#also I mean it's not like Matilda's breeder doesn't repeat her crosses more than I personally would

[grison-in-space]

man, this is often how I feel when I'm making decisions about my dogs' health. I don't do breeding--Tilly will probably be spayed in a few months, I'm clearly not going to get off my ass to go to conformation classes at this point--but even questions that don't hinge on bringing new life into the world are really hard to interpret, especially when the data just isn't all the way there yet. This is one of those things where you have to grapple with an ethical accounting where none of the answers are good, and where there are upsides and downsides to literally anything you might choose to do; you're the only person who can decide what will let you sleep at night, you know?

I'm a scientist, sure, but I'm not exactly this kind of scientist: I can evaluate the lit, and I'm fluent in population genetics, but I've never worked directly on dogs and I am only one flawed human being. That said, I have some thoughts behind the cut here.

Here's the thing: if I was you, I would not rule out breeding him to a bitch with a Grade 1 or Grade 0 back screening, given everything else about him, with full disclosure to any puppy buyers. I say this because, well, you're willing to breed a Dachshund and not march at full speed to another breed. I don't necessarily think that breeding him to a Grade 1 or 0 bitch is worse than breeding any Dachshund dog to a Dachshund bitch, and I can see some fairly strong arguments to treat this like a strong fault rather than a total disqualification from a breeding program.

The thing about dog breeding within a closed studbook, especially in a breed like Dachshunds, is that you're closed in with a lot of hard decisions to make. As You Know, Bob, there are straight up not a lot of Dachshunds with back screening to choose from out there, and most of the ones that do exist do not have ideal back scores. You've seen the genetic data re: FGF4 insertions, right? It's just straight up fixed in all of them. Additionally, IVDD screening is complicated because the phenotype we care about, which is actual disc extrusions, isn't the same as the number of calcified disks that the screen measures. This is actually really similar to hip dysplasia in that PENNhip and OFA tests don't measure the thing we really care about in dogs, which is the crippling arthritis and lameness that often (but don't always) come alongside a dysplastic hip conformation.

You've heard me talking about hip dysplasia and why loose-ligamented, extensiony GSDs suffer so much more of the associated arthritis than dogs with tight joints and ligaments that hold the joint in place, right? That does not mean that breeding for good hips isn't important, and it doesn't mean that you can't have a long-striding dog with perfect ethics, but it does mean that you have to select much more intensely on hip ball/socket fit in GSDs than you can get away with in other breeds in order to not subject the dog to the chronic pain aspect of the disorder. IVDD is going to be similar in that respect: disc calcifications correlate with disc extrusions, sure, but they are not quite the same thing. (This is especially true since interrater reliability for disc calcifications is, uh, not that good. I would personally want multiple blind expert opinions on my dog if I was trying to make decisions based on the score; I don't know if that's a feature of your scoring system offhand.)

Breeding is, by definition, about selection: not all dogs produced will be bred, right? A good breeding program selects the traits that we want to carry forward, and from there we can fix them. But when we select on good stuff, we carry a lot of things with that and concentrate it. Ideally, we can outcross to refresh our population with the neutral stuff and select from the offspring to bring back the traits we want. The trouble is that this process takes time: the looser your selection and the more you wash the general population back into your selected predictable dog population, the slower you go and the more variation you retain. The faster you move setting your type or selecting for the variants you want, perhaps by linebreeding or exercising very strong selection on your population, the more things you absolutely do not want will be carried along for the ride. For Dachshunds, propensity for IVDD has been carried along and--in the form of the FGF4 insertion--fixed. (In this case the fixation might just be a consequence of the most extreme dwarf phenotype or it might be something that is relatively neutral for Dachshunds.)

The problem with a closed studbook is that you have no way of re-infusing that neutral to positive variation once that studbook closes. So for example, if you pivot only to back-screened dogs, you lose Kermit's entire sub-population of falconry dogs into the bargain. That whole group gets washed away, along with any allelic variations specific to those dogs that might be useful later. This is what it means when people talk about dog breeds being like conservation of endangered species: they're not, you can outcross any time you want except for the cultural stuff, but if you close the studbook and ban all outcrossing permanently, you will eventually wind up with a gene pool that is too damn small very easily. Once you're there, it can be impossible to rescue the animals, and if you even can, you have to breed very carefully to maximize the genetic impact that every single individual born into the population has in order not to lose more beneficial variation. This is where Doberman breeders are at now.

Now, the trouble with building a fence around the law when breeding dogs and emphasizing breeding for a trait that correlates strongly with health but isn't the same thing as it is that this constitutes a form of very strong selection on the population. This goes triple for a breed like Dachshunds where the problematic disease state is extremely common, because you have to filter out even more of the dogs and shrink your potential pool of breeding animals. It's possible to concentrate down on a totally different disease in doing this, and that's the thing I personally would worry about from the data I've passed over in Dachshunds. (Ah, forgive me for the infodumps; I just keep checking the literature because, well, it's such an interesting frontier to investigate and I like to know how we develop new bodies of information.)

One thing this means is that purebred dogs are going to have different levels of heritability for complex traits left to work with. Heritability is not "how genetic a trait is," it's more like "how much of the phenotypic variation we see in the population (here number of disc calcifications) is attributable to genetic variation, and how much is the result of other factors?" If you've fixed all the genetic variation in the breed, heritability will be very low even if the disease is clearly genetic and found in high frequencies in your breed. I could find two estimates of heritability for disc calcification in Dachshunds, but both involve small sample sizes: Stigen 1993 estimates heritability as low as 15-20%, while Lappalainen 2015 reports a much more optimistic (for breeders) range of 43-52%.

Unfortunately, I should note that I think the Stigen paper is better even though it's older: the Lappalainen paper studies the radiographs of all the dogs submitted for spine analysis to the Finnish Kennel Club since 1997, while the Stigen paper seems to be a more general sample of dogs that has less in the way of population stratification. The Lappalainen results are also counter-intuitive to me in a few different respects: first, it seems bizarre that the genetic heritability of disc calcification should increase after 15 years of selection on this trait, and it does not seem that the intervening 15 years of selection has actually reduced the level of phenotypic disc calcification in the breed. There are a few estimations of breeding values in the Lappalainen paper also that are not clearly connected to direct values that have me peering at its conclusions with a certain amount of skepticism.

Another thing that occurs to me is that it's hard to get a sense for what other factors are creating potential for injury beyond simple disc calcification. When you reported Kermit's spine readings, I confess that my first thought is that he's an incredibly active little dude: between the agility and more importantly the ratting, his spine has a lot of insults that could develop into an injury in his every day life. That said, we know that muscle tone and ligament strength can have a protective effect on actual disc extrusions: what if the actual meaning of the spine calcification read is that Kermit's abdominal wall strength prevents discs from actually extruding even if they calcify? I really find Rohdin 2010 convincing on this: she finds that calcified discs as measured by xray are not actually more likely than non-calcified discs to cause extrusion injuries, and that the level of internal calcification is not always visible from xrays alone as confirmed by histopathology.

What I would personally do is screen the shit out of everything, use the outcomes to determine strengths needed when pairing dog to bitch, and move forward with my eyes open--not least because when you're working with the best dogs you can find for ratting ability, you're going to be limited in what you can get access to in order to breed into your lines. I would also keep detailed records with puppy buyers if at all possible and cross back and forth a fair bit across various types of lines if possible. Across dogs with Kermit's activity level, what do spine calcifications look like? We just do not have the data to ascertain that clearly right now, as far as I can see.

Jumping off what I just posted, when I said I was pulling Kermit from breeding due to the number calcifications in his back and some stuff going on with his littermates, a (American) dachshund breeder told me I was making a mistake.

I won't go into the whole conversation, but among other things she said that back screening is a pseudoscience and it's foolish to base breeding decisions on "just" calcifications. She also said that she is a veterinary neurologist who regularly deals with IVDD in her patients and she totally ignores calcifications, "so you can see how much stock I put in back screening".

And. I don't know.

On the one hand, yeah, it's not perfect. I know it's not perfect. I've been treating it like screening for hip and elbow dysplasia. The information it gives us is limited at best, it does not guarantee offspring will be ED/HD/IVDD-free, and it's basically a shot in the dark. But it's better than nothing. It's better than totally ignoring the problem.

I didn't say anything back because I'm not interested in a debate, but the people who are writing the papers on back screening are also veterinarians. It's not like they are laypeople who invented some snake-oil scheme to sell to unsuspecting plebs.

And I don't know, I feel like doing something is better than doing nothing. Nearly every American breeder I've talked to is doing nothing. And quite a few of them hold "those Europeans" in disdain. Like, don't you care? Don't you care about your dogs? Don't you want to get rid of this terrible thing plaguing your breed? Isn't selecting for more normal backs better than loudly declaring "my dogs are healthy" and vilifying owners whose dogs herniate a disc? Don't you care??

But what do I know. I'm just some rando who has an Animal Science degree but isn't a veterinarian or a breeder or a scientist. I'm just doing my best to understand.