Dairy Cattle Breeding

Are We Heading in the Right Direction?

Dr. E.B. "Ted" Burnside

Nova Scotia Agricultural College,
Truro, NS B2N 5E3 Canada
Email: Tburnside@CAdmin.NSAC.NS.CA

A Glance at the Crystal Ball

Young dairy sires being entered into progeny test programs in 1997 will set the standards for dairy improvement by the year 2000 as they achieve their first progeny tests. AI centres are maturing and consolidating into breeding companies today. Can we expect them to be as aware of the needs of the dairy farmer as they have been in the past, and will they be able to cater to individual concerns as well as they have in the past? Decisions on young bull entry into progeny test programs are based largely on the sires and cow families behind them, and these will be made globally by a handful of geneticists for these new companies. How can we be certain that they are going in the right direction? One simple way is for dairy farmers to signal with their purchases, but that leaves some companies hanging out on a limb with four years of young bulls in the pipeline who will not serve the public well. Another way is to discuss breeding directions together in a meeting like this and hope that some of the decision-makers are listening.

I am going to take a quick look at the current breeding direction and attempt to examine key traits that may have been overlooked so far. Dr. Wencan Zhang, a bright geneticist with Semex/Alliance Canada and his colleague Dr. Herman DeBoer, have recently reported on a study of global consumer preferences in Holstein semen purchases. Results are, of course, affected by the perception that dairy farmers around the world have of Canadian cattle and Semex/Alliance semen, in particular. When they come to Semex/Alliance for semen, no doubt they are shopping for type. Dr. Zhang did a factor analysis on consumer buying habits for different sires, dependant upon the number of doses taken by each country. A factor analysis attempts to identify those most important proof components that are determinants of demand, and it is commonly used in analysis of marketing data. When the results had been tabulated (Tables 1 & 2) Dr. Zhang found that there was heavy and growing emphasis on protein yield on the part of shoppers for Semex/Alliance semen. He also found that there was considerable emphasis on conformation as well. Udders and feet and legs got heavy emphasis along with protein yield. Farmers from some countries are happy to buy breed average type Canadian sire semen, provided it has high production, perhaps because they are confident that these cattle will be above average in comparison to the progeny of other available sires with respect to conformation.

Table 1. Principle components analysis of foreign buying patterns.
Percentage Variance 1990 / 1991


* Percentage Variance 1995 / 1996 Traits





Type, Fat, Protein




Milk, Udder






Type, Protein



Fat, F&L

R.U., C.E.

93% 83%

Table 2. Principle components analysis of EEC buying patterns.
Percentage Variance 1990 / 1991


* Percentage Variance 1995 / 1996 Traits



Type, Protein, Fat

Size, Cap, Rump





Production, Type

STA, Size, Cap


91% 93%

Canadian dairy farmers are also shopping for more protein, particularly in Western Canada. So, while the average sire exported by Semex/Alliance is +49 kg for protein, Canadian dairy farmers chose bulls averaging +55 kg. For mammary system the average export buyer settled for a +7 sire, while Canadians chose +9 sires, on average. Only German dairy farmers opted for more protein (+59 kg), and no country was more choosey for mammary system. Dr. Zhang's study included a profile of different countries which was also revealing. It certainly indicates a breadth of thinking about breeding goals. Who is right? No one really knows!

Let's look at some of the key traits that we have been giving attention to, as well as some that we have failed to select effectively for in the past. Will there be shifts in emphasis over the next few decades?

Protein Yield

There is no letup in the protein race, although a little evidence suggests that dairy farmers are not so interested in the sire that improves protein at the expense of good udders. A shift towards conformation is evident in some programs which have been focusing chiefly on high protein yield in the last decade. Clearly, the Aerostar sons, which are recently emerging with proofs around the world, combine exceptional protein yield improvement with outstanding well-attached udders and good feet and legs. Their sons should set the standard for exceptional yield and wearing characteristics in the year 2000.

I am doubtful that we will shift away from protein yield, in fact, I expect there will be more attention in the next decade to improvement of protein yield combined with other practical characteristics such as udders and, to some extent, feet and legs.

Fat Yield

It is doubtful that consumer preferences will allow a resurgence in selection for fat yield, however, there is no doubt we will pull fat along with protein because they are genetically correlated. The progeny test is, however, extremely powerful in breaking up genetic correlations and we might expect that selection against fat yield will continue to cause greater divergence between fat and protein yield in the future. Some bulls that are now getting extensive use have quite low fat yield in comparison to their protein yield, and their sons could cause greater divergence between fat and protein yield.

Size and Stature

We have been watching Holstein cows getting larger, heavier, and taller under the influence of the show ring for some 40 to 50 years. If we are to choose a new direction today in selection goals, I believe it will be to reduce size. It is time we stopped selecting for larger and taller cows. As an alternative, we may simply begin by paying absolutely no attention at all to size and stature. If we wish to move more rapidly we may actually discount larger cows in the selection process and this will give an advantage to the medium-sized cow. There are some very practical reasons for going in this direction:

It is time that we standardized the size of cows because, like pieces of machinery, a bigger cow requires more space in a barn and this is costly. Big cows cost a lot to keep. They don't fit into an average size conventional or free stall. We are definitely heading for the commercial era, and big cows will not fit in that era. Why not start breeding for a standard, medium sized cow. If, just by chance, you get a big one that is also pretty, you can move her from your commercial herd to the first show ring jockey that walks into your barn and is ready to pay a premium. Maintenance costs on large cows are substantially higher than those on medium sized cows. One big cow with medium or low production is going to cost you a lot more to keep around than the medium or small sized cow that delivers the same number of kilos of protein and milk in the same time.

Changing cow size is costly, particularly if we have to remodel a barn or milking system every decade. We all know that it is necessary to remodel periodically, but it makes no sense to reduce the number of free stalls in your barn to accommodate bigger cows.

Research on the genetic and phenotypic relationship between size and production, based on about 90,000 DHAS Holstein 2-year-olds by Robert Moore (2) , found a slight negative association between size and production. This means that positive emphasis on size is at the expense of production. In view of the above findings, we are doing a study of stature, size, and their relationship to production, profits, and longevity. We need to establish the optimum size of dairy cow for the future. For now, my advice is ignore stature, or standardize to a medium sized cow.

We have established that production is of greatest importance and that we should back off on size and stature. The most efficient cow is very likely to be a medium sized cow with extraordinary production. My choice among two cows with equally exceptional production and type will always be the smaller cow because she'll cost less to maintain and therefore will be more efficient.

I believe we should forget about size and stature for the next two decades in our selection programs because our cows are plenty big enough! Throw away the measuring sticks and the minimum height or weight requirements as qualifications for Excellent cows in the conformation systems. Make a note to look at cow size in 2020, but I expect we will go right on ignoring size then. Breed associations that keep on increasing cow size by setting standards for their top conformation ratings (Excellent) for size, stature or heart girth will automatically cause selection in that direction and will undoubtedly lose membership in the future.

Feed Efficiency

Research has demonstrated that plenty of genetic variation exists for efficiency of feed utilization. Pig and poultry breeding companies have been testing and selecting for feed efficiency for many generations, with very significant savings in efficiency and in feed costs for the commercial sector. Genus Genetics in the UK made a start in measuring feed intake on their dairy heifers in a nucleus breeding program in the mid 80's. The idea did not turn heads at the time. With the very recent amalgamation of Holland Genetics and the Genus Nucleus, we can expect to see feed efficiency as a major selling point in the very near future from Holland Genetics' sires. The problem is that it is costly to measure feed intake. Breeding companies which have been making significant income from young bull sales would all be well advised to go to a fence line feeder with fenders at each stall to prevent stealing of feed from the next stall, equipping these feeders with Broadbent gates and a unique numbering system so that one cow can get into one stall only. If you feed total mixed rations, monitor the feed that goes into the cow's stall, and purchase a little vacuum self-propelled unit to measure the amount of feed that the cow refuses, you can test up to, say, 40 cows in one paddock with this system. You can then record feed intake of your most productive and highest indexing cows and heifers and be ready with facts when the sire analysts come calling.

This kind of selection won't make the rapid progress we have been making by broad brush progeny testing. We are only selecting between top cow families and young females. We really need engineers and dairy farmers to work together to devise some sort of electronic system that will record feed intake on individual cows in the field. It must be low cost, simple, and accurate. This approach would allow dairy cows to be monitored on pasture or when feeding on total mixed rations.

One of the ideas I have been selling for about five years is the formation of breeding companies as a means for enterprising commercial dairy farmers to get into cattle breeding near the top. I believe the breeding enterprise of the future will be built around commercial operations which can put up 1000 or more commercial cows as recipients, and build a program around several top cows. They must have very high numbers, outstanding udders, and be able to move freely in a free stall system. Other traits of clear importance in free stalls will be medium size, feed efficiency, milking speed, reproductive performance, disease resistance, and longevity. The fancy traits go out of the breeding plan when we come to realize that the ideal commercial dairy cow is the most profitable cow. If a trait is not paying off in commercial conditions, it won't be in the selection plan.

Reproductive Performance, Longevity, and Disease


One of the most difficult series of traits to improve are those associated with resistance to disease, including reproduction, and this total package affects the cow's lifetime in the herd. Yet these traits are half as important economically, or more so, than production. Longevity alone was found to be about 60% as important economically as production in the dairy situation. We all know the economic importance of reproduction and disease resistance added to that mix makes a very important package of traits. All of these traits have very low heritability and are very hard to improve by selection using conventional progeny testing or other similar approaches. I doubt if we have effectively recorded these traits in field conditions and I am wondering if we are going to get this job done. Certainly, mastitis resistance and selection for it via monitoring somatic cell count may be the first real breakthrough in this area. However, there are scientists who debate the accuracy, or wisdom in fact, of selecting on somatic cell counts. Where do we go from here?

There can be no doubt that these traits may be improved by selection. A recently published experiment in Ottawa leads me to be convinced that there are genes which affect reproductive performance in the mammal, as well as the length of reproductive lifetime. This study involved selection and control lines of mice which were monitored for the number of litters produced in lifetime matings of paired mice. Pairs were selected which continued to reproduce steadily for the longest period of time. This experiment clearly demonstrated that the lines can diverge rapidly and steadily. Selected lines now continue to produce young for nearly twice as long as the controls.

Can we achieve similar success with the dairy cow? If we carefully record and standardize the environment by statistical methods, we can certainly achieve some progress in selection for resistance to disease, reproductive performance, and longevity. This would certainly contribute to greater profitability in the dairy cow. However, in the interim, scientists at our laboratories are looking for genes that may affect the longevity of the aforementioned selected mice, and if such genes are found, we will be developing tests to detect similar genes, sequence them, and then provide genetic monitoring for these genes for dairy cattle and other species. This kind of research is costly and time consuming. We anticipate a five year period before findings will be available on our particular experimentation. It is, however, very exciting and could unlock genetic variation that we have been unable to anticipate.


  1. DeBoer, H. and W Zhang. 1997. Global Buying Patterns for Canadian Holstein Genetics. National Holstein Breed Strategy Conference, Quebec.
  2. Moore , R. K. 1991. Parameter Estimates Between Production and Management Traits in First Lactation Using Milk Recording Data. Ph.D. Thesis . University of Guelph.