Advice Against Line Breeding
It's in the Genes!
by Andrew Merriwether, Ph.D.
Advice Against Linebreeding
Personally it is too risky. You will hear many people who say they did it and had no problems. To me this is like putting one bullet in your revolver and spinning it and putting it to the cria's head. Most of the time it will be fine, but sometimes you end up with a dead cria. The more distantly related the two animals are, the more chambers in your revolver, the less chance of getting the bullet on any one try, but still more chance then if you never load the gun by not breeding relatives in the first place.
It has nothing to do with alpacas and everything to do with the basic rules of genetics. Using people as an example, since we can all relate, every human on earth has an average of six rare lethal recessive mutations. Luckily, except for relatives, most humans each have a different set of six lethal mutations. Since they are recessive, it takes both parents passing it on for the lethal trait to be expressed. Since it is rare, the only others carrying exactly the same trait are blood relatives who share the relative in common where the mutation first happened. Particular mutations typically only happen once ever, so for most mutations, if two individuals share it, they share it because they both got it from a common relative in the past. Alpacas are no different from people in this respect.
That was just listing rare traits. There are many common recessive undesirable traits as well which are more likely to be made homozygous (two copies of the same trait) if relatives breed. So while it is true that you may also make two copies of some good or great rare recessive traits, you are making two copies randomly of traits throughout the whole genome. This is invariably a bad thing. It is why all wild species have mechanisms to avoid inbreeding. Even humans have this. Every human culture has incest taboos to prevent inbreeding. Even the most "primitive" cultures have quickly realized that terrible defects often come from relatives breeding. These "recognition" systems are subtle and very costly for the body to produce. Evidence of millennia of selection favoring outbreeding. This is true for almost every species on the planet.
Linebreeding was designed to be used on large herds (typically hundreds or thousands of animals), taking a small number of herdsires across the whole herd each year. The goal is to homogenize the herd on a similar set of genetic features. It requires VERY diligent culling to remove defects. There are only small number of large farms in the US that really could do linebreeding, and even they would have to be willing to take the losses. Breeding two animals together from your herd that are related is not really linebreeding, it is just inbreeding.
So can you do it?
Is there a risk of increased loss?
Does the risk decrease with the numbers of generations back an ancestor is shared?
Just realize over time there will be more loss in making a linebred herd than if you outbreed. There will be more defective cria born, there will be more lost pregnancies. Linebred animals (of all species) on average live less long, have less offspring over their lifetime, and have weaker immune systems than outbred animals. This is an overall effect if you compared very similar animals outbred vs linebred over a long period of time. It is called Inbreeding Depression. So there is a cost that goes along with benefits of making desireable rare traits homozygous (the prime reason for linebreeding).
Many breeds were created by linebreeding, and the entire breed descends from a single animal, the first animal to have the mutation that defines the breed’s main attributes. Newfoundland dogs are an example of this. Many breeds of horses and cattle trace back to a single male. For most species that are linebred, they can have many offspring per year, so losing 1/4 or 1/8 or 1/16th of the offspring is no great loss. Alpacas reproduce VERY slowly, one offspring max per year (the average production is far less than that, 0.8 offspring per year) and alpacas have a very long generation time (3 years typically), so linebreeding will be VERY slow and you need to start with a lot of animals, otherwise you will cull your whole herd in the first five years, or cull so many you won't have a viable program. If you don't cull, then you are making bad genes homozygyous, meaning these animals will ALWAYS pass on bad traits. In otherwords, you linebreed on all 20,000 genes in the alpaca genome, not just the fleece traits you are looking at. Many of those should not be homozygous.
Anyway, that is my take. You can certainly homogenize a herd on a few traits much better if you linebreed, but there will be a very real cost in lost cria and lost income from those cria. It makes no sense in the US market right now.
For linebreeding, the inbreeding coefficient (for animals more than one step apart in the pedigree) is 0.5 to the n-1 power, where n is the number of steps tracing through the pedigree to connect the two related animals to each other. If they share multiple relatives, you add the separate coefficients together. This tells you what percentage of your genome will be made homozygous (beyond what is already homozygous). Father and daughter share 50% of their genes and DNA sequence, on top of what they already share by chance. Breeding them together will make (on average) 50% of the cria's genome be homozygous more than it would have been if the parents were unrelated. It is random which of the 10-20,000 genes or 3 billion bases are rendered homozygous. So you are linebreeding the ENTIRE genome at the level of the inbreeding coefficient. The more distant the relatives, the lower the coefficient, the less homozygosity (for good or bad traits).
Because you are also linebreeding on bad traits, you will make more problems. If you don't start with enough animals, you will not realistically be able to achieve many goals line breeding. You have to cull, and you have to have enough offspring to be able to cull. This means either a species that has lots of offspring per year and short generation times, or lots of animals to start with. In my opinion, any other strategy is just inbreeding and is unlikely to be very successful unless you are trying to create a new breed based on a novel phenotype (like Newfoundlands or a number of horse breeds).
That is my take on it. How much loss are you willing to tolerate, and for what specific gain? You need a VERY specific breeding plan in my opinion to line-breed successfully. it is far more than just breeding relatives of a certain animal or animals together.
I can tell you that a fair number of animals in my disease database are the result of breeding relatives. Nothing statistically significant, but given how few people line breed in the US, it is intriguing that I keep getting tissue samples from dead cria of related parents. Most of my samples from unrelated parents, don't get me wrong, but an unusual number share a relative in common. I think a good number were unintentional.
So if you want to do it, the risk decreases with the number of steps needed to connect the two animals in the pedigree, with the risk dropping by a factor of 2 to 4 fold with each generation back. Just remember, 20,000 genes, so even a risk of 1:20,000 means one random gene is made homozygous, a risk of 1:100 means 200 additional genes are made homozygous.
Similarly, the odds of getting two copies of the trait you want to make homozygous go down with each additional step in the pedigree by the same factor of 2 to 4 fold with each generation needed to connect the parents. A Catch-22. The closer the relatives the greater the odds of making the gene you want homozygous, but also the greater the chance of making deleterious genes homozygous. The farther away, the safer, but with a much reduced chance of achieving the goal of the linebreeding.
I make these points because people are giving what I consider terrible advice to new breeders about line-breeding. If you have a small number of animals, it is likely you are counting on income from the cria they will make, and it is unlikely most people would want to have increased numbers of deaths, abortions, and defective cria. So I think in today’s market, where female cria sell for a lot of money, why risk losing offspring when you can greatly reduce loss by outbreeding. Try and learn the inheritance patterns of the traits you wish to breed for, and choose your breedings based upon observed phenotypes and inferred genotypes. You can greatly increase the odds of getting what you want out of your breeding program.
D. Andrew Merriwether, Ph.D. is Associate Professor of Anthropology and Biology
Binghamton University. He and his wife, Ann, own Nyala Farm Alpacas, 104 Rockwell Rd, Vestal, New York 13850 USA
Post your comments on this article on Dr. Merriwether's Blog on the Blog section under Main St.
A listing of Dr. Merriwether's seminars can be seen at: www.alpacastreet.com/merriwether.asp