Internet Bernese Mountain Dog Consortium's PandaSite @ You need to load this image!
In-breeding, Line-breeding, and Out-breeding
Adapted from post on Berner-L
Published with explicit permission of author

Each person has their own specifics when defining inbreeding, line breeding, and out-crossing. They also have their own reasons for the choices they make in their breeding program... I can give you my perspective and definitions but they're not terribly scientific and I'm sure others will differ.


I consider in-breeding to be mating first degree relatives, i.e.: father x daugher, brother x sister.

There are two primary reasons for in-breeding, one is to "set" desireable traits. In this case, you are trying to "concentrate" or "homogenize" the genes for those traits by producing offspring from parents who already carry similar genetic packages. The hope here is not only that the offspring produced will be wonderful examples of the breed, but also that they will be prepotent for these desireable traits and have a strong tendancy to pass them on to the next generation.

The second reason for in-breeding a litter is for a breeder to "flush out" undesireable genetic traits. By working with the limited genetic input of 1st degree relatives, you are more likely to double up on any lurking hereditary health problems, thereby producing an affected puppy. Should this occur, it provides the breeder (and hopefully the BMD community) with very valuable information on what undesireable genetic baggage the dogs carry.

The value of information... when shared... can't be overstated. It is the only tool that enables breeders to effectively work towards sounder, healthier, longer lived, Bernese.

Is in-breeding right or wrong?

In my opinion, it has the potential to be both. It can be a valuable tool when used on a VERY, VERY, limited basis... by a breeder with VERY, VERY extensive knowledge of the family of dogs they're working with. It's not something to be done routinely nor in a vacuum.

Here's a hypothetical scenario:

Mr. Studly is owned by his breeder who's been developing her line over the course of 20 years. She's been an active breeder with several brood bitches and stud dogs producing at any given time. Mr Studly has been a big winning dog since he was a youngster, temperament and soundness are good, he's a wonderful example of the breed.

He's now 4 years old and has been on lots of girl's dance cards, you might even consider him a matador. Many of his get are now matured to the point where they've been evaluated for hips, elbows, conformation, etc. He's produced well in all outward respects (conformation, soundness, and temperament) so many of his offspring are now in other breeding programs.

By now, the breeder has some idea of the genetic make-up of Mr.Studly and his dam... they appear to be unusually good examples of the breed in many ways. But some things require just the right mix of genes or are rare enough in the poulation that they aren't necessarily going to show up in routine breeding. There are many, many dogs entering the breeding population who were sired by this dog... is there a timebomb lurking in the genetic package he's passed on along with his good qualities?

One way to find out what lurks below the surface is to produce an in-bred litter or two and follow them closely. Say Mr. Studly is bred to his dam (who has a strong track record of free-whelping litters that grow up to be nice, sound, Bernese). One of the pups from the mother x son breeding develops PRA at 1 year. You would now know that Mr Studly is a carrier for PRA AND that ALL of his offspring have at least a 50% probability of being carriers themselves.

For some breeders, this would be sufficient reason to not breed those dogs. For others, it means that they would seek out mates with an exceptionally low probability of carrying the gene for PRA. Either way, the information gained AND shared from that in-breeding provides breeders with the means to reduce the likelihood of producing PRA affected Bernese.

So what's "wrong" with in-breeding?

There's a large body of research to support the premise that over time, in-breeding drastically reduces fertility and vigor in a population. Remember, along with concentrating those "good" qualities, you're also concentrating the genes for various hereditary weaknesses... say a so-so immune or endocrine system... and, inbreeding may affect the genetic package in a way we don't as yet fully understand.


Basically, line-breeding is mating individuals who share one or more common ancestors. It's in-breeding, but in my definition, the individuals are not as closely related. There are infinite "degrees" of line breeding. Typically, people use the term "close line-breeding" to mean matches like uncle x niece, half-sib x half-sib, grandfather x granddaughter. Loose line-breeding might be when the same dog appears as the great grandfather of the dam and the great-great grandfather of the sire. A dog is "heavily line-bred on Mr. Studly" when Mr Studly adds up to a sizable percentage of the genetic makeup, even though he's several generations back in the pedigree.

In the classic application of line-breeding, generations of dogs are produced using one ancestor as a "pivot point'. That is... each generation is planned so as to maximize the genetic input of one individual from the past. The goal is to maximize and build on the strengths of that indiviual while combining dogs in such a way as to eliminate weaknesses generation by generation. The more loose the line-breeding, the slower and less sure the progress but using this tool, breeders will plan generations ahead. Again, the specifics are infinite... a breeder may choose to linebreed on a specific mated pair in the pedigree or on 2 individuals that were never bred together.

You do have to be careful about the genetics... a relative that's not even shown on a 3 generation pedigree can actually comprise a very large part of the gene package if the ancestors were line bred on that dog. Some computor pedigree programs offer a great tool to calculate the inbreeding coefficient of a dog or prospective litter.

Is line-breeding good or bad?

I think it is a mixed blessing. Line-breeding is probably the most common formula used by hobby breeders. It allows them to learn a fair amount of the strengths and weaknesses of the dogs they're working with. It allows them a pretty good idea of what physical and temperament traits will be presented in each litter. Over time, it allows them to make progress in reducing hereditary diseases and producing typey dogs.

On the other hand, since you're still working within a very small gene pool, you risk the loss of hybrid vigor. While it may work well for an individual breeder and line, or for a limited period of time, there are valid questions about the effects of long term, extensive, line-breeding on the breed as a whole. Bernese have a long history of fertility problems and sub-par immune systems... could it be from countless generations of line-breeding on a handful of dogs? I don't know the answer... but it's a legitimate question to ask.

For another perspective on line-breeding, consider the history of European monarchies and the ill effects the houses suffered from marrying within families throughout the generations. This is similar to the concerns raised by line-breeding for many people.

Out-crossing is the mating of individuals who are, for all intents and purposes, unrelated. No common ancestors in a 3 generation pedigree or 1 common ancestor in 4 generations is an outcross in my book, but again, you'd have to check the inbreeding coefficient utilizing a pedigree program and comprehensive database to be absolutely sure.

When out-crossing, a breeder will try to select a stud that offers strength in the area(s) in which the bitch, and/or her line, are weak, that reinforces the areas in which she is strong, AND that avoids doubling up on the hereditary diseases likely to be a part of her genetic package.

Is out-crossing good or bad?

Some of each... (surprised? )

On the up side, out-crossing will tend to produce fewer offspring affected by hereditary diseases, especially if there is extensive family information available and shared.

This "dilution" of hereditary disease genes is a double edged sword though... while you may produce fewer affected dogs, you are also spreading the disease genes throughout a broader segment of the population, which may in turn increase the liklihood of producing affected individuals down the road. This is somewhat theoretical and may potentially be offset by the use of selection criteria and effective evaluation tools.

Litters are likely to be less predictable and consistent than those in a tightly controlled line-breeding program. Instead of getting, say, 30% show potential pups, you may only get 10% that you really want to go forward with. This isn't necessarily the case (I recently saw an out-cross litter that was very consistent from pup to pup, (3 out of 7 of which had valid, show/breeding potential) but I think it's probably a fair generalization.

When you do get the pup you're after for the next generation, it will likely be more heterozygous (have more diverse gene alleles) for the traits you want than an inbred or linbred pup and therefore may not reproduce those traits as consistently in its offspring.

So, out-crossing is not for those who want to get someplace fast... but when done carefully and with depth of information, it does offer the potential for producing generations of healthy pups and perhaps... improving the overall vigor of our breed.

Caveat: Out-crossing is NOT the same thing as random breeding. Many backyard breeders dis' the show & hobby breeders for "inbreeding" (which, as I've defined it, is actually very rare in this breed) and line-breeding (which they lump in with inbreeding). They will loudly proclaim that their dogs aren't related and that's supposed to be a wonderful thing. Problem is... mating two unrelated dogs without extensive knowledge of their genetic package is likely to produce far less desireable results than in- or line-breeding dogs in a well informed and well planned program.

All in all, each has their place and the breed probably benefits most from having the use of all three... in appropriate measure. I wouldn't want to see all litters in-bred or tightly line-bred... but I'm not sure I'd like to see all Bernese litters out-crosses either, especially when working with the very limited information we have available today.

Clear as mud?

Note that my definitions are fairly typical of dog breeders, NOT of scientists. People who study population genetics, etc. don't differentiate between in-breeding and line-breeding, inbred is inbred regardless of whether it's father x daughter or 1st cousins. Also, scientists tend to prioritize the macro view and consider the impact of inbreeding depression over time. Their understanding comes from the study of animals raised commercially. In that context, it's relatively easy to collect concrete data on things like milk production, egg production, and reproductive performance.

Dog breeders tend towards a more micro view. The improvements they can see in their line and in the breed in a few generations outweigh the less tangible, negative, impact of inbreeding depression. We evaluate only a small percentage of our dogs and base that evaluation on many criteria such as type, movement, hips, elbows, etc. With only a small fraction of the population evaluated and the criteria being either not quantifiable (i.e.: type) or not fully documented (i.e.: hips & elbows), dog breeding doesn't readily lend itself to the kind of analysis that's been done on cows and chickens.