I just saw a seminar that I thought may be of interest to the group. It was given by Mike Lynch (Uof Oregon), and although the talk focussed on mutation accumulation and something called Mueller's Ratchet (don't worry if you don't know what this is -- it's not important to this post), there was a very cool take-home message about captive breeding programs. As part of the experiment, Lynch's group has set up a variety of captive cultures of animals ranging from mice, rice and corn to bacteria, anemones and nematodes (and many species in-between), and have maintained lines of the animals under "ideal conditions" for as long as they can. There are two basic ways this is done: animals capable of asexual reproduction are maintained by clonal propogation (the equivalent of coral fragging) to produce each "next generation." Obligately sexual animals are maintained by breeding each family, and allowing it to contribute more-or-less equally to the next generation (a practive common in animal husbandry for conservation -- this is done to preserve the maximum amount of genetic variation present in the initial population). The experiment that Lynch's group is conducting aims to follow those family lines and examine the accumulation of mutation in each line to see what is the mean effect on fitness. They set up 1000 lines of the organism of interest and see what the fitness of the individuals are as a function of time.

The results were very interesting! Lines started to go extinct (despite ideal conditions) pretty quickly. The mean time to extinction was directly a fuinction of "generation time" (the time from one clonal propagation to the next for asexual lines or the time to reproduction for sexual ones), with the species with shorter generation times going extinct more quickly than those with long generation times. There was, of course, a distribution of performances among the lines, but there was an exponentially increasing number of extinctions in all cases they examined so far. They plotted the number of lines that died out as a function of reproductive rate (#generations rather than absolute time), and the pattern was indistinguishable for every one of their organisms - some lines went extinct in as little as 5-10 generations and others hung around for 500 generations, but the number of cumulative extinctions increased exponentially until they lost the entire experimental line (at least in the bacteria, nematodes and other organisms with very short generation times - the rest of the experiments are still underway).

So what does that mean, and why should anyone care? Well, the bottom line for us is that much of our captive breeding efforts for commercial purposes are directed along similar techniques. For example, coral propagation is generally done by fragging, and bouts of true sexual reproduction are rare, even in facilities devoted to culturing critters for our reef tanks. Thus, the results of this experiment seem directly applicable to the culture techniques currently employed in the aquarium industry. If the results from this experiment are general across many of the taxa not yet tested (which seems likely given the taxonomic breadth of the initial sampling), then it could be expected that coral propagation as we know it today is doomed to failure. It may not happen in our lifetime (although 10-500 "generations of fragging seems possible within our lifetimes), but the reproductive ability and growth rate of the fragged corals (if they follow the same pattern as the tested species) will eventually decline to below the death rate, and the animals will die out. What does that mean to us? Unless captive breeding programs can change the way they maintain their stocks, all the lines they maintain should eventually suffer the same fate as the variety of organisms in Lynch's studies.

I thought that this may be of interest to some other people on the list as well - at least it's food for thought....

Clearly there is a known problem of multiple generations of "inbred" organsisms accumulating mutations. Inbreeding is recognised as a potentially hazardous reproductive strategy for precisely that reason.

David has hit on a lot of the salient points that I would make in response to the recent posts along these lines, but there are a few I should add. The first and foremost has to do with the inbreeding comments -- these "lines," as I called them, were not inbred. The effective population size of the lines was on the order of that seen in natural populations and inbreeding was effectively zero. Regardless, inbreeding has nothing to do with mutation per se -- it has to do with the probability of expression of recessive deleterious traits. The experiment of which I spoke was looking at the accumulation of mutations which occur on a cell division basis rather than a "reproductive event" basis. With each cell division, there is a finite probability that there will be an error in the replication and a mutation will occur, and that event is irrespective of inbreeding. Inbreeding makes the problem worse because the vast majority of mutations are bad. If the mutation acts in a standard dominant/recessive fashion and that mutation is rare in the population, most individuals with that mutation will also have a copy of the dominant allele to save them from expressing that "bad gene." With inbreeding, homozygosity increases and the probability of expressing that mutation increases, but that doesn't mean that inbreeding is an explanation for the results I described.

Even more interesting is the fact that some corals are *obligate inbreeders* under natural conditions, and those populations are viable and "normal" despite inbreeding coefficients of about 0.8 (1.0 means you always breed with a brother/sister). There are natural situations in which inbreeding is both a natural and desirable thing, and mutation accumulation is an entirely separate issue. The preoccupation with inbreeding is somewhat irrelevant to this discussion, especially since the lines used in the experiment were intentionally not inbred.

Apparently a little sex goes a long way, since many organisms (many common aquarium inverts, for example) have occasional bouts of sexual reproduction in a long line of asexual reproduction.

David is correct in pointing out that a little sex goes a long way -- it only took periodic sexual reproduction among the asexual lines to eliminate the effect of which I spoke in the initial post. With bouts of sexual reproduction the lines became essential "immortal" (for the duration of the experiment) and very few, if any depending on the species, went extinct over time -- the reason is the one I just outlined above, recombination.

In the wild, a generation of offspring cloned from an individual would remain in the colony and start reproducing themselves. Clearly, there is a potential for mutations to occur and to be passed on - but only to their own offspring. In subsequent generations, additional mutations might be accumulated in some of the next generation. This may become damaging. In the context of intense competition in a clonal colony, natural selection will tend to weed out these weakened individuals, so the genetic flaws die with them. Healthy, undamaged offspring continue to reproduce, as does the original, producing more healthy offspring.

You have the basic idea, but haven't thought it through very carefully -- there is no going back, right? Sure the parents and offspring without a mutation will do fine and those with the mutation may not, but you're only thinking of a couple of generations here. How long do you think a parent can give rise to offspring without *ever* getting a mutation? If the probability of mutation is on the order of 1 in 100 (the maximum reported in the study to which I referred), you have about 100 offspring on average before you have a mutation arise. Being clonal, those offsring can't go back, and so over long periods of time the entire line begins to accumulate mutations. There is some distribution of the number of mutations in the population (0,1,2,3,etc.) but each time you move one down the scale, you're stuck there until another mutation hits and you move further along -- that is the phenomenon known as Mueller's ratchet (the ever-tightening grip of deleterious mutations) to which I referred in the original post.

Someone asked why natural populations of anemones don't go extinct -- the answer has to do with a combination of things. First, we are talking about times to extinction on the order of 500 generations, and with an organism that can live 200 years (and counting) in an aquarium, we're *pretty unlikely* to ever witness such an extinction event, even it it did occur! Second, the occasional bout of sexual reproduction seems to "reset the clock" so to speak, because some lines with accumulated mutations can, by chance, give rise to offspring without any of those mutations being present, and re-create the "0 mutations" class. Finally, there are some cases of apparently deleterious mutations in one environment proving themselves beneficial in another -- environmental variability always plays into this and is an unpredictable factor in the evolution of populations.

The extinction of a species due to a lack of genetic variability is a result of only having a limited breeding stock.

There was no difference in the amount of genetic variability (measured as degree of nuclear and/or mitochondrial microsatellite polymorphism) between the initial and final populations of the most lines. The differences that were observed overall were not significant.

Although some organisms seem to be able to clone indefinitely (especially plants), among the vertebrates, no known clones (say three times fast) are very long-lived.

Actually this is not entirely true -- there is an asexual line of true parthenogenic salamanders that has existed for as long as the Tiger Salamanders (Ambystoma) to which they are the sister taxon. But that is a famous exception to the rule, and for the most part what David said holds true. Bdelloid rotifers are the other famous example (there are no males), and it is these famous counter-examples that feul a lot of the research in the field to try to understand how it works. A few backwards polymerases that have enhanced DNA-proofreading have apparently been found in some lines, but there is no general explanation for these exceptions to the rule as of yet.

As for the experiment Rob describes--what's missing is PREDATION, a.k.a. CULLING. What we consider "ideal" conditions often means "everyone survives." While sexual reproduction undoes a lot of the ills of cloning in terms of adaptive fitness, it carries the potential for inbreeding to fix undersirable traits when fitness no longer affects the gene pool.

This is the important point for the original post -- the only selection that occurs in the experiment is viability selection. What I mean by that is the animals that develop a lethal mutation will express it and die -- they are irrelevant to the experiment, because they leave no offspring. The animals that develop a mutation that is a little "bad" however can reproduce and continue to spread that mutation throughout the population because there is no consequence to having it under "ideal conditions." It is only when the presence of a large number of little defects start to have severe consequences to the ability of an individual to reproduce that the population begins to suffer, and at some point the reproductive rate falls below the rate of natural mortality and the population dies out (extinction of that line).

FWIW, some people have argued that modern medicine and technology has moved human evolution into this framework -- people survive to reproduce in modern society that would have not been capable of surviving, much less reproducing only 100 years ago....

Did anyone ask if they thought the death as a result of certain number of generations was a result of loss of telomeres.

No one specifically asked this, but he sort of addressed it -- there was no clear relationship between the number or length of chromosomes and mean time to extincction, but there was a nice relationship between generation time and absolute time. The general conclusion was that it was simply an accumulation of slightly deleterious mutations (on the order of 1/2n effect or less) that eventually leads to viability problems.

Jake Levi

Posted to The Breeders Registry emailing list, Sunday 13th December 1998.

Fascinating post, wish I had been there, would have loved to hear the presentation, and been there for the question and answer.

The major significance that I see in his study is its bearing on the continued existence of domestic lines of livestock, and the maintenence of viable breeding lines. I would think that the protocols used within livestock breeding would be applicable to captive breeding of marine organisms, in fact, I would expect that any aquaculturist worthy of the name would be applying the normal genetic techniques utilized for centurys in breeding domestic livestock.

I well remember back in the 70s hearing Dr Graham at Ohio State lecture on developing stable breeding lines in fisheries science courses that he taught there. One of his favorite examples was the salmon breeding project of the U of Oregon. Later, within the ornamental fish industry I saw the fish farms in Fla regularly swapping breeding stock to maintain or improve pond populations of aquarium fish, so, in short, I would have had a lot of questions to ask the gentleman just from my own observations over the years within commercial aquaculture.

From my own observations of extinctions of domestic livestock breeds the factor that I have seen most operative for extinction isnot failure of a line or species due to viability but replacement by another line or species, and discontinuing the breeding of the line or species. Within the US since the turn of the century we have gone from over two dozen established breeds of dairy cattle to one breed, with three major lines in that breed representing over 90% of the dairy cattle in the US. More then a dozen of the other breeds have become extinct from being dropped in place of other breeds. Six other formerly major breeds now account for less then 10% of cattle numbers, and one strain has less then 200 breeding animals. It is even more drastic among poultry breeds with most of the presnt existing breeds being relegated to hobbyist or fancier breeds and three crossbred lines producing the majority of the fryer chickens and one breed producing the egglayers.

Just more food for thought. Economic or aesthetic tastes being responsible for the greater number of domestic livestock breeds extinction then viability of lines. I cannot come up with a single breed of domestic livestock that has become extinct through a lack of viability of the breed.

Rob Toonen

Posted to The Breeders Registry emailing list, Tuesday 15th December 1998.

The danger to avoid in captive breeding programs is the economic temptation to keep, raise, and sell every single offspring produced. That is the primary message of the study -- there has to be either an active culling of "irregular" offsring from the stock or the accumulation of mutations in the lines is an unavoidable factor. I think that livestock lines are pretty good at doing this because the market is for dinner rather than for ornamental live animals. If you have a livestock situation, the breeding animals are those few that are "most desirable" within the context of the breeding design. Culling the remainder means profit, and those few "most fit" individuals kept as breeders are actually non-productive in terms of the economic base of the trade. In the ornamental/conservation industry, there is an effort to keep, raise and sell *every* offspring (barring significant deformity) produced, and it is this practice that leads to the accumulation of deleterious mutations and eventually to viability selection.

This is the important point for the original post -- the only selection that occurs in the experiment is viability selection. What I mean by that is the animals that develop a lethal mutation will express it and die -- they are irrelevant to the experiment, because they leave no offspring. The animals that develop a mutation that is a little "bad" however can reproduce and continue to spread that mutation throughout the population because there is no consequence to having it under "ideal conditions." It is only when the presence of a large number of little defects start to have severe consequences to the entire population that the reproductive rate begins to fall, and at some point the reproductive rate falls below the rate of natural mortality and the population dies out.

FWIW, some people have argued that modern medicine and technology has moved human evolution into this framework -- people survive to reproduce in modern society that would have not been capable of surviving, much less reproducing only 100 years ago....

Jake Levi

Posted to The Breeders Registry emailing list, Tuesday 15th December 1998.

I am in quite agreement with your post, especially the last paragraph, unfortunately there are a great many folks having children today that would be better to adopt, but, thats another subject.

A point in support of your comments is what is happening within the ornamental freshwater fish today, and the marketing of 'balloon bodied' fish in many species, and other defectives.. These are ones that would have been left on the pond bank just a few years ago, but, the 'drive for something different' leads people to buy them and the farms to produce them. The problem being that there isnot an infinite number of fish farms or ponds on them and these defective fish wind up in the other breeding populations. Murphys Law is alive and well in farming as all other enterprises.

I have also seen many deformed and defective marines, mostly clowns offered for sale that should have been used to feed triggers or anemones. But not sold. Culling, for the good of the species has to be done, whether its a calf, or a guppy, or a trigger fish. This is just one factor I was referring when I made the general statement of 'using the protocols within livestock breeding'. I should haved been more specific with some examples.

To a certain real extent we are on the thresh-hold of marine mariculture and it is a great temptation to keep all the fry or cultivars of a breeding or division, but culling has certainly got to be learned and practiced for the benefit of all of our species that we keep. Only the best should be our breeding stock. The rest culled, by leaving out of the breeding population.

Without this tool we accelerate the risk of the destruction of those species that we try to produce.

© 1998 -