Your opinion on inbreeding

Originally Posted by bill38112

All modern animal husbandry is based on inbreeding. Basically all the food sources we have all heavily inbred. That's why we have turkeys so heavy they can't fly and cattle that are nearly twice the bulk of their ancestors. It's a good thing. But food animals are not bred for longevity. Hopefully pet animals are.

The examples you cite are not because of inbreeding - they're because of human selection, and one of the methods of selective breeding is inbreeding. A nit-picky distinction, but an important one all the same. Careful inbreeding can also be used to extend lifespan though - it all depends what you're selecting for.

Originally Posted by bill38112

By definition developing new morphs requires inbreeding. What I haven't seen discussed here is "Hybrid Vigor" or the tendency that multi generation inbreeding results in smaller, weaker, less healthy specimens. I bred dogs, cats, birds, and fish before acquiring the snakes. All of those species showed the negative effects of inbreeding after five generations.

Hybrid vigor is the absolute opposite of inbreeding depression (which is what you describe), and inbreeding depression has been mentioned in this thread already. I disagree that after five generations you necessarily can see the negative aspects of inbreeding - do you mean five generations brother x sister, or looser breedings than that - and if the matings are not as close as bro-sis then how can you decide it definitely causes problems after five generations?

I don't really know what the basis is for saying it's okay to mate father->daughter or mother->son but not brother->sister as it relates to deformities. Parents are related to their offspring to the same degree that a full brother and sister are related to one another. In other words, parents share 1/2 their genes with their offspring, and full sibs (brother and sister) share 1/2 their genes. From that stand point, there is no difference in any of those crosses because they are all equally related.

Not quite.
Because of the way genes mix up in gametes, an offspring always has 1/2 of its genes from its parent, but may have MORE than 1/2 of its genes in common with a sibling. To take an extreme example, identical twins share 100% of their DNA with eachother, but only 50% with mom or dad. You have anywhere from 0%* to 100% of your DNA in common with your sibling, though on average it will be somewhere in the middle.

That's why it's sometimes preferrable to pair an offspring with a parent, rather than a sibling. Let's take the example of a parent carrying a recessive gene X, which by itself is harmless. Let's say there's another gene, Y, which also by itself is harmless. But when X and Y are homozygous in the same animal, the offspring dies in the egg. Parent A is XXYy (het for y), Parent B is XxYY (het for x). The offspring will be a mix of XXYY, XxYY, XXYy, and XxYy. Obviously the XXYY's are fully healthy and will never produce deformed offspring. The XxYY and XXYy snakes are likewise fine, because their offspring will never produce the double-homozygous form needed for offspring death. It's only the XxYy snakes that have a potential for disaster. But if the XxYy snake gets bred to either parent, the offspring will all still be healthy. If the XxYy snake gets bred to an XXYY, XxYY, or XXYy sibling, all the offspring will also be healthy. But if the XxYy snake gets bred to another XxYy snake by accident, you will get some dead offspring. (This is a ficticious example designed to illustrate the theory, for those who are wondering.)

Conversely, if x and y were amel and anery, you'd want to breed siblings if you wanted snows out of the mix. It's simply the same theory in reverse.

-Kat

*I'm referring to the areas of DNA which differ between individuals in a species. There's a large percentage of DNA which is identical between members of the same species, so the percentage of total DNA that differs between siblings will actually be quite small... but for the part we care about, the explaination holds.

Originally Posted by Kat

Not quite.
Because of the way genes mix up in gametes, an offspring always has 1/2 of its genes from its parent, but may have MORE than 1/2 of its genes in common with a sibling. To take an extreme example, identical twins share 100% of their DNA with eachother, but only 50% with mom or dad. You have anywhere from 0%* to 100% of your DNA in common with your sibling, though on average it will be somewhere in the middle.



*I'm referring to the areas of DNA which differ between individuals in a species. There's a large percentage of DNA which is identical between members of the same species, so the percentage of total DNA that differs between siblings will actually be quite small... but for the part we care about, the explaination holds.

I was exlcuding identical twins. There is always the chance that sibs don't fit the .5 average, but unless we know the genetic makeup of the individuals our best estimate is r=.5. In any equations or calculations one would never randomly assume r to be equal to .325 or .768. I just feel like the average is more useful to illustrate the point.
A parent may have exactly 50% in common with it's offspring, but statistically, a brother or sister has 50% in common as well. I don't know if I'm making sense or not. I'm just saying that deviations from .5 for full sibs is not the norm, and the best value we can use in everyday talk about relatedness is the average or .5.

Zach, my point is it's the fact that it CAN differ that causes the difference (however slight it may be) between offspring to parent and sibling to sibling pairings.

It's like the difference between .00001 and 0. Sure, .00001 is pretty close to 0, and for most applications it doesn't matter which one you use, but if someone asks you what the difference between .0001 and 0 is, responding that "They're both the same." is wrong. And more importantly, it's not a useful answer.

-Kat

Originally Posted by Kat

Zach, my point is it's the fact that it CAN differ that causes the difference (however slight it may be) between offspring to parent and sibling to sibling pairings.

Agreed.

It's like the difference between .00001 and 0. Sure, .00001 is pretty close to 0, and for most applications it doesn't matter which one you use, but if someone asks you what the difference between .0001 and 0 is, responding that "They're both the same." is wrong. And more importantly, it's not a useful answer.

I just feel like you're analogy is misleading. I'm really not just trying to be difficult or argumentative. Using a value of r=.5 to describe full sibs is not the same as what you are suggsting. Using such a value is certainly more proper than just picking an arbitrary value between 0 and 1. Of course, this method may be based on a certain assumption, but how many theories/calculations/etc in biology are based on assumptions but still widely used because they generally hold true under a wide range of circumstances. Hardy-Weinberg takes making many assumptions that could change the numerical outcome, but it is still is a very useful tool for making inferences (even when some of those assumptions are violated).

I agree that we should always use the correct/absolute values when we know them. My point is that, when we don't know the specific relatedness of two full sibs, they are typically designated as being .5 because that is the most accurate estimation we can use. It's true that the value may actually deviate from the mean, but we don't know that for sure and so we don't just randomly assign some value as the relatedness.

My point is that, when we don't know the specific relatedness of two full sibs, they are typically designated as being .5 because that is the most accurate estimation we can use.

My point is that the original question was what the difference between parent->offspring and sibling->sibling was. In this case, using the approximation of .5 for sibling relatedness is incorrect, because it doesn't show the difference. For MOST applications you can use .5 because it simplifies things. But in this case, you cannot. It is not useful to say that parent->offspring and sibling->sibling pairings are identical, because they are NOT identical. For parent->offspring comparisons, the percentage of genes related is ALWAYS .5. For sibling->sibling comparisons, the percentage of genes related is ALMOST NEVER EXACTLY .5, but more than likely something close to that number.

-Kat

The degree of relatedness, or shared genes is going to be an inexact quantity, given that you will never know the exact degree of relatedness of the original pair.
Inbreeding, whether parent to offspring or between siblings is going to increase the degree of shared genes over generations, this will increase the chance of desirable traits being reproduced, which is why it's such common practise in animal breeding. It will only increase the chances of genetic defects if those defects are present in the original stock.
Unfortunately the presence of 'bad genes' isn't something you'll know about until they are expressed, but it's seen as a chance worth taking to improve most domestic stock, be it poultry, cattle etc or pets

Originally Posted by Kat

My point is that the original question was what the difference between parent->offspring and sibling->sibling was. In this case, using the approximation of .5 for sibling relatedness is incorrect, because it doesn't show the difference. For MOST applications you can use .5 because it simplifies things. But in this case, you cannot. It is not useful to say that parent->offspring and sibling->sibling pairings are identical, because they are NOT identical. For parent->offspring comparisons, the percentage of genes related is ALWAYS .5. For sibling->sibling comparisons, the percentage of genes related is ALMOST NEVER EXACTLY .5, but more than likely something close to that number.

-Kat

Okay, now we're getting somewhere! Would you agree that any given full sib pair is just as likely to have r<.5 as they are likely to have r>.5? From that perspective, wouldn't it be inaccurate to say that a parent->offspring mating is necessarily "safer" than a full sib mating?

My, people do like to make things difficult. To clarify my previous remarks:

In breeding is a form of selective breeding and ALL animal Husbandry uses it. The Cat Fanciers Association is 100 years old this year; The American Kennel Club is 126 years old; As Bill Love's excellent documentary, REPTILEMANIA, points out, the domestication of snakes is in its infancy. There are almost no studies on snake genetics available. I am, therefore, relying on almost 40 years of my personal experience and research with other species.

As for the application of selective breeding and in breeding (and by in breeding I mean brother/sister and parent/offspring pairings. Less closely related pairings, I consider line breeding). In breeding specimens with the visible desired trains will produce the trait in off spring more reliably than when breeding unrelated specimens exhibiting the trait. I stand by the turkey and cattle example.

As for the discussion of hybrid vigor or "inbreeding depression" I suspect that there are very few of you out there who have five generations of your own breeding to test my theory. I am basing the five generation rule on a study done with birds by the San Diego Zoo in the 70's. I found that data to be reliable in dogs, cats, birds and fish, all of which I have bred to the fifth generation or more. As far as the hetero/homo-zygosity of your initial population, tell me, who has viability data on their snakes for even three generations? The five generation rule is based on the "probability" that fertility, viability issues OTHER THAN specific disease will manifest within five generations.

Bottomline, In breed to develop the traits you want, then line breed to maintain that trait, with an occasional outcross to maintain vigor.

Originally Posted by bill38112

As for the discussion of hybrid vigor or "inbreeding depression" I suspect that there are very few of you out there who have five generations of your own breeding to test my theory. I am basing the five generation rule on a study done with birds by the San Diego Zoo in the 70's. I found that data to be reliable in dogs, cats, birds and fish, all of which I have bred to the fifth generation or more. As far as the hetero/homo-zygosity of your initial population, tell me, who has viability data on their snakes for even three generations? The five generation rule is based on the "probability" that fertility, viability issues OTHER THAN specific disease will manifest within five generations.

One study hardly makes it a rule. Personally, in all seriousness, I would love to read the findings of that study if you have a link or article title. 3 generations of viability data? Well, I'm sure there are lots of members who have stats of such things. Many members track how many fertile and infertile eggs they have, hatchling deaths, etc... Do they have the same degree of inbreeding as you are proposing? No, probably not quite as much. Changing the subject to viability does nothing to address the issue of intitial hetero and homozygosity. I believe it would have been more accurate to say "The five generation rule is based on the assumption that fertility, viability issues OTHER THAN specific disease will manifest within five generations." because a few limited observations just doesn't establish a "probability".

There's no precedent to warrant saying that "There is a probability that x or y issues will be present by generation 5." At least not based on such limited observations. Like I said before, I'm not discounting that you observed such things. It wouldn't make any sense for you to lie about it, especially if people like me are gonna give you a hard time It's just that when I feel like something is being misrepresented, I want to put my .02 in. I just felt that implying rules or probabilities based on limited info is premature.

And to go so far as to even give your own theory support...absence of evidence is not evidence of absence. There may well be some such thing as a 5 or 10 generation rule with certain species when it's applied to certain parameters such as some sort of viability or whatever. However, I still believe that it would be at least partly dependent on the species in questions and the intial population size and genetic structure.