Published December 5, 2014
Quantification of factors affecting semen traits in artificial
insemination boars from animal model analyses1
J. Wolf2 and J. Smital
Institute of Animal Science, PO Box 1, CZ 10401 Prague Uhříněves, Czech Republic
ABSTRACT: The objective of this study was to investigate individual fixed effects in an animal model
for breeding value estimation for semen traits of pig
sire breeds. Data (151,755 ejaculates collected from
2000 through 2007 from 2,077 Duroc, sire line of Large
White, Piétrain, and single cross boars between these
breeds) were from 20 AI centers in the Czech Republic. Traits considered per ejaculate were semen volume,
sperm concentration, motility, percentage of abnormal
sperm, total number of sperm, and number of functional sperm. Fixed effects in the animal model were
month of collection, age of the boar at collection, interval between subsequent collections, combined effect
of AI center and year, and breed or crossbred combination. Semen volume was greatest from October through
December and was least in March and April. Sperm
concentration was greatest in winter and early spring
and least in late summer and early autumn. Both total
sperm number and number of functional sperm were
greatest in winter and least in summer. Semen volume
increased until about 2 yr of age and remained relative-
ly constant thereafter. Sperm concentration increased
sharply until 11 mo of age, followed by a long-term
moderate decrease until 3 yr of age and stabilization
thereafter. Motility decreased steadily with age, whereas the percentage of abnormal sperm increased over the
entire productive lifetime of the boar. There were initial steep increases with advancing age in total sperm
number and number of functional sperm, both reaching their maxima at about 2 yr of age and then dropping slightly to the end of the time scale investigated.
The interval between subsequent collections had a large
effect on sperm concentration. Motility tended to decrease and the percentage of abnormal sperm tended to
increase with lengthening time interval between collections. Both total sperm number and number of functional sperm rose as the interval between collections
increased to 10 d. Although boars will continue to be
selected mainly for their breeding values for production
and female reproduction traits, AI centers should also
place economically optimal emphasis on boars with favorable estimated breeding values for semen traits.
Key words: boar, breeding value estimation, fixed effect, pig, semen trait
©2009 American Society of Animal Science. All rights reserved.
INTRODUCTION
By allowing greater use of genetically superior sires,
AI plays an important role in animal breeding (Oh
et al., 2006b). Currently, boars of sire breeds (lines)
selected for commercial use as AI sires are evaluated
on grow-finish performance and carcass characteristics
(Oh et al., 2006a). Boars of the sire breeds kept in the
Czech Republic, for example, are evaluated on ADG
from birth until the end of the performance test and on
1
The research was supported by the project MZE 0002701401 of
the Ministry for Agriculture of the Czech Republic. Thanks are due
to the Pig Breeders Association of the Czech Republic for making
the data available and to W. D. Hohenboken (Philomath, OR) for
editing the English of the paper and for valuable comments.
2
Corresponding author: [email protected]
Received August 4, 2008.
Accepted January 20, 2009.
J. Anim. Sci. 2009. 87:1620–1627
doi:10.2527/jas.2008-1373
lean meat content at the end of the test. However, an
AI center cannot restrict itself to selection on production traits for the pig producer, but must also consider
factors that influence efficiency of the center such as
conformation, temperament, and sperm quantity and
quality. Basic semen traits that affect AI center profitability are volume, sperm concentration, and gross
sperm morphology (Robinson and Buhr, 2005).
Semen traits are known to be heritable (Grandjot et
al., 1997; Smital et al., 2005; Oh et al., 2006b; Wolf,
2008). Therefore, genetic evaluation of boars for semen
traits and selection based upon estimated breeding values is possible. On this basis, an animal model was developed for the genetic evaluation of semen traits of pig
sire breeds kept in the Czech Republic. The objective
of the present investigation is to present and discuss the
effects of environmental and genetic factors affecting
semen traits that were identified from test runs of the
animal model in which a large data set was analyzed.
1620
1621
Breeding value estimation for semen traits
Table 1. Summary statistics for semen traits
D1
Variable
Number
No. of boars
No. of sires/dams of boars
No. of ejaculates
Average No. of ejaculates per boar
Mean
Semen volume, mL
Sperm concentration, 103 sperm/mm3
Motility, %
Percentage of abnormal sperm, %
Total No. of sperm, 109 sperm
No. of functional sperm, 109 sperm
204
73/156
10,691
52
LW2
All
purebreds
P3
607
202
204/415
69/99
46,169
12,050
76
60
200
491
73.6
10.8
93.7
61.5
270
401
76.6
11.2
101.3
69.3
275
453
76.8
11.8
118.7
80.3
1,013
68,910
68
260
424
76.2
11.3
103.1
70.0
D × LW
D×P
LW × P
196
356
512
49/103
65/207
131/335
17,671
27,190
37,984
90
76
74
236
431
71.6
13.1
95.1
59.1
241
445
74.2
10.8
102.1
67.5
All
crossbreds
1,064
82,845
78
282
407
76.6
10.8
107.4
73.8
258
424
74.8
11.3
103.0
68.6
1
D = Duroc.
LW = sire line of Large White.
3
P = Piétrain.
2
MATERIALS AND METHODS
Animal Care and Use Committee approval was not
obtained for this study because the data were obtained
from an existing database (database of the Association
of Pig Breeders in the Czech Republic).
Animals and Traits
Data from the years 2000 through 2007 were collected from 20 AI centers in the Czech Republic. The
151,755 ejaculates originated from 2,077 boars of the
Duroc, sire line of Large White, and Piétrain breeds
and of single crosses between them. The data set was
made available by the Association of Pig Breeders in
the Czech Republic. The number of boars, number of
ejaculates, and average number of ejaculates per boar
for each breed and crossbred combination are summarized in Table 1. These numbers refer to the edited
data set that was used for all calculations (see below
for details).
The following semen characteristics were measured
on each ejaculate: semen volume or ejaculate volume
(Vol) in milliliters (i.e., volume of the sperm rich fraction) measured by graduated cylinder; sperm concentration (Con, in 1,000 sperm cells per mm3) measured
by photocolorimetry; motility (Mo, progressive motion
of spermatozoa in per cent, i.e., proportion of sperm
cells actively moving, evaluated microscopically); and
percentage of abnormal sperm (Ab, percentage of deformed sperm cells, also evaluated microscopically).
The total number of sperm in the ejaculate (Ntotal, in
109 sperm cells) was calculated as
Ntotal = Vol × Con/1,000,
and the number of functional sperm (Nfunc, in 109 sperm
cells) was estimated as (Smital et al., 2004):
Nfunc = Ntotal(Mo/100)(1 − Ab/100).
Statistical Analyses
A model similar to that reported by Wolf (2008) was
used for (co)variance and breeding value estimation:
yijklmno = monthi + agej + intk + center_yearl
+ breedm + pn + an + eijklmno,
where yijklmno is the sperm characteristic measured on
the oth ejaculate of the nth boar of the mth breed or
crossbred combination, monthi is the effect of the season (month), agej is the effect of the age class of the
boar, intk is the effect of the interval between the present and previous semen collection, center_yearl is the
combined effect of the AI center and year, breedm is the
effect of the breed or the crossbred combination of the
boar, pn is the permanent environmental effect of the
boar, an is the additive genetic effect of the boar, and
eijklmno is the residual effect. The pedigree was traced
back to approximately the year 1985.
To form age classes, the age of each boar in months
at each collection was calculated. Ejaculates from animals less than 8 mo of age or greater than 48 mo were
excluded from the data set. Age classes with monthly
intervals were used up to an age of 28 mo. For animals aged 29 to 38 mo, 2-mo intervals were formed. For
animals over 38 mo of age, the following 3 classes were
formed: 39 to 41 mo, 42 to 44 mo, and 45 to 48 mo.
Preliminary analyses showed that all measured traits
were most sensitive to changes in the interval between
2 semen collections from the same boar when that interval was short. Therefore, for intervals less than 11 d,
classes were formed with an interval of 1 d. For intervals
of 11 d and more, the following 3 classes were formed:
11 to 12 d, 13 to 15 d, and 16 to 21 d. The first semen
collection of each boar and semen collections with an
interval of 1 d or more than 21 d also were not included
in the analyses.
Data were excluded from further analyses if any one
of the following conditions was not fulfilled: the mini-
1622
Wolf and Smital
Table 2. Estimates of genetic parameters used for breeding value estimation of semen traits1
Item
Vol
Variance component
Additive genetic variance
Permanent environmental variance
Residual variance
Proportions of variance
Heritability
Permanent effect
Repeatability
Correlation
Additive genetic
Permanent environmental
Residual
2,011.5
1,196.9
4,009.9
0.28
0.17
0.44
Vol − Con
−0.60
−0.61
−0.34
Con
4,505.9
3,593.2
14,344.1
0.20
0.16
0.36
Vol − Mo
−0.12
−0.01
−0.01
Mo
Ab
0.9
2.9
12.2
4.8
8.2
17.4
0.05
0.18
0.24
Vol − Ab
−0.20
−0.01
−0.01
0.16
0.27
0.43
Con − Mo
0.15
0.14
0.06
Ntotal
Nfunc
209.632
143.742
858.409
108.503
77.278
410.591
0.17
0.12
0.29
Con − Ab
0.11
0.01
−0.01
0.18
0.13
0.31
Mo − Ab
−0.57
−0.28
−0.07
1
Vol = semen volume (mL); Con = sperm concentration (103 sperm/mm3); Mo = motility (%); Ab = percentage of abnormal sperm (%); Ntotal
= total number of sperm (109 sperm); and Nfunc = number of functional sperm (109 sperm).
mum number of ejaculates per AI center was 100, the
minimum number of ejaculates per AI center and year
subclass was 20, and the minimum number of semen
collections per boar had to be 5. Furthermore, trait
values for individual ejaculates had to be within the
following intervals: semen volume 50 to 600 mL, sperm
concentration 50 to 900 thousand sperm per mm3,
motility 50 to 100%, percentage of abnormal sperm 0
to 30%, total number of sperm 5 × 109 to 200 × 109
sperm. The means for all traits of the final data set are
given in Table 1.
Restricted maximum likelihood and optimization by a
quasi Newton algorithm with analytical gradients (Neumaier and Groeneveld, 1998) as implemented in VCE
5.0 program (Kovač et al., 2002) were used to estimate
the variances and covariances. Program PEST (Groeneveld et al., 1990) with the SMP solver was used for
prediction and estimation of random and fixed effects,
respectively, in the model given above. A 4-trait animal
model was used for simultaneous analysis of semen volume, sperm concentration, motility, and percentage of
abnormal sperm. Total sperm number and number of
functional sperm were mathematical functions of the
4 measured semen traits; therefore, they could not be
included in the multiple-trait animal model and singletrait animal models were used for the 2 derived semen
traits.
Effects of fixed factors are presented as deviations
from their average effect, with the exception of the
breed effects, which are expressed as deviations from
the corresponding effect of the Piétrain breed. Heterotic effects were calculated from the estimated effects
for each crossbred combination and its purebred parent
breeds in the usual way. Let breedA×B equal the effect
for the crossbred combination A × B and breedA and
breedB equal the effects of breeds A and B. Then the
heterotic effect for combination A × B, hA×B, is
1
hA´B = breedA´B - (breedA + breedB ).
2
Each heterotic effect was also expressed as a percentage
of the midparent phenotypic value.
To get an impression of the environmental trend
across time, the effect of the year of collection was calculated from the combined effect for the AI center and
year using the GLM procedure of SAS Institute Inc.
(Cary, NC). Average breeding values of boars born in
the same year were the basis for estimation of the genetic trend.
RESULTS
Genetic Parameters
The estimates of genetic parameters used in breeding value estimation are summarized in Table 2. Semen volume showed the greatest heritability (near 0.3).
With the exception of motility, the heritabilities for the
remaining traits were in the range from 0.15 to 0.20.
The proportion of variance caused by the permanent
effect was mostly between 0.10 and 0.20; only for the
percentage of abnormal sperm was this range exceeded.
The repeatability was approximately in the range from
0.25 to 0.45, with motility at the lower limit and semen
volume at the upper limit.
High negative genetic correlations were observed between semen volume and sperm concentration and between motility and percentage of abnormal sperm. The
correlations caused by the permanent environmental
effect of the boar behaved similarly as the genetic correlations. The residual correlations were mostly near 0,
with the exception of the moderate negative correlation
between semen volume and sperm concentration.
Seasonal Effects
Seasonal variability for the 6 semen traits, expressed
as deviations from their overall annual averages, are
presented in Table 3. Semen volume had the greatest values from October to December and was least in
March and April. Sperm concentration was greatest in
winter and early spring (December to April) and least
in late summer and early autumn (August to October). Motility was relatively constant throughout the
year. Seasonal differences in the percentage of abnor-
1623
Breeding value estimation for semen traits
Figure 1. Effect of age at collection on semen volume (mL) and
sperm concentration (103 sperm/mm3).
Figure 2. Effect of age at collection on motility (%) and percentage
of abnormal sperm.
mal sperm also were small (less than 1%), but because
the trait average was as low as 11%, differences in the
order of magnitude of 1% may be of some importance.
Both total sperm number and the number of functional
sperm were greatest in winter and least in summer.
Effect of the Interval Between
Subsequent Collections
Effect of Age at Collection
The dependence of semen traits on the age of the
boar at collection is shown in Figures 1 to 3. Semen volume increased until an age of about 2 yr and remained
more or less constant thereafter. The total increase in
volume was approximately 100 mL. Sperm concentration increased until 11 mo of age, followed by a longterm moderate decrease until 3 yr of age and relative
stabilization thereafter.
Motility decreased steadily with age, but the overall
decrease was less than 1% (Figure 2). In contrast, the
percentage of abnormal sperm increased over the productive lifetime of the boar, achieving a difference of
nearly 4% between the youngest (8 mo) and the oldest
(48 mo) boars. Initially, both total sperm number and
the number of functional sperm increased sharply with
age, reaching their maxima at about 2 yr and then
decreasing slightly through the end of the age range
investigated (Figure 3).
The interval between subsequent collections had a
large effect on sperm concentration (Figure 4). An interval of 2 d resulted in 100 × 103 fewer sperm cells per
mm3 compared with an interval of 6 d. Intervals from
10 to 21 d yielded 40 × 103 to 50 × 103 more sperm
cells per mm3 per ejaculate than collections separated
by a 6-d interval. The influence of the interval between
collections on semen volume was considerably less than
its effect on sperm concentration. A slight increase in
semen volume was observed when increasing the interval from 2 to 7 d; for longer intervals, the values did not
change markedly.
Motility tended to decrease and the percentage of
abnormal sperm tended to increase with lengthening
interval between collections, but the changes were relatively small (Figure 5). Total sperm number and number of functional sperm both increased with increased
collection intervals up to 10 d (Figure 6). At longer
intervals, these values decreased slightly.
Breed and Heterotic Effects
Differences in semen traits among purebred boars are
shown in Table 4. Duroc had the least semen volume
Table 3. Effect of month of collection (as deviation from overall annual average) on
semen traits1
Month
January
February
March
April
May
June
July
August
September
October
November
December
Vol
Con
Mo
Ab
Ntotal
Nfunc
1.1
−7.6
−13.8
−16.0
−9.2
−7.4
−6.5
−2.6
7.0
18.2
23.3
13.6
9.1
11.7
16.0
12.8
4.0
−5.8
−7.0
−13.4
−16.7
−12.5
−8.4
10.2
0.17
0.25
0.29
0.38
0.19
0.17
−0.14
−0.24
−0.29
−0.25
−0.31
−0.22
−0.25
−0.46
−0.41
−0.32
−0.48
0.05
0.00
−0.12
0.32
0.51
0.54
0.62
3.4
0.2
−1.7
−3.0
−2.8
−4.3
−4.1
−4.3
−1.7
3.8
6.6
7.9
2.6
0.7
−0.6
−1.4
−1.3
−2.9
−2.8
−2.9
−1.7
2.0
3.7
4.6
1
Vol = semen volume (mL); Con = sperm concentration (103 sperm/mm3); Mo = motility (%); Ab = percentage of abnormal sperm (%); Ntotal = total number of sperm (109 sperm); and Nfunc = number of functional
sperm (109 sperm).
1624
Wolf and Smital
Figure 3. Effect of age at collection on the total number of sperm
(109 sperm) and the number of functional sperm (109 sperm).
and greatest sperm concentration, whereas the sire line
of Large White had the greatest semen volume and
the least sperm concentration, with intermediate values
for both traits in the Piétrain breed. Duroc and the
sire line of Large White showed more favorable values
than Piétrain for motility and percentage of abnormal
sperm. The total number of sperm and the number of
functional sperm were greatest in the sire line of Large
White and least in Duroc.
Estimates of heterotic effects as absolute values and
as percentages of midparent phenotypic value are shown
in Table 5. The percentage of abnormal sperm was from
10 to 26% less, and the number of functional sperm was
from 3 to 13% greater in crossbred than in purebred
boars. There was also a beneficial effect of crossing on
total sperm number. In both crossbred combinations
with Piétrain, there was 6 to 7% heterosis in semen
volume. Crossing had no or a slightly negative effect on
sperm concentration. The most favorable expression of
heterotic effects summarized across traits occurred in
the Duroc × Piétrain cross.
Environmental and Genetic Trend
The environmental trend of the semen traits expressed as the effect of the year of collection is shown
in Table 6. All values are expressed as deviations from
the year 2000 averages. Semen volume increased by 39
mL between 2000 and 2007, but the increase was not
equally distributed over years. No clear tendency was
Figure 4. Effect of the interval between subsequent collections on
semen volume (mL) and sperm concentration (103 sperm/mm3).
Figure 5. Effect of the interval between subsequent collections on
motility (%) and percentage of abnormal sperm.
observed in sperm concentration. Motility decreased by
approximately 1%, and abnormal sperm increased by
approximately 1.5% over the time interval investigated.
A trend toward greater values was observed both for
total sperm number and for the number of functional
sperm. No clear genetic trend was found for any trait
(data not shown).
DISCUSSION
The estimates of effects presented in this paper were
derived from animal model analyses that were designed
to account simultaneously for all identifiable genetic
and environmental factors potentially affecting the
semen traits. In most comparative studies, only phenotypic means (for boars in different age classes, for
example) are reported, which may be unadjusted for
other effects or only partially adjusted in linear model
analyses. Both of these types of studies are likely to
suffer a greater degree of confounding among modeled effects than our animal model analyses, in which
additive genetic relationships among boars have also
been accounted for. Our more thorough accounting of
relationships among causes and effects may partially
account for differences in results and conclusions between our study and other results from the literature.
Furthermore, it should be emphasized that the effects
reported herein were estimated not from small samples
Figure 6. Effect of the interval between subsequent collections on
the total number of sperm (109 sperm) and the number of functional
sperm (109 sperm).
1625
Breeding value estimation for semen traits
1
Table 4. Effect of breed or crossbred combination on semen traits
Breed or crossbred combination
Duroc (D)
Sire line of Large White (LW)
Piétrain (P)
D × LW
D×P
LW × P
Vol
Con
Mo
Ab
Ntotal
Nfunc
−39.8
29.4
0.0
−6.7
−2.8
31.9
56.7
−16.2
0.0
19.2
28.6
−17.7
0.53
0.39
0.00
0.68
1.22
0.57
−1.95
−1.94
0.00
−3.03
−3.86
−2.28
−7.0
6.2
0.0
0.6
5.8
7.0
−2.7
6.2
0.0
3.7
7.8
6.9
1
Vol = semen volume (mL); Con = sperm concentration (103 sperm/mm3); Mo = motility (%); Ab = percentage of abnormal sperm (%); Ntotal
= total number of sperm (109 sperm); and Nfunc = number of functional sperm (109 sperm). The effects were defined as deviation from the effect
of the Piétrain breed.
or experimental populations, but from all AI boars of
the analyzed breeds and single crosses that were used in
the Czech Republic from 2000 through 2007.
General Discussion of the Model
The basic principle for creating an animal model
should be to include all traits in one model because all
of the traits are measured on the same experimental
unit (boar). Therefore, at first glance, a 6-trait model
should have been used. However, total sperm number is
a function of semen volume and sperm concentration,
and the number of functional sperm is a mathematical function of the 4 traits that are directly measured.
Therefore, these 2 traits were not included in the overall model because to do so could have caused problems
in numerical analysis. Furthermore, interpretation of
correlations between traits where one is a clear mathematical function of the other is problematic.
Although our overall objective was to design methods for breeding value estimation of purebred boars,
data from crossbreds were included in our analyses.
This added information increased the precision with
which environmental fixed effects were estimated and
provided boars with larger numbers of relatives to increase accuracy of breeding value estimation.
Genetic Parameters
Our results have shown that semen traits are heritable with values between 0.05 and 0.28. These values
are in a similar order of magnitude or greater than
heritabilities for litter size traits. That means they are
sufficiently large to allow for selection on these traits
using an animal model. Functions of these traits, such
as the total number of sperm in the ejaculate or the
number of functional sperm, may also be used for selection purposes.
Repeatability is a measure of stability of the trait
values over collections. The greater the repeatability
value, the greater the stability. From this point of view,
it is of practical importance for AI stations (prediction
of the repeated performance of the boar). However, for
selection decisions, the heritability estimates are more
important.
The negative genetic correlation between semen volume and sperm concentration is unfavorable for selection on the total number of sperm. On the other hand,
the negative correlation between motility and the percentage of abnormal sperm is favorable.
There are only a very limited number of literature
sources on estimates of genetic parameters for semen
traits of the boar. These sources are summarized and
discussed in Wolf (2008).
Seasonal Effects
Like their wild counterparts, most farm animal species at mid and high latitudes show seasonal variation
in reproductive phenomena such as ovulation frequency, spermatogenic activity, gamete quality, and sexual
behavior (Chemineau et al., 2007). According to Colenbrander and Kemp (1990), sperm production of boars
may fluctuate as much as 25 to 30% throughout the
Table 5. Heterotic effects1
Breed or crossbred combination2
Absolute value
D × LW
D × P
LW × P
Relative values, % of phenotypic midparent
D × LW
D × P
LW × P
Vol
Con
Mo
Ab
Ntotal
Nfunc
−1.5
17.1
17.2
−1.1
0.2
−9.6
0.2
1.0
0.4
−1.1
−2.9
−1.3
1.0
9.4
3.9
1.9
9.1
3.7
−0.6
7.2
6.3
−0.2
0.0
−2.2
0.3
1.3
0.5
−10.0
−25.7
−11.3
1.0
8.9
3.5
2.9
12.8
4.9
1
Vol = semen volume (mL); Con = sperm concentration (103 sperm/mm3); Mo = motility (%); Ab = percentage of abnormal sperm (%); Ntotal
= total number of sperm (109 sperm); and Nfunc = number of functional sperm (109 sperm).
2
D = Duroc; LW = sire line of Large White; and P = Piétrain.
1626
Wolf and Smital
Table 6. Effect of year of collection on semen traits1
Year
Vol
Con
Mo
Ab
Ntotal
Nfunc
2000
2001
2002
2003
2004
2005
2006
2007
0.0
9.9
25.2
21.3
26.8
24.3
31.5
39.1
0.0
−10.6
−10.5
21.8
−7.5
9.3
13.5
3.9
0.00
0.12
−0.11
−0.35
−0.32
−0.42
−1.25
−0.92
0.00
−0.45
−0.43
0.17
0.76
1.35
1.47
1.63
0.0
0.8
6.2
13.0
8.3
12.4
16.7
18.2
0.0
0.9
4.6
9.0
5.1
7.0
9.2
10.3
1
Vol = semen volume (mL); Con = sperm concentration (103 sperm/mm3); Mo = motility (%); Ab = percentage of abnormal sperm (%); Ntotal = total number of sperm (109 sperm); and Nfunc = number of functional
sperm (109 sperm). The effect was defined as the deviation from the effect of the year 2000.
year, forcing AI centers to keep additional boars to
compensate for these fluctuations. In the investigation
of Grandjot et al. (1997), the greatest values in total
number of sperm occurred, as in our investigation, in
the last quarter of the year. Rutten et al. (2000), Smital
et al. (2004), and Smital (2009) found that the number
of usable doses per collection and the number of functional sperm exhibited clear seasonality with the greatest values from autumn to winter and the least values
from spring to summer. This is in good agreement with
the finding that the main season of rut in the wild boar
is in late autumn (Delcroix et al., 1990; Kozdrowski and
Dubiel, 2004).
Effect of Age at Collection
In agreement with our results, Clark et al. (2003)
reported a dramatic increase in average total sperm
numbers from 8 to 10 mo up to 14 mo of age with little
change thereafter. Smital (2009) also observed a rapid
increase of sperm output with advancing age of the
boar, but in that study, maximum output was achieved
at a later time (3.5 yr of age). The increase of sperm
output with age is probably caused mainly by testis
growth and development.
The results of Rutten et al. (2000) showed that the
number of usable doses per collection increased only
slowly with age, which seemingly is in contradiction
with the results of the 2 papers cited above and with
our investigation. The slow increase may be explained
by the fact that the interval between subsequent collections in the Rutten et al. (2000) experiment decreased
with age, which decreased the influence of the age effect
on total sperm number.
Effect of the Interval Between
Subsequent Collections
In agreement with Rutten et al. (2000), Frangež et al.
(2005), and Smital (2009), our investigation suggests
that a time interval of 7 to 10 d between collections is a
good choice for optimizing all semen traits from the biological point of view. Rutten et al. (2000) investigated
collection intervals from 1 to 10 d and found that the
greatest number of doses per collection can be generated for intervals of 7 to 10 d. Frangež et al. (2005) reported that smaller ejaculate volumes, decreased sperm
concentrations, and decreased total sperm counts per
ejaculate were obtained at collection frequencies of 7
and 3 than at 2 and 1 times per week. Significantly decreased progressive sperm motilities at 7 than at 3, 2,
and 1 time per week collections also were observed.
Though longer intervals yield better results for individual semen traits, economic analysis by Rutten et
al. (2000) showed that the greatest profits could be
achieved from the shortest intervals between subsequent
collections. However, that analysis did not take into account that, in the long-term, high ejaculation frequency
leads to gradual deterioration of the biological value of
the spermatozoa and induces changes in essential indices of semen quality (Strzezek et al., 1995). Those authors concluded that high semen-collection frequencies
stimulate an array of specific biochemical damaging
changes in the spermatozoa that are similar to apoptosis of somatic cells. Pruneda et al. (2005) reported
that a high semen-collection frequency brings about an
altered resorption and secretion pattern of epididymal
fluid, which results in defective sperm maturation and
abnormal development of sperm motility.
Breed and Heterotic Effects
Our finding that Piétrain boars produced semen with
decreased volume and total sperm number, but greater
sperm concentration per ejaculate, than Large White
boars was confirmed by Ciereszko et al. (2000), Smital
et al. (2004), and partially by Smital (2009). The latter
author estimated a decreased total number of sperm
for Large White compared with Piétrain. Furthermore,
the result that Piétrain boars produced semen with
greater volume and number of sperm per ejaculate and
decreased sperm concentration than Duroc boars was
supported by Smital et al. (2004), Kondracki et al.
(2006), and Smital (2009).
Very limited information is available from large data
sets on heterosis for semen traits. The most exhaustive
data on this topic are provided by Smital (2009). In
agreement with our findings, Smital (2009) reported
Breeding value estimation for semen traits
relatively high negative (favorable) heterotic effects for
the percentage of abnormal sperm, especially for crossbreds involving Piétrain.
Environmental and Genetic Trend
Environmental trends for the semen traits were monitored over a relatively short interval of time. Nevertheless, results suggested a tendency for increasing sperm
quantity, but deteriorating sperm quality. To the best
of our knowledge, there are no generally available investigations on the genetic trend for semen traits in swine.
In bull semen, in agreement with our results, no genetic
trend has been observed (Taylor et al., 1985; Ducrocq
and Humblot, 1995; Van Os et al., 1997). In conclusion,
although selection in sire breeds of swine will continue
to be based primarily on improving growth, carcass
merit, and female reproduction, breeding value estimation for semen traits provides a tool for AI centers
wishing to control costs by increasing the number of
doses of semen per boar per unit of time.
LITERATURE CITED
Chemineau, P., B. Malpaux, J. P. Brillard, and A. Fostier. 2007.
Seasonality of reproduction and production in farm fishes, birds
and mammals. Animal 1:419–432.
Ciereszko, A., J. S. Ottobre, and J. Glogowski. 2000. Effects of season and breed on sperm acrosin activity and semen quality of
boars. Anim. Reprod. Sci. 64:89–96.
Clark, S. G., D. J. Schaeffer, and G. C. Althouse. 2003. B-mode
ultrasonographic evaluation of paired testicular diameter of
mature boars in relation to average total sperm numbers. Theriogenology 60:1011–1023.
Colenbrander, B., and B. Kemp. 1990. Factors influencing semen
quality in pigs. J. Reprod. Fertil. Suppl. 40:105–115.
Delcroix, I., R. Mauget, and J. P. Signoret. 1990. Existence of synchronization of reproduction at the level of the social group
of the European wild boar (Sus scrofa). J. Reprod. Fertil.
89:613–617.
Ducrocq, V., and P. Humblot. 1995. Genetic characteristics and evolution of semen production of young Normande bulls. Livest.
Prod. Sci. 41:1–10.
Frangež, R., T. Gider, and M. Kosec. 2005. Frequency of boar ejaculate collection and its influence on semen quality, pregnancy
rate and litter size. Acta Vet. (Brno) 74:265–273.
Grandjot, G., H. Brandt, and P. Glodek. 1997. Genetische und
phänotypische Untersuchungen zu Eigenleistungs-, Spermaund Fruchtbarkeitsmerkmalen von Besamungsebern. 1. Mitteilung: Systematische Einflußfaktoren auf genetische Parameter.
Arch. Tierz. 40:421–432.
1627
Groeneveld, E., M. Kovac, and T. Wang. 1990. PEST, a general purpose BLUP package for multivariate prediction and estimation.
Proc. 4th World Congr. Genet. Appl. Livest. Prod., Edinburgh,
Scotland XIII:488–491.
Kondracki, S., A. Wysokińska, D. Banaszewska, and J. Zajda. 2006.
Breed-related variation in ejaculate traits of AI boars. Anim.
Sci. Pap. Rep. 24(Suppl. 3):121–129.
Kovač, M., E. Groeneveld, and L. A. García-Cortés. 2002. VCE-5
package for the estimation of dispersion parameters. Proc. 7th
World Congr. Genet. Appl. Livest. Prod., Montpellier, France
33:741–742.
Kozdrowski, R., and A. Dubiel. 2004. The effect of season on the
properties of wild boar (Sus scrofa L.) semen. Anim. Reprod.
Sci. 80:281–289.
Neumaier, A., and E. Groeneveld. 1998. Restricted maximum likelihood estimation of covariances in sparse linear models. Genet.
Sel. Evol. 30:3–26.
Oh, S. H., M. T. See, T. E. Long, and J. M. Galvin. 2006a. Estimates of genetic correlations between production and semen
traits in boar. Asian-australas. J. Anim. Sci. 19:160–164.
Oh, S. H., M. T. See, T. E. Long, and J. M. Galvin. 2006b. Genetic
parameters for various random regression models to describe
total sperm cells per ejaculate over the reproductive lifetime of
boars. J. Anim. Sci. 84:538–545.
Pruneda, A., E. Pinart, M. D. Briz, S. Sancho, N. Garcia-Gil, E. Badia, E. Kadar, J. Bassols, E. Bussalleu, M. Yeste, and S. Bonet.
2005. Effects of a high semen-collection frequency on the quality of sperm from ejaculates and from six epididymal regions in
boars. Theriogenology 63:2219–2232.
Robinson, J. A. B., and M. M. Buhr. 2005. Impact of genetic selection on management of boar replacement. Theriogenology
63:668–678.
Rutten, S. C., R. B. Morrison, and D. Reicks. 2000. Boar stud production analysis. Swine Health Prod. 8:11–14.
Smital, J. 2009. Effects influencing boar semen. Anim. Reprod. Sci.
110:335–346.
Smital, J., L. L. De Sousa, and A. Mohsen. 2004. Differences among
breeds and manifestation of heterosis in AI boar sperm output.
Anim. Reprod. Sci. 80:121–130.
Smital, J., J. Wolf, and L. L. De Sousa. 2005. Estimation of genetic
parameters of semen characteristics and reproductive traits in
AI boars. Anim. Reprod. Sci. 86:119–130.
Strzezek, J., W. Kordan, J. Glogowski, P. Wysocki, and K. Borkowski. 1995. Influence of semen-collection frequency on sperm quality in boars, with special reference to biochemical markers. Reprod. Domest. Anim. 30:85–94.
Taylor, J. F., B. Bean, C. E. Marshall, and J. J. Sullivan. 1985.
Genetic and environmental components of semen production
traits of artificial insemination Holstein bulls. J. Dairy Sci.
68:2703–2722.
Van Os, J. L., M. J. De Vries, N. H. Den Daas, and L. M. K. Lansbergen. 1997. Long-term trends in sperm counts of dairy bulls.
J. Androl. 18:725–731.
Wolf, J. 2008. Genetic parameters for semen traits in AI boars estimated from data on individual ejaculates. Reprod. Domest.
Anim. doi:10.1111/j.1439-0531.2008.01083. In press.