Purebred dogs is a dear subject for many. But what does it mean to be purebred? Large numbers of dog breeds were influenced by other populations at one point or another, even if enthusiasts of those breeds don’t always remember this. For example the crosses between Belgian Tervueren and longhaired Dutch Shepherd made in the 80’s: when I was first introduced to the breed early 2000’s, the R-registered/”not pure” generations still existed (ie. my first Dutch Shepherd) and I quickly learned about many opinionated breeders around the world actively avoiding these impure dogs and their descendants with atypical traits. Now that another 20 years is gone I don’t really see this anymore, and it is even starting to feel like the Tervueren crosses are now, almost 40 years later, finally a universally accepted part of this variety. Now “pure longhairs” already include the dogs with Tervueren from the 80’s in their pedigrees: before they were called F-line and only dogs with shorthair in their background were able to be pure. Not many know it, but the shorthairs had already introduced another Tervueren cross from the 50’s through their pedigrees! The fact that this was forgotten just enhances the idea that 30-40 years in people’s minds will allow any foreign blood to be integrated into the purebred population. Of course, otherwise we could never call any Dutch Shepherd a purebred!
History always repeats itself, and today we have similar, but different prejudices. Nowadays we have “pure longhairs” and “longhairs with shorthair background”. There are two types of shorthair background: dogs descended from non-working line / pure shorthairs who can pretty much be considered pure longhairs as soon as they get a long coat, and longhairs descended from working-line / impure shorthairs who will be able to be called purebred in 40 years time. 😉 All longhair dogs born out of shorthair parents are considered a part of the latter group as well, starting from 1999 Pacha/Polka offspring (even though the offspring of these two is already much better accepted than some of the newer dogs). Before going any further in this post I want to note that when I talk about pure or impure Dutch Shepherds, I’m not talking about how the dog is registered, but how clean or unclean the bloodlines are seen as in the breed circles (due to false pedigrees, lookalikes etc.).
How big of a percentage of each contributing breed a dog has is based on knowledge of foundation dogs. My website has a history section (in Finnish, unfortunately) introducing the idea that longhaired Dutch Shepherds have six foundation dogs (a foundation dog always has unknown parents – they are the first dogs to be called a specific breed), so if I propose that a dog has x% longhair, it means x% inherited directly from these first dogs. A purebred longhair also has some shorthair (regular crosses starting from the 60’s), roughhair (through shorthairs), German Shepherd (also through shorthairs) and Belgian Shepherd mixed in through both long and shorthairs. The Belgian Shepherds crossed into Dutch Shepherds were always Tervueren (they did do a Malinois cross for the shorthair, but this line died out quickly), but if you look into that breed’s history a little more you will notice that the Tervueren is historically very similar to the Dutch Shepherd longhair: it has very few foundation dogs in its own variety, and most Tervuerens are by large shares just descendants of longhairs born into Malinois litters and yellows born into Groenendael litters. Because of this, and because the breed distributions are calculated from the foundation dogs, even purebred Dutch Shepherds will have more Malinois, Groenendael & Laekeneois than Tervueren.
For this post I collected the data of longhair litters from the last three years (born 2017-2019), their inbreeding coefficients and breed distributions. A couple of litters were removed because I couldn’t confirm they were FCI registered, or their pedigrees were so largely unknown it would have twisted the calculations unrealistically. The final count is 62 longhair litters (litter with longhair puppies + at least 1 longhair parent), and on top of that there were 4 shorthair litters with one or more dogs registered as longhairs.
These aren’t direcly related to the main subject of this post, but here’s some interesting data:
62 litters in 9 countries:
Finland 17, France 16, Netherlands 13, Germany 9, Austria 2, Switzerland 2, Belgium 1, USA 1, Slovenia 1
44 unique breeders, of which 12 produced two or more litters
Kennels with most litters: Tarita’s 6 (Finland), du Mont Brabant 4 (France)
Average coefficient of inbreeding: 28.6%
Highest COI: 43.6% (Finland)
Lowest COI: 1.0% (Finland)
Average COI in countries with most litters: Finland 26.2%, France 20.4%, Netherlands 36.5%, Germany 34.8%
I divided these litters into four groups of “purity” based on their three generation pedigree:
– “LH” Pure longhairs (has only longhairs born into longhair litters in the pedigree)
– “SH” Longhairs with pure shorthair (has pure shorthairs or crosses to them in the pedigree)
– “XSH” Longhairs with impure shorthair (has longhair out of shorthair or working line shorthair in the pedigree)
– “XH” Longhairs with x-blood (has lookalikes or breed crosses in the pedigree)
This pie chart shows the distribution of each type in the litters of the last three years. For now, most litters are still pure, but the use of longhairs born out of shorthair parents is increasing rapidly. For example in France, where Dutch Shepherds of this type first originated, there was only two “purebred longhair litters” born during this time, and even these were labeled pure only due to the “impure” dogs being just out of 3 generation pedigree. To balance it is out, Holland had all but one litter either pure, or descended from mixes with pure shorthair.
To allow anyone loving this type of data to really get into it, I listed the averages of all four groups in inbreeding coefficient (COI) and breed distributions. The breeds are listed as follows:
LH/SH/RH = long, short, roughhaired Dutch Shepherd
BM/BG/BT/BL = Belgian Malinois, Groenendael, Tervueren, Laekenois
GSD = German Shepherd Dog
XH = x-line, unregistered shepherd dog
O = other breed, ie. Bouvier des Flandres or a molosser type
(Please ignore the final bar sized 80 units, I used it to standardize the scale for easier visual comparison)
If I was clever, I would end the post here and wouldn’t start analyzing this data any further. The risk is that this turns out to be yet another hours-long rant and I’ll end up burying it before it’s ever finished. But here we go anyways…
Looking at the two first graphs it is immediately clear that longhairs in pure and “with pure shorthair” are pretty much exactly the same thing. Their inbreeding coefficients are very close to each other, as are the breed distributions. On average there was only 2% more shorthair in the dogs with cross background! Because of the regular crosses to shorthair, the pure longhair and shorthair have now become closely related populations. With around 10% inbreeding in the first generation crossing a pure longhair with a pure shorthair cannot even be called an outcross anymore, and crossing back the numbers rise faster every time – by third generation the inbreeding is again as high as it was before the cross. This same story is told by the genetic diversity evaluation of MyDogDNA service as well (GD is directly comparable to a correctly calculated inbreeding coefficient, look to this post), revealing that when pure lines are used, already second generation SH x LH crosses (G1 generation) fall below 25% (= the breed median) in their genetic diversity. It can easily be said that genetic diversity cannot be expected to rise by mixing these groups or breeding within them.
Significant differences are only found when moving to the “impure” groups frowned upon by many today (but always remember the rule of 40 years!). The COIs are on average below 20%, the share of longhair foundation dogs is less than half and the dogs have more novel blood from Belgian Shepherds and x-shepherds. How bad is it for a longhair to have 10% Malinois or 15% XHH? Both of these populations, much overlapping, are so extremely diverse compared to the six-dog foundation of the longhairs that it is possible to gain infinitely more from them than from crossing to pure shorthair.
How big of a share is needed from some other breed before the longhaired Dutch Shepherd isn’t a longhaired Dutch Shepherd anymore? This is a question that everyone will have their own answer to. What we do know is that the more longhair foundation dogs the dog has in it’s makeup, the higher the inbreeding coefficient and the lower the genetic diversity it will have. After plotting the data on a new graph not included here, we get a beautifully fitting trendline suggesting that if the share of longhair was 100% the inbreeding could be nearly 60% – remember that a pairing between full siblings or parent-child is 25%. Getting below that 25% (which is of course still too high) would with this same trend require the share of foundation dogs to be around 50%. Actually this was already established with real-life results, just look at the graph of group XSH again! If we wanted to keep COI under the generally considered “safety limit” 10%, the share of pure longhair should be kept below 30-40%.
I did look for a correlation between inbreeding and the other shares as well, but understandably nothing else gave such clear results. Only the share of x-blood had a clear enough positive link to genetic diversity. Only this fact is a bit harder to use as sort of a “diversity index” than the share of longhair foundation dogs in the pedigree, because x-pedigrees are still much less represented and they are nowhere near as well documented as registered longhair pedigrees.
Finally, another graph looking like the ones before. This one displays the averages for the longhairs born out of shorthairs in 2017-2019.
The future of the variety can be seen here. These are the types of dogs that will be used more and more in longhair breeding. Working line shorthairs truly are more Malinois and XHH than “pure” Dutch Shepherds at all. It is up to everyone to decide for themselves how bad they think this is.
My personal opinion is that using dogs like this carefully effectively removes the urgent need to plan for crossbreeding projects. Why would we need to make another Belgian Shepherd cross (let alone some other breed) when we have working line shorthairs? Why would we need to cross into some distant breed with even more special faults and widely different characteristics, when we have this source of variety available? The Malinois and KNPV lines are genetically so rich we can keep adding from those populations for decades before they lose value to longhairs. And to the people who don’t want to cross at all, I can only say that there is nothing left to fear. It has already happened. In the future the longhaired Dutch Shepherd as well will have a larger share of Malinois and XHH than it has now. In exchange, we’re getting so much more genetic variety and with it, the possibility for improved selection.
Of course the share of “pure longhairs” (the share of the original longhaired foundation dogs) should definitely be conserved, but we should always keep in mind that this was only a handful of dogs, and their influence in the breed should mirror their numbers. Furthermore, these dogs were naturally not “finished products”, but instead many generations of work was needed before it could be said that the longhair existed again. Continuing this thought, my wish is that introducing new bloodlines in the name of genetic diversity should not mean that the breeders taking action want to change the breed to be copies of those new dogs. Personally I believe that every person adding new blood has a plan, and that plan is to use that new material to improve the longhaired variety. Not to change it to something else or to create a new breed, but to improve the traits we already have by carefully selecting for what is needed and leaving the unwanted stuff behind. Looking at those bar graphs before, it is of no use to just bounce those same ratios around for decades: it will lead to nothing but loss of vitality, health problems and degeneration of character traits.
Even if the four pedigree types in this post were invented by me, they’re trying hard to mimic the common views of owners and breeders of the breed in a manner that can be measured. My next, biggest wish would be that in the future we didn’t have yet another set of different groups (worst case: “show” and “work”), but in forty years we would only have one longhaired Dutch Shepherd. It would be a Dutch Shepherd with a long coat, and we could call it the same no matter what amounts of which breed it had in its pedigree.
In the main image of this post I have my Pip, who is a whopping total of 10% longhaired Dutch Shepherd, 30% shorthair, 23% Malinois and 31% x-Dutch Shepherd. If used in breeding she – and other dogs like her – could change the proportions of breed distribution to all kinds of directions, but her use would also lower inbreeding in the population and through that allow a more effective selection for her offspring. Conformation-wise she is not the most typical Dutch Shepherd, but this doesn’t mean her offspring or grand-offspring couldn’t be if the partners were selected accordingly. I am one of those breeders who believe it is well worth it to sacrifice easily obtainable traits for a while, just to be able to make the future for the whole breed a bit more sustainable first.