BREEDING & GENETICS                               PIH-39


     Crossbreeding Systems for Commercial Pork Production

William T. Ahlschwede, University of Nebraska
Charles J. Christians, University of Minnesota
Rodger K. Johnson, University of Nebraska
O. W. Robison, North Carolina State University

Russell and Jane Clark, Frankfort, Indiana
Walter A. Gross, University of Rhode Island
James and Shirley Mitchell, Eaton, Ohio
Irvin Omtvedt, University of Nebraska

     Years of experience and  research  show  that  crossbreeding
pays. The hybrid vigor gained through crossbreeding improves per-
formance of both the breeding herd and individual  pigs.  Surveys
indicate  that  more  than  90%  of market hogs in the U.S.A. are
crossbreds. However, crossbreeding is not  sufficient  to  assure
profitable  performance. Rather, specifying which breeds to cross
and how to cross them is needed to describe a crossbreeding  sys-
tem.  Choices among crossbreeding systems can dramatically affect
profit levels.

|Table 1. Heterosis advantage for production traits.               |
|_________________________________________________________________ |
|                            First       Multiple                  |
|                            cross         cross                   |
|                          purebred      crossbred     Crossbred   |
|Item                         sow           sow           boar     |
|_________________________________________________________________ |
|                            Percentage advantage over purebred    |
|Reproduction                                                      |
|  Conception rate            0.0           8.0           10.0     |
|  Pigs born alive            0.5           8.0            0.0     |
|  Littersize 21 days         9.0          23.0            0.0     |
|  Littersize weaned         10.0          24.0            0.0     |
|Production                                                        |
|  21-day litter weight      10.0          27.0            0.0     |
|  Days to 220 lb.            7.5           7.0            0.0     |
|  Feed/gain                  2.0           1.0            0.0     |
|Carcass composition                                               |
|  Length                     0.3           0.5            0.0     |
|  Backfat thickness         -2.0          -2.0            0.0     |
|  Loin muscle area           1.0           2.0            0.0     |
|  Marbling score             0.3           1.0            0.0     |

     The value of crossbreeding depends upon hybrid vigor. Hybrid
vigor, or heterosis, is the superiority of the crossbred compared
to its parental breeds.  In pigs, hybrid vigor appears  important
for pig survival and growth, litter size and mothering ability in
sows and gilts and boar breeding performance (Table  1).  Differ-
ences  in  performance  among  breeds  can  be  utilized  by some
crossbreeding systems to take advantage of the best attributes of
breeds while minimizing the impact of their deficiencies.

     Crossbreeding became  a  common  practice  as  a  result  of
crossbreeding  research  in the 1930's and 1940's. Crossbred off-
spring of purebred parents were indeed  superior.  Following  the
lead  of  corn growers, crossbred females became the base of com-
mercial production. However, with  pigs,  development  of  inbred
lines  did  not  prove feasible. Rotational crosses, which nicely
fit the most common  styles  of  production  in  the  1950's  and
1960's, were generally adopted.  Changes in the way hogs are pro-
duced have led to increased adoption of terminal crosses  in  the

Rotational Crosses

     Two general types of crossbreeding  systems  are  described.
Rotational  crossbreeding  systems utilize replacement gilts from
the market crosses and change breeds of boar each  generation.  A
three-breed-rotation  uses  three  breeds  of  boars,  rotated in
order, one breed per generation (Figure 1).  A  rotation  of  two
breeds  is  called  a  criss-cross. Rotations of up to six breeds
have been used to advantage.

Table 2. Heterosis percentage in rotational crosses.
                              Generation number          Equilib-
Crossbreeding system    1     2     3     4     5    6     rium
Two breed rotation
  ``crisscross''      100.0  50.0  75.0  62.5  68.9 67.2   66.7
Three breed rotation  100.0 100.0  75.0  87.5  87.5 84.4   85.7
Four breed rotation   100.0 100.0 100.0  87.5  93.8 93.8   93.3
Five breed rotation   100.0 100.0 100.0 100.0  93.8 96.9   96.8
Six breed rotation    100.0 100.0 100.0 100.0 100.0 96.9   98.4
Crisscross with F1
  boars (four breeds) 100.0  75.0  87.5  81.3  84.4 82.3   83.3

     Rotations using crossbred boars  take  advantage  of  hybrid
vigor  in  boar  breeding  ability. After the first full round of
breeds in a rotation, some heterosis is lost. To the degree  that
genes  of  the  breed of the service boar are present in the sow,
heterosis is reduced. As shown in Table 2, rotations of three  or
more  breeds  retain relatively high levels of heterosis. Serious
losses of heterosis in rotations occur when the planned order  of
breed  use  is  not  followed. This is likely when generations of
sows are not kept separate or when color is used as an  indicator
of sow breed.

     The level of heterosis in advanced generations  of  a  rota-
tional  cross  depends upon the number of breeds in the rotation.
As shown in Table 2, initial crosses all express 100%  heterosis.
With  succeeding  generations, the heterosis level modulates (the
first six generations are shown) until an equilibrium is reached.
The  heterosis  equilibrium, characteristic for each rotation, is
shown in the right hand column of Table 2.

     Rotational crosses are general purpose crosses. Offspring of
each generation are used for market production and as replacement
gilts. Since breeds appear first as the sire of the  market  hogs
and  a generation later as the sire of the sows, performance both
as market hogs and as breeding stock is  considered  in  choosing
breeds.  There  is  little  opportunity  to take advantage of the
differences among the breeds used in the  cross.  Performance  in
rotations  would  be maximized if there were no breed differences
in performance.  We would want them to all be as good as the best
in every trait.

Terminal Crosses

     Terminal crosses take their name from the fact that the sys-
tem  terminates  with the market cross. Replacement gilts are not
saved from the matings designed for market production.  A  simple
terminal cross would be a Yorkshire sow mated to a Hampshire boar
with all of the offspring marketed. Replacement  gilts  are  pro-
duced  by  matings  other than the market crosses. Because breeds
are used in specialized roles, (in this  example,  the  Hampshire
boar  sires  market  hogs  but  not  replacement  gilts) terminal
crosses can take advantage of breed differences.

     A more common terminal cross is mating Landrace x  Yorkshire
first  cross  (F1)  females to Hampshire x Duroc first cross (F1)
boars (Figure 2). The sow in this case is  a  cross  between  two
breeds  which have reputations as good sows.  The sire is a cross
of two breeds with reputations for efficient lean gain and  fast,
efficient  gain. In this way, terminal crosses allow breeds to be
used in roles that  take  advantage  of  their  strengths,  while
minimizing their deficiencies. This allows terminal crosses to be
more productive than rotational crosses. In this example, maximum
use  of  heterosis  is  realized.   The boar and the sow are both
first crosses with 100% heterosis, important  for  breeding  per-
formance,  littersize,  and maternal performance. Since there are
no common breeds in the parents, the  offspring  also  have  100%
heterosis, important for piglet vigor, survival, and rapid gain.

     In terminal crossbreeding systems, replacement gilts are not
selected  from  the market crosses. They must either be purchased
or produced in special matings. Because of this, terminal crosses
were  generally  not  used by commercial pork producers. Histori-
cally, rotational crosses fit the  production  systems.  However,
production  systems  have  changed. Many intensively managed pork
producing units today are able to program the special  gilt  pro-
ducing  matings  into their schedule without difficulty. Terminal
crosses are becoming more popular.

     Opportunities  exist  to  combine  rotational  and  terminal
crossbreeding   systems.  Many  producers  have  used  rotational
crosses among breeds that excel in maternal performance  to  pro-
duce  replacement  gilts  for  terminal  matings.  These systems,
called rotaterminals, combine the ease of operation of a rotation
with   the  specialization  and  high  productivity  of  terminal

Crossbred Boars

     The use of crossbred boars has sparked discussion  and  con-
troversy in the industry in recent years. In the mid 1970's their
use was sufficient to stimulate stories in the popular press.  By
the  late  1970's, research trials had been conducted that clari-
fied their attributes. Crossbred boars were found to excel  pure-
bred  boars in breeding performance as measured by farrowing rate
of sows and gilts to which they  were  exposed.  They  were  more
vigorous,  active breeders. While this advantage may appear small
(Table 1), timely pregnancies are crucial to intensively  managed
pork producing operations. In addition to the improved conception
rates identified in research trials,  commercial  pork  producers
report  fewer  problems  and  longer  useful lives with crossbred

     Crossbred boars can be part of a crossbreeding system. Pure-
bred  boars,  crossbred  boars,  and combinations of purebred and
crossbred  boars  are  being  used  in   productive,   profitable
crossbreeding  systems,  which maintain high levels of heterosis.
The availability of both  purebred  and  crossbred  boars  allows
increased  flexibility  in  planning  and operating crossbreeding
systems.  In the discussion that follows, both types of boars are
used.  Whether  purebred or crossbred boars are used, the success
of the breeding systems depends upon following the system.

Evaluating and Choosing
Crossbreeding Systems

     One would like to choose the crossbreeding system that would
provide  the  most  favorable economic outcome. However, the many
factors involved in determining the outcome of a system  make  it
difficult to effectively evaluate.  Farm objectives vary, facili-
ties differ, and desired breeds and crosses are not always avail-

     Recent studies using computers to model  crossbreeding  sys-
tems  have helped clarify the differences among systems. At least
five separate computer programs which analyze crossbreeding  sys-
tems  have  been  utilized  in  these  studies.  The computerized
models, some of which are very complex, take into  account  breed
differences  for  each of the traits studied, the mode of inheri-
tance  for  specific  traits,  relationships  among  traits,  the
economic  impact  of  each  of  the  traits, and market prices of
inputs and production.  These studies offer guidance in  choosing
among various systems.

Table 3. Expected outcome for crossbreeding sytems.
                        Proportion Percent    Pigs   Conception
Mating                   matings  offspring marketed    rate    
systemSire     Dam      in system heterosisper litter    %      
Purebred matings
      HampshireHampshire              0       5.94       85     
      Duroc    Duroc                  0       6.34       85     
      YorkshireYorkshire              0       7.78       72     
      Landrace Landrace               0       8.40       69     
      Chester  Chester                0       7.35       85     
Purebred average                                                
1  Three breed rotation
      HampshireYxD,H..     33.3     85.7      8.56       80     
      Duroc    HxY,D..     33.4     85.7      7.96       83     
      YorkshireDxH,Y..     33.3     85.7      8.02       85     
System average, weighted by proportion matings in system        
2  Three breed terminal
      YorkshireYorkshire    5         0       7.78       72     
      Duroc    Yorkshire    15       100      8.08       72     
      HampshireDxY          80       100      8.81       81     
System average, weighted by proportion matings in system        
3  Four breed terminal
      YorkshireYorkshire    5         0       7.78       72     
      Landrace Yorkshire    15       100      8.08       72     
      HxD      LxY          80       100      9.55       87     
System average, weighted by proportion matings in system        
4  Rotaterminal with two breed sow
      YorkshireLxY,L..     7.5      66.7      9.10       72     
      HxD      YxL,Y..     42.5      100      9.08       87     
      Landrace YxL,Y..     7.5      66.7      9.34       72     
      HxD      LxY,L..     42.5      100      9.11       86     
System average, weighted by proportion matings in system        
5  Rotaterminal with three breed sow
      YorkshireCW,L,Y..     5       85.7      9.18       81     
      HxD      Y,CW,L..     29       100      9.22       89     
      Landrace Y,CW,L..     5       85.7      9.55       78     
      HxD      L,CW,Y..     28       100      9.30       88     
      Chester  L,CW,Y..     5       85.7      9.32       75     
      HxD      CW,L,Y..     28       100      9.12       90     
System average, weighted by proportion matings in system        
6  Rotaterminal with two-way, three breed sow
      YorkshireDL,Y,DL..   7.5      66.7      8.81       88     
      HampshireY,DL,Y..    42.5      100      8.81       76     
      DxL      Y,DL,Y..    7.5      83.3      8.93       88     
      HampshireDL,Y,DL..   42.5      100      8.83       78     
System average, weighted by proportion matings in system        
7  Two way rotation with four breeds
      HxL      DY,HL..      50      83.3      8.61       90     
      DxY      HL, DY..     50      83.3      8.41       90     
System average, weighted by proportion matings in system        
8  Three-way rotation with four breeds
      HxL      Y,D,HL..    33.3     92.9      8.93       89     
      Duroc    HL,Y,D..    33.4     85.7      8.61       79     
      YorkshireD,HL,Y..    33.3     85.7      8.30       83     
System average, weighted by proportion matings in system        

Table 3. (Continue..)
                         Days  Fat  Growing-   Net
Mating                    to  thick-finishing  per
systemSire     Dam      market ness    F/G    litter
Purebred matings
      HampshireHampshire 183   1.00   3.30   ($19.62)
      Duroc    Duroc     172   1.20   3.33   ($16.53)
      YorkshireYorkshire 177   1.20   3.35    $23.97
      Landrace Landrace  180   1.25   3.40    $33.05
      Chester  Chester   185   1.30   3.45    ($4.25)
1  Three breed rotation
      HampshireYxD,H..   169   1.10   3.26    $75.94
      Duroc    HxY,D..   165   1.16   3.27    $51.29
      YorkshireDxH,Y..   166   1.19   3.28    $50.61
2  Three breed terminal
      YorkshireYorkshire 177   1.20   3.35    $23.97
      Duroc    Yorkshire 162   1.22   3.27    $46.35
      HampshireDxY       166   1.12   3.25    $86.24
3  Four breed terminal
      YorkshireYorkshire 177   1.20   3.35    $23.97
      Landrace Yorkshire 166   1.25   3.31    $51.58
      HxD      LxY       166   1.19   3.28   $107.28
4  Rotaterminal with two
      YorkshireLxY,L..   170   1.23   3.32    $73.48
      HxD      YxL,Y..   165   1.18   3.27    $91.07
      Landrace YxL,Y..   171   1.25   3.34    $78.29
      HxD      LxY,L..   166   1.19   3.28    $90.07
5  Rotaterminal with thr
      YorkshireCW,L,Y..  169   1.26   3.33    $77.61
      HxD      Y,CW,L..  166   1.19   3.28    $94.57
      Landrace Y,CW,L..  169   1.26   3.33    $87.42
      HxD      L,CW,Y..  166   1.19   3.29    $96.28
      Chester  L,CW,Y..  171   1.29   3.36    $71.28
      HxD      CW,L,Y..  167   1.21   3.30    $87.25
6  Rotaterminal with two
      YorkshireDL,Y,DL.. 168   1.22   3.31    $68.58
      HampshireY,DL,Y..  167   1.13   3.26    $82.53
      DxL      Y,DL,Y..  166   1.24   3.30    $77.77
      HampshireDL,Y,DL.. 167   1.13   3.26    $83.31
7  Two way rotation with
      HxL      DY,HL..   169   1.17   3.29    $69.27
      DxY      HL, DY..  167   1.19   3.29    $64.94
8  Three-way rotation wi
      HxL      Y,D,HL..  167   1.18   3.28    $84.84
      Duroc    HL,Y,D..  165   1.20   3.28    $68.96
      YorkshireD,HL,Y..  166   1.21   3.29    $57.63

Economic projections based on unit farrowing  100  litters  at  a
base cost of $300 per litter.

  Base conception rate of 80%, herd cost adjusted by $28 per  sow
  above or below 80%.

  Base litter size marketed was 7.5 pigs, litter cost adjusted by
  $8 per pig above or below base.

  Base age to market at 220 pounds was 180 days. Nonfeed costs of
  $14 per pig, adjusted by $0.05 per day above or below base.

  Feed for growing-finishing charged at $140 per ton.

  Hogs were marketed at 220 pounds at a base  price  of  $45/cwt.
  with  1.15  in.  last rib fat thickness.    Premiums and discount
  based on fat thickness were 1% for each 0.1 in.  below  or  above
  base value.

Table  4.  Performance  averages  assigned  to  breeds  used   in
crossbreeding systems analysis procedures.
               Concep- Litter  Piglet Days to    Fat     Feedlot
Breed         tion rate size  survival220 lb. thickness feed/gain
                  %     no.      %               in.
Hampshire        85      9.0     66     183      1.0      3.30
Duroc            85      9.6     66     172      1.2      3.33
Yorkshire        72     10.8     72     177      1.2      3.35
Landrace         69     10.0     84     180      1.3      3.40
Chester White    85     10.5     70     185      1.3      3.45

Table values refer to purebred performance.

(Source: NC-103 research project  publication  and  test  station
Johnson, R.K. 1980. Heterosis and Breed Defects in Swine.   North
Central Regional research publication No. 262.

     Results from one of the crossbreeding systems analyses  pro-
grams  mentioned  above are shown in Table 3. Matings are grouped
into systems. For each type of mating, projections of performance
for  six traits were made using the breed averages shown in Table
4 and the heterosis values given in Table 1. An expected economic
outcome  for  each  mating type was also calculated. The economic
system represents conditions experienced  in  the  early  1980's,
with credit given for faster and more efficient gain, higher con-
ception rates, larger litters weaned,  and  for  leaner  hogs  at
market.  An  expected outcome for each system was calculated as a
weighted average of the mating outcomes, weighted by the  propor-
tion  of  the  mating  in  the  system. With a $45 per cwt market
price, purebred performance with Hampshires, Durocs,  Yorkshires,
Landrace and Chester Whites averaged $3.32 per litter profit when
market hogs were sold at 220 lb.  (first  section  of  Table  3).
Credit for leaner pigs at market was based upon the National Pork
Producers Council Lean Value System.

     The analysis system assumes that all boars are purchased and
replacement  gilts  are valued as market hogs. Boars are taken to
be of average genetic merit. By including the gilt producing mat-
ings  in the system average, replacement gilt costs are accounted

     THREE  BREED  ROTATION-The  three   breed   rotation   using
Hampshire,  Duroc, and Yorkshire is shown as System 1 in Table 3.
These projections suggest $63 per litter advantage over the aver-
age  of  the purebred performance of the three breeds involved in
the cross. The net return as purebreds  of  Hampshires  ($19.62),
Durocs  ($16.53),  and  Yorkshires ($23.97) averaged a $4.06 loss
per litter; the three generations  of  the  rotation  averaged  a
profit  of  $59.27.   Three  generations  are shown, representing
advanced generations of the rotation.  Gilt  offspring  from  one
generation are used as the sows in the following generation.

     With rotational crosses, the breed composition changes  each
generation.   After  several  generations,  57% of the genes come
from the last sire breed used, 29% from  the  grandsire  and  14%
from  the great grandsire. If the breeds in the rotation are dif-
ferent for important  traits,  wide  swings  in  performance  and
profit  can  be  expected  from generation to generation. This is
apparent in Table 3. Littersize raised is considerably higher  in
the Hampshire sired generation because of the Yorkshire influence
in the sow, and the pigs are leaner. These differences led to  an
expected  economic  advantage  of  $25 more profit per litter for
this generation. Swings in performance and profit of this  magni-
tude are expected with rotational crosses.

     TERMINAL CROSSES-System 2 in Table 3 is a three-breed termi-
nal  cross, with Duroc-Yorkshire F1 sows mated to Hampshire boars
to produce the market hogs.  These matings are on the third  line
of  the  system.  The  second  line  is  the matings that produce
replacement gilts, Yorkshire sows bred to Duroc boars.  The first
line  in  the system shows the matings that produce the Yorkshire
gilts used as sows in the second line. The average for the system
is  calculated  on  5%  purebred  Yorkshire matings (line 1), 15%
Duroc x Yorkshire matings (line 2) and 80% terminal matings (line
3).  This  system  projects  an advantage of about $18 per litter
over the rotational cross in System 1. The same breeds  are  used
in  both  systems.  The  advantage  is  due  to  higher levels of
heterosis in both sows and pigs, the beneficial effect on litter-
size of a high percentage of Yorkshire in the sows of all matings
and the improved carcass value due to the Hampshire sire  in  the
terminal crosses.

     System 3 in Table 3 is a four breed terminal using  both  F1
sows  and  F1  boars to produce the market crosses. The system is
similar to System 2, except that Landrace-Yorkshire F1  sows  are
used and F1 Hampshire-Duroc boars sire the terminal crosses. This
system combines two good maternal breeds  in  the  sow  producing
most  of  the  pigs  and takes advantage of the heterosis in boar
breeding ability in the terminal matings. This system projects  a
$35 per litter advantage over the rotation in System 1. The addi-
tional $17 per litter advantage over the three breed terminal  is
due  to  the inclusion of a second superior maternal breed in the
sow and the improved conception rate  because  of  the  crossbred

     ROTATERMINAL SYSTEMS-Rotaterminal systems are shown as  Sys-
tem  4,  5, and 6 in Table 3. In these systems, replacement gilts
are produced with rotational  crosses.  These  females  are  then
mated to terminal boars for market production.  These systems are
run similarly to rotations, except that most of the matings  pro-
duce  only  market  crosses. Depending upon the choice of service
sire, any sow can produce  either  replacement  gilts  or  market
hogs.  Production in the terminal matings of the rotaterminal are
generally not as profitable as in the analogous terminal  matings
in  System  3  primarily because maternal heterosis is lower. For
the system, some of this difference is made up because gilt  pro-
duction is less expensive, being based on crossbred sows.

     System 4 is a rotaterminal  based  on  a  Yorkshire-Landrace
crisscross.  The  first line in the system shows the matings pro-
ducing Yorkshire sired replacement gilts. The gilts  produced  in
this mating appear in the second line as the sow producing termi-
nal crosses by a Hampshire-Duroc F1 boar, and in the  third  line
of  the  system  mated  to  Landrace boars to produce replacement
gilts.  The Landrace sired gilts are used as the sow for terminal
matings  in line 4 of this system and as the mothers of Yorkshire
sired gilts in line 1. This system projects a $30 a litter advan-
tage  over  the three breed rotation. While the breed composition
of System 4 is similar to System 3, the reduced level of maternal
heterosis  in  terminal matings (66.7% vs. 100%) accounts for the
$6 per litter disadvantage for the system.

     System 5 and System  6  are  rotaterminals  that  use  three
breeds  in  the sow.  System 5 is based on a three breed rotation
of Yorkshire, Landrace and Chester White. The increased  maternal
heterosis  and  individual  heterosis  in  gilt producing matings
accounts for the $2.26 per litter advantage over System 4.   How-
ever, an additional breed of boar and type of mating is required.
System 6 also uses three breeds in the sow but relies on only two
types of boars to produce replacement gilts. Sows are produced by
a crisscross between the Yorkshire  and  the  Landrace-Duroc  F1.
Compared  to  System  4,  this gives a sow with less Landrace and
increases maternal heterosis in half of the matings.  The reduced
littersize  of  the  Duroc  and  the lower conception rate of the
purebred terminal sire reduce the expected  net  income  of  this
System. Since the Duroc is reputed to be better adapted to exten-
sive production conditions than the Landrace, the sow in System 5
would  be classified as a hardier sow than those in System 4, and
might be better adapted to extensive production systems.

     ROTATIONS WITH CROSSBRED BOARS-The last section of  Table  3
shows  two  rotations  using  crossbred  boars.  System  7  is  a
crisscross among four breeds using two types of  F1  boars.  This
gives  the  breed  balance  found in the four breed terminal. The
heterosis level in the sow and pig is similar to  a  three  breed
rotation.  This  system  projects  nearly $8 per litter advantage
over the three breed rotation. The maternal contribution  of  the
Landrace  and improved conception rate of crossbred boars account
for the difference.

     System 8 is a three way rotation, using  purebred  Yorkshire
and  Duroc  boars  along with an F1 Hampshire-Landrace boar. This
system attempts to balance white and  colored  breed  influences,
while  reducing some of the generation to generation fluctuations
in  performance  characteristics  of  rotational  crosses.    The
increased  level  of  heterosis accounts for the $3.37 per litter
advantage over System 7.

     Large differences in performance and profit potential  among
the  crossbreeding  systems  are  indicated in Table 3. While the
numbers in Table 3 are produced by computer simulation, the  cal-
culations  are  based  upon results of real breeding experiments.
The budget used was based on costs reported by pork producers who
keep  records. Further, the results reported here seem to closely
follow the experiences of producers who use the various systems.

     Large  differences  in  projected  profits  should  not   be
translated  directly into a recommendation that all producers use
the mating system with  the  largest  estimated  profit  reported
here.  The  projections  shown are for a small number of possible
crossbreeding systems. The systems shown were  chosen  to  illus-
trate  differences  and  represent  the most common applications.
However, it appears that the advantages of the  terminal  systems
can be realized by most pork producers.

     Rotation crosses have generally been recommended for commer-
cial  production  since  the  advantages  of  crossbred pigs were
demonstrated in the 1940's. The attributes  of  terminal  crosses
were  known, but production methods favored the use of rotations.
Seasonal farrowing coupled with replacement of the whole breeding
herd at one time made rotations easy to use. Extensive production
systems minimized the advantage of specialized  maternal  breeds.
Small  production units did not allow subdivision for replacement
gilt production.   And  health  programs  were  not  sufficiently
strong  to  encourage  purchasing replacement gilts. All of these
factors are different today. Terminal crossbreeding systems  more
closely fit the production systems being used for commercial pro-
duction than do the rotations.

     Operating rotational crosses properly has  become  difficult
on most pork producing farms. Rotational crosses save replacement
gilts from the market crosses, changing the breed  of  boar  each
generation  in  a  prescribed  order.  With multiple farrowing in
modern facilities, it is desirable  to  replace  cull  sows  with
gilts  after  each  farrowing. The gilt, being of the new genera-
tion, is to be bred to the next breed of boar  in  the  rotation.
Many  producers  lack  the breeding pens, sow identification, and
personnel necessary to assure that sows are bred to  the  correct
breed of boar. Too often the replacement gilt is bred to the same
breed of boar used to breed the group of sows she will join,  the
breed of boar of her sire. These mistakes (the mating of sows and
gilts to boars of the breed of their sire) are common and costly.
In  a  rotational  system using purebred boars, backcross matings
cut heterosis in half. In a three breed  rotation,  heterosis  is
cut from 85.6% to 42.8%, with accumulated production losses which
average $40 per backcross mating.

     The  breeding  management  required  to  operate  rotational
crosses  on intensively managed pork producing farms will support
terminal crossbreeding systems as well. With a three breed  rota-
tion,  the  sows  must  be identified according to their breed of
sire and provision made, either through adequate breeding pens or
hand  mating, to breed with three breeds of boars simultaneously.
The same is required of the  terminal  and  rotaterminal  systems
described  in Table 3. The operational advantage of the rotations
observed in the 1940's is generally absent  today.  In  addition,
production systems are in place which can take advantage of breed
differences, particularly in sow productivity.

     The key to the operation of terminal  crossbreeding  systems
is  the  acquisition of sows. Unlike rotations, replacement gilts
are not kept  from  matings  designed  to  produce  market  hogs.
Replacement  gilts are produced in special matings. The high lev-
els of productivity experience with gilts  and  sows  from  these
special  gilt  producing  matings give the terminal crossbreeding
systems  their  advantage.  Operationally,  the   production   of
replacement gilts is the key to the terminal crosses.

     A  second  determiner  of  the  success  of   the   terminal
crossbreeding  systems is the ability to meet the nutritional and
environmental needs of the more productive sow herd. Feeding pro-
grams and housing systems that supported rotational crosses might
not sustain performance with terminal crosses, which wean 1 or  2
pigs  more  per litter. Adjustments in the care and management of
the breeding herd are in order.

     If the pork producing industry  followed  the  lead  of  the
poultry  industry,  replacement  females  would be purchased. The
practice of buying replacement gilts has expanded during the last
decade.  Buying  gilts is functional for every farm that farrows.
However, questions of health  risk  and  cost  must  be  answered
before  judgments of feasibility can be made. Generally, the mat-
ings that produce replacement gilts for terminal crosses are  not
as  productive as the terminal crosses. Hence, the purchase price
represents, in part, the lost productivity in the breeder's herd.

     Most pork producers who farrow  have  preferred  to  produce
their  own  replacement  gilts. With rotational crosses, that was
built into the system.  With terminal  crosses,  special  matings
are  required.  As an example, if Landrace-Yorkshire F1 gilts are
desired,  purebred  Yorkshire  sows  and  a  Landrace  boar   are
required. Both the number and timing of these matings are coordi-
nated with the production schedule so that  gilts  are  available
when  needed.   Yorkshire gilt replacements for these matings can
either be purchased or produced. If they are produced,  an  addi-
tional  type  of  mating  is required.  Yorkshire sows mated to a
Yorkshire boar are needed. While it is possible to perform  these
matings, the small breeding groups, the need to schedule matings,
and the performance differences experienced  with  purebred  sows
and F1 animals complicate management of the system.

     The rotaterminal systems attempt to make the job of  produc-
ing  home  raised  gilts  easier.  In  the case of the Yorkshire-
Landrace crisscross for replacement  gilts,  both  Yorkshire  and
Landrace  boars are needed to sire replacement gilts.  The advan-
tage of this system is that all of the sows in the herd are simi-
lar.  They are all Yorkshire-Landrace crosses with similar levels
of heterosis and performance characteristics.  All  sows  in  the
herd  are potential mothers of replacement gilts. For replacement
gilts, Yorkshire sired sows are bred to a Landrace boar, or Land-
race  sired sows are bred to a Yorkshire boar. Most sows are bred
to a terminal boar for market production. This system  allows  an
easier flow of replacement gilts into the system, because sows to
produce them are always available. And, because all of  the  sows
are  similar  crossbreds, matings producing replacement gilts are
nearly as productive as market crosses.  The disadvantage is that
the  sow  herd  expresses  only 66.7% of the available heterosis.
This accounts for the difference in profit projected between  the
two systems.

     The competitiveness of the  terminal  crossbreeding  systems
depends  in  part upon the proportion of matings committed to the
production of replacement gilts. The number of  matings  required
to  produce  replacements  depends  upon  the  litter size in the
gilt-producing litters, the number of gilts needed (controlled by
culling  in the sow herd), and the proportion of replacement gilt
candidates selected. The proportion of each  type  of  mating  in
each of the crossbreeding systems in Table 3 was assigned to pro-
vide ample replacement gilts. In System 3, reducing  the  propor-
tion  of  nonterminal  matings to 10% would increase the expected
net for the system by $5 per litter. In the rotaterminal systems,
where the gilt producing litters are more productive, the propor-
tion of replacement gilt producing matings is not as critical.

     The proportion of litters producing replacement gilts for  a
terminal system is an operational characteristic of a farm. It is
suggested that initial proportion of litters  producing  replace-
ment  gilts be chosen to supply an excess of gilts. As the opera-
tion becomes familiar with the attributes of the sows and  deter-
mines a functional culling rate, the proportion of gilt producing
matings can be reduced. Since the cost of too  many  home  raised
replacement  gilts  is  generally  less than the cost of too few,
starting with plenty is suggested.

Breed Evaluation

     The genetic  merit  of  the  breeds  used  in  crossbreeding
represents  the  foundation  upon  which  performance  is  based.
Heterosis effects and the breeding values of  parents  are  addi-
tions  to  this  base  in  determining final performance.  Hence,
choice of breed can have a marked  effect  on  profitability.  As
demonstrated   in   Table  3,  placement  of  the  breed  in  the
crossbreeding system is also important. Knowledge of  breed  per-
formance  is  essential  in  planning  crossbreeding  systems and
predicting performance.

     Notions and estimates of breed performance  come  from  many
places.  The  most  reliable  estimates  come  from crossbreeding
experiments where several breeds and crosses are tested together.
In  these  experiments, attempts are usually made to include pigs
from many sources and lines within the breeds.  Table 4 is a sum-
mary  of  the performance of five breeds adapted from the evalua-
tion compiled by project NC-103, a committee of the swine  breed-
ing researchers in the U.S.A. A more general summary of the find-
ings of the NC-103 committee is in Table 5.

|                                                                  |
|Table 5. Relative performance of breeds.                          |
|_________________________________________________________________ |
|                 Concep-   Litter size   21-day   Age at    Back- |
|Breed           tion rate    raised      weight   220 lb.    fat  |
|_________________________________________________________________ |
|Berkshire           +           -          -                      |
|Chester White       +          ++          -        --        A   |
|Duroc               A           A          -         +        -   |
|Hampshire           A           -          A         -       ++   |
|Landrace           --          ++          ++        A       --   |
|Poland                                               A        +   |
|Spotted                        --          --        +        -   |
|Yorkshire           -          ++          +         +        -   |
|_________________________________________________________________ |
|                                                                  |
|Based on NC-103 review.                                           |
|Blank cell indicates data unavailable.                            |
|A indicates performance near average of breeds studied.           |
|+ indicates performance superior to average.                      |
|++ indicates performance substantially superior to average.       |
|- indicates performance inferior to average.                      |
|-- indicates performance substantially inferior to average.       |

Table 6. Breed averages of purebred barrows tested  at  the  1985
and 1986 National Barrow Show.
                        ADG     Age at    10th rib    Loin muscle
Breed           No.   lb./day   220 lb.   backfat        area
                                 days       in.         sq. in.
Berkshire       47     1.66     176.40      1.15         4.53
Chester White   66     1.64     175.94      1.26         4.50
Duroc           161    1.73     171.89      1.22         4.72
Hampshire       75     1.60     177.53      0.91         5.22
Landrace        34     1.65     172.00      1.22         4.49
Poland          47     1.64     173.47      1.12         5.17
Spotted         65     1.68     171.98      1.20         4.72
Yorkshire       93     1.61     178.24      1.15         4.51

|                                                                  |
|Table 7. Performance of boars in  central  test  stations  during |
|1984, 1985, and 1986.                                             |
|_________________________________________________________________ |
|                                 ADG          Feed/       Backfat |
|Breed               No.        lb./day        gain          in.   |
|_________________________________________________________________ |
|Berkshire           349         2.10          2.68         0.83   |
|Chester White       326         2.08          2.59         0.82   |
|Duroc               4294        2.27          2.49         0.81   |
|Hampshire           1587        2.17          2.49         0.74   |
|Landrace            685         2.20          2.53         0.78   |
|Poland China        219         2.11          2.68         0.80   |
|Spotted             801         2.09          2.63         0.79   |
|Yorkshire           4028        2.22          2.49         0.79   |

     Another source of comparative information is test  stations,
including  barrow  tests.  Since breeders choose the pigs tested,
these results are less reliable. Table 6 is a  summary  of  breed
performance  based  upon results of the National Barrow Show per-
formance test of market hogs. Table 7 is based upon results  from
boar  testing  stations.  Word  of mouth reports of how a certain
kind of pig performed are also abundant. Since these reports sel-
dom  are  based  upon comparative information, they are of little

     Breed averages as presented in these tables have the appear-
ance of accuracy and stability. However, there are limitations to
the usefulness of the numbers. The  NC-103  estimates  are  based
upon  studies  completed during the 1970's. While it is difficult
for a breed to change rapidly, more current  estimates  would  be
desirable.  The  data  from test stations and the National Barrow
Show are current but selected. If all breeders are equally skill-
ful  in  selecting pigs to put on test and those who test at sta-
tions are a random sample of breeders in a breed, the data  might
be  quite  reliable.  However, these assumptions are not met, and
the data are not very useful.

     The breed evaluation data presented  here  are  intended  to
give  guidance  in  making  decisions. Other factors deserve con-
sideration as well.

     o    Source of stock: An increasing proportion of commercial
          pork  producers  buy  breeding stock from breeding com-
          panies. The companies have private breeds  and  private
          lines  from  the  breeds.  The  performance evaluations
          listed here might not apply.

               Within a  breed,  differences  in  breeding  value
          exist.  While  these  differences in breeding value are
          additions to the breed average, and usually not  large,
          in regional situations they can be important.

     o    Changes in breed average: Rapid changes in  a  national
          breed  average  are  unlikely.  Basically, such changes
          come about in two ways. The first is through selection.
          For  this  to be a factor, selection must be based upon
          the same performance criteria and followed uniformly by
          all  breeders. The second is gene migration. An example
          of this is the  importation  of  European  Large  White
          breeding  stock  into the Yorkshire breed. If the Large
          White and Yorkshire  are  different  genetically,  this
          migration  could  substantially  change  the  Yorkshire
          breed average. A similar situation can be found in  the
          Landrace  breed.  Importations  during  the last decade
          from several  European  and  Canadian  Landrace  breeds
          could have changed the Landrace breed average.

     o    Regional  Availability:  Although  a  national   market
          exists  for breeding stock, some breeds are less avail-
          able in some areas. When  health  status  and  breeding
          value  are  considered,  further  restrictions on breed
          availability in  the  area  of  production  are  to  be


     Large differences among crossbreeding systems in  levels  of
performance  and  profitability  are due primarily to differences
among breeds, the beneficial effects of heterosis, and the place-
ment  of breeds in systems. Generally, the use of one of the ter-
minal crossbreeding systems  with  superior  sow  performance  is
recommended.  Success  with any crossbreeding system depends upon
carefully following the system and adopting production  practices
that  enhance  the chosen system. With all crossbreeding systems,
performance can be improved by  purchasing  breeding  stock  with
superior genetic merit.

REV 12/87 (5M)

Figure 1. Three breed rotation crossbreeding system.

Figure 2. Four breed terminal crossbreeding system.

% Figures are available in hard copy

Cooperative Extension Work in  Agriculture  and  Home  Economics,
State  of Indiana, Purdue University and U.S. Department of Agri-
culture Cooperating. H.A. Wadsworth,  Director,  West  Lafayette,
IN. Issued in furtherance of the Acts of May 8 and June 30, 1914.
It is the policy of the Cooperative Extension Service  of  Purdue
University  that  all  persons  shall  have equal opportunity and
             access to our programs and facilities.