Ram breeding needs incentives

It may be paradoxical, but Dorian Garrick believes that the success of NZ’s breeding value maternal worth has hampered farmer-led innovations.

My research activities expose me to all kinds of technological developments. Many have real potential to provide cost-effective improvements in the rates of genetic gain. It has been particularly rewarding to have played a role in the adoption of some of these science-led technologies.

Starting in the early 1980s, we were able to estimate genetic gains for individual breeders. In the late 1980s, once some leading breeders agreed to exchange rams across flocks, we were able to directly compare merits of animals from one flock to another. By the 1990s, the computers were powerful enough to tackle the dairy industry data using the same approaches, and the across-breed breeding worth was born. Later developments in computing and statistics meant that genomic data could be analysed.

I am proud to say that I have had a hand in many of these developments. Some of which have been exported to many other places in the world.

The farmer-led innovations of yesteryear

Over the same 40 year period, there have been many impressive farmer-led innovations. Whereas traditional ram breeding was limited to selecting only among those animals that were registered and recorded in stud books, known as a closed-nucleus breeding scheme, Tony Parker and the Romney Development Group pioneered the use of open-nucleus breeding programs which screened high-performing commercial ewes and used them to produce sons. Successive generations of those unregistered sons greatly improved the fertility of dual-purpose Romneys.

Russell Emmerson in the Lindis Pass installed an on-farm mainframe computer to his specifications to store a database to his specifications and innovated optical reading of eartags, conveyor belt systems for animal handling, and automated drafting to improve the value of the wool crop from his fine-wooled Merinos, long before these technologies were commercially available.

Some Eastern North Island ram breeders prototyped the use of ram exchanges over a number of years to link their flocks in a sire reference scheme. Some of that same group went on to develop a progeny testing scheme to determine the terminal worth of their Romney rams, including novel traits such as meat yield.

Mac Hanna noticed he had families that produced triplets and infertile animals, which led to the discovery of the Inverdale gene for improving sheep reproduction.

Gordon Levet focused on the environmental challenges of Haemonchus internal parasites and facial eczema, two serious management and welfare problems that at the time were more apparent in his Northland environment than elsewhere in the country.

Bay and Hamish De Lautour challenged their flocks by ceasing anthelmintic drenching of lambs at weaning and then selecting for performance in that challenging post-weaning environment. Many of these innovations led to the widespread adoption of these technologies by new entrants to ram breeding who are likely unaware of this history.

Has the success of the maternal worth index killed innovation?

Sadly, there are now fewer recent examples of these kinds of farmer-led innovation and entrepreneurial activities that benefit the industry and nation. Paradoxically, I believe one of the disincentives to innovation is the success of the NZ maternal worth – a single index that is supposedly comparable across all rams of all breeds.

In principle, I am in favour of a national breeding objective that attempts to combine the relative superiority and/or inferiority of several different economically important traits into a single value. Such an approach is widely used with considerable success in all dairying nations.

However, I have several concerns about its application for a nationwide ranking of sheep. It assumes that sire rankings are consistent across every environment. It also assumes that the evaluation system provides sound comparisons of every sire. It is derived from an economic outlook based on some measure of average national values rather than the farmers whose businesses are on the line. It leads to mating for connectedness in elite ram breeders flocks that can reduce the efficiency of their own selection. It reduces efficiencies by disrupting the long-term service relationships between each ram breeder and their ram buyers. It discourages elite ram breeders from trying something new.

When the relative performance of offspring of sires is different from one environment to another, we refer to this phenomenon as genotype-environment interaction. This phenomenon is very common, most markedly when animals are compared across environments that differ in some form of challenge.

There are three kinds of environmental challenges that cause rerankings: nutritional, due to variation in the quantity and/or quality of feed in different environments; climatic, due to extreme heat, extreme cold, and variation in rainfall; and disease challenges including facial eczema, viral pneumonia, and internal parasites.

One of the reasons that there are so many sheep breeds in the world is that there were environments that differed with respect to these challenges and many breeds developed by local adaptation. Research undertaken by MAF in NZ compared Romneys, Perendales and Coopworths in different NZ environments and demonstrated that these breeds ranked differently in the different environments.

Connectedness is key to a single index

Dairy evaluations use a single index to compare bulls, but those bulls produce large numbers of progeny that are distributed across all the farming environments throughout NZ. In progeny testing programs, good breeding scheme management ensured that any young dairy bull had at most a few progeny in any one herd, but with progeny in many herds. Comparisons between sires were therefore comparisons across many environments so that sires were evaluated for average progeny performance across many different herds. The sheep breeding scheme is very different. Most rams are not used across flocks or environments.

Who really wins out of a connected flock?

Maternal worth combines trait evaluations for reproduction, survival, growth, mature size, and wool. Even if these trait evaluations for every sire were reliable across every environment, the economic importance of these traits is not equal on every farm. Selection is a tool for improving future worth, and farmers need to make their own decisions as to future economic circumstances. After all, it is their own business that will benefit or be compromised in future years by strategic decisions made today. A national index should be a guideline for individual farmers, not the yardstick for comparison.

In order to reliably compare rams in one flock to rams in another flock, genetic linkages are required. The most reliable comparison occurs when a portfolio of sires of interest have offspring competing with each other in the same flocks, and when there is no reranking of sires between those flock environments.

Generating connections with outside flocks is great for new entrants to the ram breeding industry, as it enables them to benchmark their flock against sires from several other flocks. However, generating connections uses up female resources that could have been used for generating additional candidates for within-flock selection. Elite ram breeders who have concentrated on certain traits, such as facial eczema or worm resistance, and who’ve done a better job of selection than others in the industry find that reference sires tend to produce below-average offspring in their flock.

It is an expensive business for breeders to retain males for sale in anticipation of future sale opportunities. However, if ram buyers jump between breeders from season to season based on national rankings of a single index, this creates inefficiencies in the ram supply chain. Those inefficiencies increase ram purchase costs, and reduce ram breeder profits, reducing innovations.

We want diversity in our ram breeders. We want ram breeders to be innovative in a variety of ways, selecting for new traits they believe will be important in future. If the choice of ram breeder is largely made because of a single index such as maternal worth, then a breeder is forced to continue emphasising that index in their selection in order to retain their national ranking. Any selection emphasis on some other trait will compromise the ranking. This may explain one of the reasons why breeders have failed to be proactive in producing a high-performing, dual-purpose, fine-wooled sheep, or a high-performing shedding sheep.

If I were a ram buyer, I would think about what attributes I would want in my sheep for the future, and I would align with a ram breeder who farms in a similar environment and shares my vision of the attributes that matter.

If I were a ram breeder, I would need to make the choice between being a commodity ram breeder or trying to be innovative. Either way, more than anything, I would want to be confident that every year’s ram crop was better than the previous one.

A better ram crop means that their descendants on my clients farms are better suited to their needs than the rams they bought the previous year. There are plenty of existing and upcoming technologies that can help the industry achieve that.

What we don’t need is disincentives to that innovation.

• Dorian Garrick is Massey University professor of animal breeding and genetics.