Healthy Choice - How Breeding for Fitness and Well-Being Would Pay Off in the Future

When talking about using herd management data, a veterinarian once asked why people dwell on reasons for disposing of dairy cows and why they die. Instead, he suggested, why not concentrate on what happens to make them sick while they're still alive? Then you'd have a lot better chance of fixing the problems. His comments had a lot of merit. In dairy genetics, we've spent far more time analysing why cows die than we have on how we might keep them alive and well.

A recent paper, summarizing work done by University of Wisconsin researchers Nate Zwald and Kent Weigel, marks a tremendous leap forward in how dairy breeders look at genetic improvement programs. Reporting on their work at the annual meeting of the American Dairy Science Association recently in St. Louis, Missouri, the two researchers looked at the possibilities of improving health traits through genetic selection. Their work also raises the question of how we should consider selection for many traits that affect the fitness and well-being of dairy cattle.

Zwald and Weigel analysed data collected from about 160,000 cows in 579 herds. These herds used one of three management programs: DairyCOMP 305, DHIPlus or PCDart. All three on-farm software systems let producers routinely record health and treatment events.

Traits the researchers analysed were clinical mastitis, lameness, metritis, ketosis and displaced abomasum. The numbers varied per trait since it appeared that not all producers recorded all of the traits. Incidence levels, or the number of times a condition occurred in a herd, varied: clinical mastitis, 14.1 per cent; lameness, 10.4 per cent; ketosis, 5.4 per cent; metritis, 4.9 per cent; displaced abomasum, 2.2 per cent.

Heritabilities, or the amount of variation in a trait due to genetics, were low to medium, ranging from 0.04 to 0.14. Low heritability of under 0.10 usually means that any progress made through selection will be very slow. Low heritability can also be due to not having records as accurate as we would like or not describing the trait accurately.

For example, even though the incidence level of displaced abomasum was the lowest, this trait had the highest percentage of useable records as well as the highest heritability. This condition possibly has the highest genetic component because its diagnosis is objective and leads to a more standardized recording of the trait across herds.

Most of the other traits dealing with health are subject to interpretation. Reporting levels can vary, depending on whether the people recording the data log a clinical case, or just that the cow appears to have that particular problem. This tends to make it more difficult to have reliable data and contributes to low heritabilities.

Zwald and Weigel calculated genetic correlations among the traits suggesting that some cows may be prone to having more than one disease condition. As well, they estimated sire values by predicting the probability of daughters having disease conditions. Lactation incidence rates for daughters of the worst sires were two to five times higher than for the best sires.

This kind of data can be developed into a national disease recording system and genetic evaluation system for health traits. Scandinavian countries, which have had such systems for the past 20 years, have shown how it can be done. Development of the Norwegian Red breed, for example, has been based over the years on a selection index with heavy emphasis on health and fertility traits.

Although health traits have a complex nature, are influenced by many genes and have low heritability, the Scandinavian experience shows a dairy industry can make progress through long-term genetic selection.

Selection programs there claim to have reduced traits such as cystic ovarian disease and stillbirth incidence.

The concept of a complex selection index including fitness or health traits isn't really foreign to Canadian breeders. For instance, the Lifetime Profit Index, commonly used to rank artificial insemination sires in Canada, includes weights on functional type traits as well as production traits. The index also includes scores for somatic cell counts and calving ease. Canadian emphasis on breeding for a balance of traits may have prevented such a decline in fertility and health traits as seen in other countries more focused on production alone.

Including other disease resistance or health trait measures in Canadian selection programs wouldn't be a big move and would payoff in dairy barns across the country. Genetic variation for disease incidence does exist, is economically important and justifies including disease in breeding programs, in spite of low heritabilities for disease traits.

Genetic studies generally tend to find low heritabilities for disease incidence traits such as retained placenta and cystic ovaries. However, other studies have found medium heritabilities for dystocia, metritis, milk fever and mastitis. Maybe the reason for low heritabilities is that we haven't defined or identified these traits accurately.

Studies have found a negative relation between milk production and diseases such as clinical mastitis and some fertility measures. Emphasis on production traits and increased value of production has resulted in genetic productivity gains of more than 1.5 per cent per year. Much less emphasis has been placed upon genetic selection to reduce production costs. Some research suggests that genetically, health and fitness traits may be changing at an equal and opposite rate to production.

Some disease conditions occur in early lactation, also a time of maximum loss of body condition. There is research to suggest that the interrelation of these disease conditions, along with body condition score and high production may indicate a cow's ability, or inability, to cope with the demands of high production.

To deal with such issues, we need high-quality, reliable data to generate meaningful genetic measures and parameters on which we can base selection programs. U.S. herds providing data for the Wisconsin study participated in either progeny testing programs or were known to routinely record health disorders. The researchers applied various edit checks to the data to eliminate obvious errors.

In Ontario, many DHI {dairy herd improvement) members enter detailed and complete health and disease information through the Dairy COMP program as an integral part of their management program. Over 25,000 pieces of health information streamed through the computer system of CanWest DHI last year alone. The present system doesn't summarize this information or give back benchmarks or standards as it does with production information.

Another shortcoming is that health data recorded on farms can be messy. It's sometimes recorded inconsistently, and there's no standardized definition of what's to be reported. Reporting levels depend upon the person doing the recording.

Reinforcing this view, in Ontario at least, was a preliminary analysis conducted of health data entered during 2001 and 2002 from DHI herds that used an on-farm copy of the DairyCOMP program. Of these 150 herds, some entered no health data and some appeared to be interested only in certain traits. Incidence of various health traits ranged from three per cent to 10 per cent per year, depending upon the trait. Similar to the Wisconsin study, this analysis found clinical mastitis was most frequently reported, followed by reproduction-related disorders.

Since this was a historical look at the data, we don't know which pieces of information were reported accurately. There's probably no way to go back and truly find out. That's why developing a strategy to collect and use this information is so important to the dairy industry at this time.

To address this issue, Dairygen and the Agricultural Adaptation Council have funded a study to come up with a strategy for recording health and disease incidence traits for genetic improvement. Can we in Canada have a voluntary system where producers enter disease information to use for genetic evaluation as milk records and type information are used now? Or, can this information only be obtained through contracts with certain herds that provide the information as a service?

Several factors make it important that we come up with a strategy so we can look at improving dairy cow health through genetic selection:

  • concern that constant selection for productive traits has had or is having a negative effect on health and fitness;
  • ongoing pressure to reduce antibiotic use for treating farm animals;
  • improved animal welfare through reduced health problems and conditions;
  • lower costs by reducing treatments, veterinary bills and lost production.

New quality assurance and food safety programs could provide one opportunity for recording the needed data. These programs already do or will require such items as identifying individual animals for National Livestock Identification, recording all treatments of animals with livestock medicines, and keeping disease surveillance information. Why not design a system that also uses this information to help you manage your herd and breed for more trouble-free cows?

U.S. and Scandinavian studies have shown that there is genetic variation in dairy cattle health and disease traits. If we have let performance slip by not selecting for these traits, then it's logical that we can improve them if we do select for them.

The challenge is to come up with a standardized, workable system to record and measure some of these traits. Then we can make genetic progress in disease resistance and health. Progress may be slow, but it's possible. In the long term, it will pay off for the entire industry and you as an individual producer.

This article first appeared in the Milk Producer magazine.

For more information:
Toll Free: 1-877-424-1300
Author: Blair Murray - Dairy Genetics Improvement Specialist/OMAFRA
Creation Date: 1 September 2004
Last Reviewed: 3 June 2010