestation period one to three days longer than younger sows seems
to have no very good basis; in fact, the one-year-old sows carried
their pigs slightly longer than the average for the 549 litters from
Sows of all ages, which was 114.58 days from the time of service.
However, it is perhaps worth noting that the average age of all sows
farrowing with gestation* periods of 114 days or less was 2.24 years,
while that of the sows which carried their litters 115 days or longer
was 2.31 years ; which indicates that there may be a slight variation
due to age. Also, many of the sows four years old or above had
noticeably longer gestation periods than the average, but there were
so few litters under observation among the very old sows that these
differences are of doubtful significance.

A variation of twenty-six days in the length of gestation is some-
what longer than is expected. However, the fact that 93.6 percent


of the litters were farrowed between the lllth and 119th days after
being bred indicates that the extreme cases shown in the table are
riot duplicated with very great frequency. A further study of the
same distribution indicates that the date of farrow can be predicted
with but reasonable accuracy since 73.8 percent of the litters were
carried by the sows for periods varying from 113 to 117 days. Less
than 20 percent of the litters were farrowed on the 114th day, which
was the day of greatest frequency of farrow.

Table 2 shows the relation of the size of litter to the length of
gestation for the same 549 litters which are shown in Table 1. From
this it seems that there is a tendency for some of the litters which
are carried longer than the average to be smaller than would
be found in a chance distribution. This is perhaps more clearly
brought out in Table 3, which shows that the average number of
pigs (7.6) farrowed in gestation periods longer than the average
was less than the grand average number of pigs (8.00) per litter in
466 litters.

Table 4 shows a very even distribution of pigs of the various birth
weights according to the length of the gestation period. The heav-
iest pig in the whole population of 4,115 pigs studied (4.9 pounds)
was farrowed in a litter of the shortest gestation period (98 days),
but this pig seemed to stand out alone, there being none other nearly
so heavy. The lightest pig which was farrowed (0.2 pound) came
in one of the litters farrowed on the 115th day, or near the average
for the length of gestation. Aside from the two pigs to which ref-
erence has just been made, and perhaps four or five others, none of
which are outside of a distribution such as one would expect to find,
the 4,115 pigs on which the individual weight as well as the length
of gestation period was secured are very evenly centered about the
114- to 115-day gestation period and the 2.5 pounds weight per pig.

The distribution of 5,188 pigs as to individual weight, which is
the number regarding which individual weights were secured, is
given in the graph shown on page 70. This shows a variation
from 0.2 pound to 4.9 pounds, with a grand average or mean weight
of 2.55 pounds and with 66.9 percent of the pigs weighing between
two and three pounds. It will be noted that the greatest number
of pigs usually occurs at the pound and half-pound frequencies.
This is doubtless due to the fact that the scales were graduated by
pounds and tenths and the pounds and half-pounds were more clearly
marked than the other divisions, thus making it more natural thai
the nearest pound or half-pound should be read. It is doubtful, how-
ever, whether this error would noticeably alter the true average
weight of the whole population, the inequalities probably balancing
one another.





G 1.4 1.5


A study of the litters as regards the relation of breed to
the length of gestation period (Table 5) reveals the fact that two
breeds, Berkshire and Poland China, stand out alone as having
longer gestation periods than the average for the 549 litters. The
Poland Chinas carried their pigs an inappreciable fraction of a day
longer than the grand average, but the Berkshires ran over more
than three-fourths of a day. The fact that the Berkshires seem to
have longer normal gestation periods than any other breed involved
in this study, together with the fact that there were more Berkshire
litters studied than litters of all other breeds combined, explains in
part the reason for the grand average gestation period under con-
sideration being somewhat longer in these 549 litters than is gen-
erally considered to be normal. Outside of the Hampshires, of which
there were but two litters, the Chester Whites had the shortest
average gestation period. Following these in order of increasing
gestation period are the Duroc-Jerseys, Large Yorkshires, Tarn-
worths, Poland Chinas, and Berkshires.

A study of the sex of the pigs in relation to the length of gesta-
tion period (Table 6) shows no apparent tendency for either sex to
predominate more than normal. In a majority of cases there were
more males than females. Of the 4,363 pigs on which the sex as
well as the gestation period was recorded, 2,074 were females and
2,289 males, or 47.5 percent females and 52.5 percent males.

Table 7, showing the sex of the first and last born according
to length of gestation period, indicates that there is possibly a
tendency for males to predominate among the first farrowed in any
given litter, and for the sex to be rather evenly distributed in the
case of the last farrowed. However, from this table it is seen that
in litters farrowed on the 114th day, females predominated among
both the first and last farrowed, notwithstanding the fact that there
were more males than females among the total number farrowed.
The study of the sex of the first and last farrowed involves so few
individuals that no definite conclusions can be drawn.


The number of pigs per litter was obtained in connection with
each of 720 litters containing a total of 5,840 pigs. Table 8 shows
the sex of these pigs as they varied within litters of different sizes.
Altho there were 183 pigs the sex of which was not obtained, a sex
distribution of the remaining 5,657 is interesting and seems to in-
dicate that there is a decided tendency for males to predominate,
since of this number 2,933 were males and only 2,724 were females,
or 51.9 percent males and 48.1 percent females. There seemed to be
no noticeable correlation between the size of the litter and sex, a

72 BULLETIN No. 226 [May,

predominance of males being found in a majority of different-sized
litters rather than merely a tendency for this sex to predominate in
those of certain sizes, either small or large.

Table 9, showing the number of pigs farrowed dead or immature
in the litters of different sizes, seems to indicate that there is a
tendency for the larger litters to contain a larger proportion of
dead or immature pigs than are found in the smaller litters, since in
litters of eight pigs or less the percentage of dead or immature at
birth was 7.7, whereas in the litters of more than eight pigs, 10.5
percent is the corresponding figure. Among those farrowed dead
there was a greater proportion of males than females 56.0 percent
males and 44.0 percent females, which is a significant difference.

A study of the distribution of 5,774 pigs according to the size
of the litter as well as the age of dam, as given in Table 10, indicates
that sows under two years of age produce smaller litters than do
sows two to four years of age. The line of averages at the bottom
of the table shows a gradual increase in the size of litter as the sows
grew older up to the time that they were three years old. After
the sows had attained the age of three years they had a tendency
to produce fewer pigs in each succeeding litter, altho the few sows
which were kept until they were over five years of age seemed to
maintain their fecundity better than the average even at younger
ages. There was, however, at least one force operating in the herd
which would tend to maintain or possibly increase the size of litter ;
namely, the fact that no sows were retained in the breeding herd
unless they proved to be producers either of large litters or of litters
containing pigs of special worth. As a result of this selection some
of the gilts which produced small litters were discarded and did
not have an opportunity to exert an influence as older sows. On
the other hand, this sort of selection is just what would probably
be made in any herd, and the results here are therefore somewhat
comparable to those which might generally be found.

The distribution of 2,483 sow pigs showing the relation of the
weights of the pigs to the size of the litter in which they were far-
rowed (Table 11), as well as a similar distribution for the 2,705
boar pigs on which the individual weights at farrowing time were
secured (Table 12), shows a marked regularity in centering around
the averages. The succeeding table (Table 13) shows the same data
in a form which may be more easily studied. In these comparisons
the average weight of the 2,483 sow pigs is shown as 2.51 pounds
and that of the 2,705 boar pigs as 2.59 pounds, the males being
heavier by 0.08 pound. A study of the average weight of all pigs
farrowed in litters of different sizes shows rather clearly that in
litters smaller than the average (8.0 pigs) the average weight per
pig (2.67 pounds) is likely to be greater than the grand average


weight for all litters (2.55 pounds for the 5,188 pigs on which the
individual weights were obtained). Those farrowed in litters of
more than eight pigs weighed an average of 2.47 pounds each. In
every case where the litter contained more pigs than the average,
the average weight of all pigs in litters of that size was noticeably
less than the grand average. Making a similar comparison by sex,
we find that likewise the pigs of either sex in litters larger than
the average have a tendency to be lighter than the average for pigs
of that sex, and in litters smaller than the average the tendency is
for the pigs to be heavier than the average for that sex.


Sows which are two years old or older are generally considered
to be producers of larger litters and stronger pigs than are gilts or
sows one and one-half years old.

Table 14, showing the relation of the age of the dam to the birth
weight of the pigs, indicates rather clearly that the young sowjs
farrow noticeably lighter pigs than do the older mothers. The
average weight of pigs from sows under two years of age was 2.44
pounds, as compared with an average weight of 2.61 pounds for
the pigs from sows two years old and older. There seemed to be
a fairly general increase in the weight of the pigs as the sows grew
older, even until they were eight years old, altho there were so few
sows over five years that the weight of the pigs from sows past
maturity may not be representative. On the whole, however, the
belief that the young sows produce smaller pigs is substantiated by
these data. Table 10, given in connection with the discussion on
the size of litters, likewise substantiates the popular belief that
up to a certain age, which is about three years, there is an increase
in the size of litters as the sow grows older.

One frequently hears statements to the effect that even tho the
older sows farrow larger litters than the gilts, they do not raise as
large a percentage of the pigs farrowed because more of them are
farrowed weak or dead. Table 15, in which the number of dead or
immature pigs is given in litters from sows of different ages, shows
that the old sows farrowed a greater percentage of dead pigs than
did young dams. Eight and one-tenth percent of the pigs farrowed
by sows under three years of age were dead or immature, whereas
12.3 percent (an increase of 4.2 percent) of those from dams three
years or more in age were in similar condition at birth. There was,
however, marked irregularity among the proportionate number ofr
dead or immature at birth among the litters from sows three years
old and over. There were also few litters of these classes.

The sex of pigs in litters from sows of different ages (Table 16)
is about as evenly distributed as it was in litters of different sizes

74 BULLETIN No. 226 [May,

(Table 8) or in litters of different lengths of gestation (Table 6).
It shows a uniform tendency for males to be present in slightly
greater numbers than are the females.


Any study which can be made of the different breeds from the
data available in this research cannot be conclusive because for
accurate comparisons there have been too few litters from any one
breed, with the possible exception of the Berkshires. At best, breed
comparisons of any kind are of doubtful value and are dangerous
indulgences. However, these data were collected and are given here
as a possible supplement to any other which may be available at
present or obtained at any future time. Table 17 shows the total
number of litters of each breed in the 720 litters under observation.
From this table it is evident that more Berkshire litters were far-
rowed in the herd during the years that these data were being col-
lected (1903-1916) than all other breeds combined.

The compilation showing by breeds the average age of sows,
the average length of gestation period, the number of pigs per
litter, and the average weight per pig at farrowing, as pre-
sented in Table 18, gives an opportunity for comparison of breeds,
so far as the available material herein presented will permit. (All
litters of which the breeding was not definitely known or concern-
ing which there was any doubt were omitted.)

The average age of the sows at the time the litters were farrowed
did not vary widely, there being some sows of each breed that were
kept until they were four or five years old ; consequently any varia-
tion which may be found cannot logically be explained by variations
in age.

In the discussion in connection with Table 5 the variation in
length of gestation was considered, and therefore will not be re-
peated here.

The largest variation which was found among the litters of the
different breeds was in the size of litters, altho there was some no-
ticeable variation in the weight of the pigs. Arranging the breeds
according to the descending number of pigs per litter, as well as
the descending weight per pig, we have the following :

Pigs per litter Weight per pig

Large Yorkshire 11.58 Berkshire 2.61

Chester White 9.59 Large Yorkshire 2.60

Tamworth 9.43 Chester White 2.59

Duroc-Jersey 8.74 Tamworth 2.58

Crossbred 8.00 Poland China 2.50

Berkshire 7.42 Hampshire 2.50

Hampshire 7.00 Crossbred 2.46

Poland China 6.57 Duroc-Jersey 2.25

Average 8.00 Average 2.53


In this at first there seems to be some evidence, which is con-
tradictory to that found in the study of the size of the litters and
its effect on the weight of the pigs, for the Large Yorkshire litters,
altho they contained more pigs as an average than the average for
litters, were composed of pigs which were heavier than the average.
A similar statement is also true of the Chester White and Tamworth
pigs. Contrary to this, we find that the Hampshire and Poland
China litters, with fewer pigs than the average, had lighter pigs than
the mean of the whole population. Further, it is to be noted that
the Duroc-Jersey pigs weighed over one-fourth pound (0.28 pound)
less than the average for all pigs. Such findings are a concrete
illustration of the differences due to breeding.

The foregoing tabulation shows the results of but 457 litters, the
litters containing a total of 3,658 pigs, an average of 8.00 individuals
per litter. The sows averaged 2.35 years in age.


The time of the year at which pigs are farrowed does not
seem to exert any very noticeable influence, with any regularity,
upon the length of gestation period, size of litter, or weight of pigs,
altho possibly there is a tendency for larger litters and heavier pigs
to occur in the months of normal farrowing seasons. This, of
course, might be expected since sows farrowing in other months
are, in many cases, ones which were re-bred, having been originally
bred for the normal farrowing season. Such sows are doubtless
somewhat abnormal. There are no positive indications that the sows
will carry their litters longer during summer months than in winter,
or vice versa. Summarizing here the data presented in Table 19, and
arranging it by months in descending order, we have :

. , Length of gestation Pigs per Weight

period litter per pig

July September October July

April May August September

March February April August

September March March October

October June February February

August October September April

June April ' May March

May July June May

February November July November

November August November June


Table 20, showing the relation of order of farrowing to sex and

weight of pigs in the 261 litters on which the order of farrowing
as well as the birthweight of all the pigs in the litters was ob-

76 BULLETIN No. 226

tained, indicates rather strongly that among the first farrowed, males
tend to predominate more than normal, for there were 155 males to
106 females among the first born. The sex of the last farrowed was
in about the normal ratio.

The same table also indicates that there is a slight tendency for
the first farrowed (average weight 2.60 pounds) to be heavier than
the last farrowed (average weight 2.43 pounds). This is perhaps
in part due to the predisposition of the males, which predominated
here, to be heavier than the females. In the case of the first far-
rowed here, the males averaged 2.67 pounds and the females only
2.50 pounds. There was a somewhat similar tendency for the males
to predominate and to be somewhat heavier than the females, among
the last farrowed, but the tendency was not so marked as among
the first farrowed.


In Table 21 is shown the average age, length of gestation period,
number of pigs per litter, and weight of litter and pigs in succeding
litters from the same sow.

The gestation period of the first litter is rather consistently
longer than for those following it, as was shown in Table 1. There
is an increase in the number of pigs per litter in most cases until
the fourth litter, at which time the sows are three to three and one-
half years old. The total weight of the litter increases with each
succeeding litter up to the fourth owing to the increase in the num-
ber of pigs per litter as well as in the weight of the individual pigs.

The age at which a sow should farrow her first litter, as well as
the optimum frequency of succeeding litters, is a much debated
subject. No positive conclusions can be drawn from Table 22, which
gives the succeeding litters from the same sows at different ages,
for there are doubtless other factors than those in the table which
should be taken into consideration in determining the age at which
a sow should farrow. One would infer from the first part of the
table that sows farrowed their first litters at one and one-half years
of age had shorter gestation periods and farrowed more and heavier
pigs than sows that farrowed their first litters at one year of age.
Upon studying other parts of the table, it is readily seen that con-
flicting conclusions can be drawn. The sows that farrowed their
first litters at one and one-half years of age rather consistently
farrowed heavier pigs in the first litter than those which farrowed

1 3

Online LibraryW. J. (Wilbur Jerome) CarmichaelVariations in farrow : with special reference to the birth weight of pigs → online text (page 1 of 3)