Previous Page Table of Contents Next Page


C.E. Terrill


The breeding programme proposed is very simple and easy to apply. It consists of three essential steps and, if possible, a fourth step to accelerate progress further.

  1. Mate all prospective replacement females to give birth at about one year of age.

  2. Retain only females for further breeding that wean offspring from this first mating. In subsequent years discard any that fail when they fail.

  3. Mate only twin males born from mothers at an age of about one year, or if no twins are born select the heaviest singles from yearling mothers. Mate all males so that the sire is about one year of age when offspring are born.

  4. Add research and development in nucleus flocks on experimental or government farms to follow the above steps more exactly and fully, to add selection for success of artificial insemination with frozen semen, and to use embryo transfer to increase selection differentials and to decrease generation interval on the female side.

Success is assured by the high selection differential obtained by using the early weaning performance as an indicator of lifetime performance and of genetic merit for high production, and by turning over male generations every year. Experience has shown that the first step will give 1 percent each year plus about 5 percent immediate, one-time phenotypic gain. Ercanbrack has been applying the first three steps in selected groups at Dubois, Idaho since abour 1977. He states in his 1981 progress report-“Percentage superiority of lines selected solely for reproduction criteria, over the unselected controls, is generally increasing at an absolute percentage rate annually of from 3 to 5 percent”.

Room 418 Building 005, B.A.R.C.W.,
Beltsville, Maryland 20705, U.S.A.

Heritability can be disregarded as the goal is a rapid increase in economic gain which might not be the same as genetic gain if a highly heritable trait like fleece weight would receive more emphasis even though in many cases it is much less important economically. It would be better to let fleece weight remain constant and make more progress on lamb production. Of course it would be fortunate if heritability of weaning production were high, but it does tend to be low, particularly weaning rate. This is one reason the rate of progress is as low as 3 percent per year. However, 3 percent per year is very important in regard to net returns as was shown earlier. Certainly the attitude that selection for lamb production is too slow to be worthwhile is not justified.

Most of the research leading to the above proposal has been done with sheep and not with goats. However, the principles will apply equally to goats as with sheep. Rate of progress might even be higher in goats as the initial weaning rate is generally higher in meat goats. The plan is designed for within flock selection only as this will be more effective.

Many farmers with sheep and goats in developing countries probably cannot go beyond step 2, as identification of individuals preferably with eartags and some record keeping at birth and at weaning would be essential. However, steps 1 and 2 can be taken without records or identification providing those females that fail to wean offspring can be detected and removed. If mating is once per year, this can easily be done by going through the flock at weaning time and sorting out those females which are not suckling young. Failures can be sorted out earlier if this permits them to be sold at a higher price. Pregnancy tests may be practical especially on lambs or kids. Some females that fail may be detected at lambing or kidding time. If shearing is between birth and weaning, some may be found then. If young are born throughout the year, marking those that have young may be needed. If feasible an ear notching system for young born may be useful if the mother can be marked without disturbing her maternal behaviour.

Farmers would not need to be concerned about step 3 if they obtain rams from a government farm where nucleus ram breeding flocks of their breed were following step 3 and hopefully, step 4.

The proposed breeding programme is the one action that will give maximum increase in biological efficiency, in net returns and in total meat produced per female maintained, probably greater than for all other means of increasing efficiency combined, over the long run. This advantage is even more important in light of its wide applicability in every country, whether developing, developed or with a central planned market economy. It is applicable under the most intensive to the most extensive conditions, and under any environmental situation. It enhances and does not conflict with or duplicate any other means of increasing efficiency. It enhances further genetic progress for efficiency because selection differentials can be increased and generation lengths can be decreased as the weaning rate increases. It enhances selection for other traits, although progress may be reduced if much attention is given to other traits. It enhances improvement in adaptability even though such will be slow. The proposed breeding programme should be given priority, both in production and research, over any other means of increasing efficiency because of its promise of greater immediate and permanent gains, especially in net returns to farmers.


Success in predicting lifetime performances by breeding females giving birth at about one year of age depends in part on the proportion that wean offspring. An average of 50 percent may be optimum, because this will divide the population into those that are above average and those that are below average phenotypically and genetically, of course with some overlap. The numbers that are needed for replacement will be important. The interaction between weaning rate, the number of years a group needs to remain in the flock and average generation interval are shown in Table 1. A small proportion weaning offspring in the first year would probably give a higher selection differential while a higher proportion would permit a shorter generation length. However, where flock weaning rates are generally below 100 percent, the number of females weaning offspring in the first year would need to be 50 percent or better to provide sufficient replacements. This might not be attainable on farms so that selection gains on experiment stations and on government farms who provide sires to producers would be highly important. These values are only a rough guide and actual performance could vary considerably.

If the weaning percentage and/or the number having young at one year is too low to follow the plan completely, the farmer should go as far as he can as even a little effort will be worthwhile. Also, if it is impractical to mate offspring at 7 months of age, it may be done later or at 19 months of age, but the gain will be less. This programme can be effective in increasing net income even if only followed in part. The number reproducing at one year of age may be discouragingly low at first, but it will likely improve every generation and possibly every year.

The proportion of females weaning lambs in the first year could be increased by feeding the lambs from weaning to breeding so that they could reach 60 to 80 percent of their mature weight before breeding. The extra production might pay for the feed. Again this could be done better on government farms than by private farmers. Actual results would vary with breed, location, time of year, quality of feed and possibly other factors so that research is needed in each area to determine the optimum procedure for each breed in that area. This supplemental feeding of female offspring would not be the same as supplemental feeding for production. The objective of feeding these offspring would be to reveal genetic variation and to approach an optimum selection level.

If more female replacements are needed than the number weaning offspring in their first year, the next choice would be females in their first year which had live lambs but did not wean them. Next would be those that became pregnant in their first year but failed to produce live lambs. Keeping high producing ewes in the flock more years might be preferable to retaining females that failed to become pregnant in their first year. At Dubois, Idaho, under certain conditions, some of the highest producing ewes would maintain production through 10 or 11 years of age. If possible, females should always be replaced with younger ones with higher genetic merit as revealed by production at one year of age in a subsequent year.

If more females weaning offspring in their first year are available than are needed for replacements, the oldest ewes with the lowest performance might be culled, thus reducing the generation interval or if indexes are available, the females with the lowest production indexes should be culled regardless of age. If genetic progress is being made, the younger females will have higher average genetic merit than the older ones but judgement needs to be used to retain the ones with highest genetic merit regardless of age.

The recommended selection index is the average lifetime production of a female in total kilogrammes of liveweight weaned annually plus the equivalent weight of wool (actual fleece weight adjusted for the relative value of wool to meat) and with milk production added in the same way if the females are milked. Thus, the females may be ranked at any time on their production index, adjusted to a lifetime equivalent, so that those with the lowest indexes are the first to be culled and sons of those with the best indexes should be used in breeding.

The selection index for females may be easily calculated if individuals are identified, preferably with eartags, and if records of birth dates and weaning weights and dates are taken. Weaning weights may be adjusted to an average age at weaning on the basis of average daily growth from birth. Adjusted lifetime equivalents may be calculated based on the deviations of individuals from age group average, including a correction for number of records, and expressed as a deviation from the lifetime average or by adding or subtracting the deviation to the lifetime average for the entire flock.


Selection of males is much more effective than selection of females as a much smaller proportion of males need to be used and male generations in sheep and goats can be turned every year. The rate of progress from selection of males alone will probably be about 2 percent per year or twice as great or more than with females. This assumes that males selected within the flock are the very best for that flock, which cannot be achieved if males are purchased. Individual identification and keeping of some records is essential for selection of males on the farm. If this is impractical then the gains will need to be made on governmental farms or in nucleus flocks from which farmers would obtain rams.

Males should be selected on their mother's performance, because this permits turning generations every year. Annual gain from using male offspring under one year of age is greater than could be obtained from more accurate half-sib records or progeny tests, because male generation length would have to be two years for half-sib records to be used, or three years for progeny records to be used. Therefore, the selection differential would need to be twice or three times as great to offset the longer generation length, and this seems impossible.

Turning male generations every year is responsible for the major gain over traditional methods of selection towards increasing the rate of improvement of lamb production. Male twins from females giving birth at one year of age would have an environmental disadvantage both from their type of birth and young mothers. Therefore, it is essential that the selected males and some substitutes be given supplemental feed from weaning through breeding. Semen tests should be made prior to breeding to see that they are producing ample numbers of sperm with normal motility. Semen can be collected on farms, and the rapid swirling motion of normal sperm can be seen with the naked eye. If group breeding is practiced with at least 3 sires, this test for semen motility is certainly adequate, and it is probably adequate for individual sires to ensure high fertility.

Selection of rams within flocks may introduce danger from inbreeding, although frequent changing of sires tends to slow inbreeding. One sire inbred lines increase in inbreeding at the rate of about 1 percent or more per year (0.5 to 1.7percent) while lambs weaned of ewes mated tend to decrease about 1 percent or more with each 1 percent increase in inbreeding. Therefore, one and also two sire farm flocks under within farm selection are not recommended even though the selection gain might more than offset the loss from inbreeding. Three sire breeding flocks would probably be satisfactory if 3 sire lines (original unrelated sires) were maintained and if female offspring were rotated among sire lines so that parents would always have the lowest possible relationship within the closed flock. This does put flocks smaller than about 60 females at a disadvantage. Small flocks should probably change sires with other small farms following the same practice or obtain sires part or all of the time from nucleus or government breeding farms, also following the same practice.


Small fanners are often unable to practice selection most effectively even if they do identify individuals and keep essential records. Optimum selection groups should have around 200 head for each breed, and most small farmers have less than this. Small farmers rarely have the facilities for satisfactory record keeping or for single sire matings. Also, it is difficult for any farmer to select for only one trait, to believe environmental effects, to select for something like lamb production that he cannot see when making the selection, and to avoid being influenced by appearances that are striking at the time of selection. Professional Animal Breeders sometimes have the same problems, but they are more likely to be entirely objective in selecting the superior animals. Therefore, an experimental or government nucleus flock, with the help of a professional geneticist or animal breeder can probably make progress with selection at a more rapid rate than farmers. They should not be influenced by economic returns, as a farmer must be, and should be able to take more risk and be more ruthless in culling. Where government breeding farms are established, they provide the most effective means of increasing production efficiency through more rapid genetic improvement.

Nucleus or government breeding flocks should be within the ecosystem area in which the farms are located which they are to serve. They should have nucleus flocks for each breed that is important in the area. Methods of feeding and management should be as comparable as possible to farming practices in the area. Plane of nutrition should be the same as on farms. Adaptability is the main genetic asset which native breeds have. It will be improved constantly but slowly by natural selection providing environmental influences are much the same as on the farm.

Farmers should breed females to give birth at about one year of age whether or not they obtain sires from a government farm. The phenotypic gain is important, but it is not permanent like the genetic gain, and therefore, the selection of females must go on without interruption. The additional genetic gain from female selection is well worth making and should be continued indefinitely the same as that for males.

Government breeding farms are well justified because of the public gains from increased efficiency of production. The supply of high quality food is increased and the price usually goes down as efficiency goes up. Rural income goes up which tends to benefit everyone in each rural community. In many countries, the public may benefit even more than the farmer. It is reasonable that the government provides the breeding farm as the gains might not be made otherwise.


The selection for other traits, along with selection for meat production efficiency, can best be done on government farms where professional animal breeders can supply the highly technical input that is needed. Selection for wool quality, efficiency of feed use and even fleece weight and milk production require measurements that may not be practical on the farm. They require a more sophisticated selection index than is recommended for meat production alone so that net returns will be further increased. They may require some research where unknown parameters might be obtained. Research to increase the rate of progress from selection will always be needed.

In areas where breeding seasons are limited, in some or all breeds, and where breeding out-of-season may be desirable, selection for the ability to breed successfully any time of the year should be done in the nucleus or government flocks. This may be done by mating all eligible females every month or every other month. Synchronization should be used so that young are coming in restricted periods. Further procedures would be the same as for breeding once per year. Offspring born at the more difficult times of the year should be favoured in selection.


Artificial insemination offers the best means of distributing gains from nucleus or government breeding flocks to many small flocks within each ecosystem area. Use of frozen semem will permit distribution to many places over a relatively wide area compared with the limitation of only a few places with fresh semen. Use of synchronization of oestrus and ovulation will allow the process of artificial insemination to be planned and programmed. However, frozen semen gives such low pregnancy rate (average around 40 percent) that improvement in this trait must be made before A.I. can be used practically.

The most plausible way of improving success of these techniques is through genetic selection for success. Success of synchronization of oestrus and ovulation without reduced fertility is better for some females than others. Likewise, a few rams produce semen with high fertility after freezing. These variations undoubtedly have genetic components. Simply using these techniques in nucleus and government breeding flocks will lead to progress as all replacements will come from those that succeeded. Male selection could emphasize both lambs weaned per ewe mated and success with frozen semen. There would likely be no genetic conflict between the two traits. Degree of success could be predicted from freezing and insemination tests along with early pregnancy diagnosis before the prospective sires were seven months of age. Of course research would be essential but as a part of the selection process. Selection should be started immediately and then enabling research could be done to accelerate progress in selection already underway.

The objective should be to increase the success rate with frozen semen sufficiently to permit farmers to do the insemination themselves rather than depend on technicians. Varied results because of technique of insemination would be much less likely with a high than with a low success rate. Farmers would need to do both the synchronization and insemination techniques themselves to ensure that these would be timed to suit their farming operations rather than the schedule of a technician.

About ten years of selection or ten male generations would be required before success from artificial insemination with frozen semen might reach 60 to 70 percent at which time, use of the technique on selected farms might begin. As the reproductive rate is increased by selection, a lower success rate from synchronization and artificial insemination can be accepted because there would be more room for culling of ewes which fail. Pregnancy diagnosis soon after mating would permit failures to be marketed, many before one year of age, before the costs of feeding to heavier weights, or through the pregnancy interval are incurred. Those that succeed in weaning offspring after birth at one year of age from synchronization and insemination with frozen semen, would probably show a high success rate throughout their lives.

Artificial insemination not only gives the small farmer the advantage of superior sires for increasing meat production, but also he can now economically use more than one sire so that his best females can produce potential replacements and other females could be used to produce market offspring. It seems probable that the cost of using synchronization and artificial insemination might be offset by not having to purchase or to withhold a prospective sire from market or to keep a male year round. The small farmer would thus have the advantage in flexibility and of use of superior sires, probably with no added cost as compared to natural mating.


Embryo transfer, which requires surgery in sheep, may not become practical in production, but it offers great promise in enhancing genetic progress. Fertilized ova may be transferred from females superior for reproductive rate to less superior foster mothers. Thus, selection gains may be more rapid, because this will increase the selection differential and will decrease the generation interval on the female side. Use of superovulation is not favoured for this procedure in order to give maximum attention to genetic improvement of the entire reproductive process including the natural ovulation rate. Embryo splitting is favoured, however, to increase the number of embryos per collection. Research along with the selection procedure, might result in turning generations every two years with a corresponding increase in the selection differential to about 70 percent of that for males. Further progress could no doubt be made with more research. With selection one should never conclude that current practice is the best that can be done. The goal of research with selection is to find a way to make more rapid progress.


Selection for total weight of young weaned per female mated can be added to selection for milk production in small ruminants without loss in milk producing ability. Improvement in one is consistent with improvement in the other, because high milk production contributes to heavy weight of offspring weaned. Of course, milk consumed by the young cannot be sold but management can be used to obtain the optimum yield of each product.

Initial mating of females to give birth at one year should be done. If the weaning rate is such that some which produce offspring at one year could be culled, the ones with the lowest milk yields might be culled. However, it would be more efficient to calculate indexes of adjusted liveweight of offspring weaned times a standard value plus adjusted yield of milk times a standard value. The records available would be adjusted to a lifetime equivalent including correction for the number of records. Males would be selected on their mother's index and females could be selected on their own index. If some females failing to wean offspring after birth at one year would need to be selected, this could be done on their mother's records.

Angora kids produce mohair of higher value than older animals. Thus, it is an added incentive for selecting for those that wean more offspring at a young age. As the rate of reproduction increases, the mothers and females kept for mohair production could be culled at a younger age thus increasing the yield of meat. With much higher weaning rates, it might not be necessary to keep males or mothers after their kid mohair was produced thus still further increasing the economic efficiency of meat and mohair production.

Efficiency of production of slaughter animals from fine wool sheep can be increased in the same way as for milk or mohair simply by selecting for most economical combinations of the products and then following the same procedure as for the production of slaughter animals alone. This procedure has been very successful at Dubois, Idaho with the Rambouillet breed.

TABLE 1. Approximate Guide to Optimum Number of Females weaning Offspring in First Year

Offspring weaned per female mated in flockOldest age to remain in Flock by proportion of females weaning offspring in first yearAverage generation interval in years for females by proportion of females weaning offspring in first year
.7     10     6.0
.8    109    5.55.0
.9   1098
1.0   986
1.1  10875
1.2  9765
1.3 108654
1.4 107644
1.5 96544
1.6 86544
1.7 85444


  1. Sex ratio of 0.5
  2. Mortality from weaning to breeding of females at 0.05
  3. Mandatory culling of female offspring of 0.10for defects, small size, and unacceptability
  4. Annual mortality of adults 0.05
  5. Annual mandatory culling of adults 0.05 for spoiled udder, unthriftiness, unsoundness and failure with age.

Previous Page Top of Page Next Page