Contents Updated: Monday, September 13, 1999
Richard Dawkins argues that virtually all the mutations studied in genetics laboratories—which are pretty macro because otherwise geneticists wouldn't notice them—are deleterious to the animals possessing them. Since he says "virtually all", the implication is that some macro mutations studied in the laboratory are beneficial, or at least neutral, to the animal possessing them.
Even if the virtually were erased, one is entitled to ask how an evolutionist could imagine that, however many experiments were carried out in laboratories, the vastness of the range of nature's experiments could be reproduced. Even accelerated evolution usually depends upon passages of many millennia, impossible to simulate in a laboratory. This is especially true of saltation precisely because beneficial macromutations are, as all agree, rare.
But Dawkins readily accepts that the type of mutation he refers to as the stretched DC8 macromutation is not unusually rare. Like the DC8 aeroplane lengthened by adding a new section of hull, mutations of this sort would put a complete additional segment into a millipede, extra ribs into a snake or an extra digit to a human hand.
According to Dawkins these are really micromutations. Although they are major changes for the complete organism, they are only small changes to the genetic instructions.
If you were to write a computer program to print five asterisks in a row you could write a program to print one asterisk then put the program into a loop telling the computer to repeat the instruction five times. If you found that six asterisks had actually appeared you would not think that the complete set of instructions to print another asterisk had miraculously been duplicated. You would realize there was one small error—you had mistyped six for five in your program which had then looped six times not five.
Dawkins says that disposes of saltation—when it occurs it is not a macromutation at all. But it actually justifies it. It makes macromutation at the whole organism level more likely, depending as it does on only a micromutation at the level of the program, the genetic code.
Small changes in genes that are subtly linked together to influence a complex of apparently unrelated features can have profound changes on the organism. The environment effects natural selection on the whole organism not on the genes themselves, so it is the macro effects that matter in the selection process.
Some micromutations can simply be the re-expression of a previously unexpressed gene or group of genes. The interesting aspect of this effect is that it can cause apparent violations of Dollo's Law in that features which have apparently disappeared can reappear.
When a macro feature is lost during evolution it does not follow that its genetic blueprint has gone, simply that it has switched off. The complete genetic blueprint is never fully expressed in advanced animals. The inoperative bit of genetic code will be lost eventually. Mutation will lead to reprogramming of the redundant piece, although that might be millions of years later.
A feature which has disappeared has done so for a reason. It is unlikely that conditions would turn again in its favor before the underlying code had been altered—thus upholding Dollo's Law. But, though usually so, it need not always be true and lost features can reappear, albeit rarely. If the section of genetic code remains intact, it can occasionally be expressed by accident causing atavisms or throwbacks. If the throwback happened to be again of value to the organism, it could again be selected.
Atavisms, as one would expect, mainly look bizarre and are not of any benefit—quite the opposite. Three toed horses are not particularly uncommon, the extra toes not being a crippling deformity, but whales with rear limbs are more unusual. Both represent throwbacks of tens of millions of years of evolution.
No birds today have teeth but experiments have shown that birds are capable of growing them given the right conditions. Surgical manipulation in a chicken's foetus can induce structures to grow that normally would not. The genetic instructions for their growth must be present even though they are not usually expressed.
Furthermore, the growth of some structures induces the growth of others. The fibula in a chicken normally has atrophied to a splinter. Yet if it is encouraged to grow until it reaches the ankle, lo and behold, ankle bones, that normally do not grow at all in a chicken, appear.
The hoatzin chick is, in some respects, a throwback to the archaeopteryx. It has the three grasping claws on its forelimbs seen in the archaeopteryx but not present in other birds including the hoatzin's near relatives. They recede in the adult form. Evolution seems to have partly revived an otherwise discarded plan by bringing it out of genetic storage because for one odd bird it has proved advantageous. The hoatzin chick can cling to reeds, roots and branches above the marshes where it lives.
The apparent loss of the wishbone or collar bones in dinosaurs seemed to preclude them as ancestors of the birds. Consequently Heilmann in 1925 proposed that primitive reptiles called pseudosuchians were the ancestors of both birds and dinosaurs. Birds retained the wishbone but dinosaurs, in the main, lost it. What renders this theory untenable is that the archaeopteryx with its wishbone is so uncannily like some dinosaurs without it. Possibly some of the bird branch and some of the dinosaur branch evolved in parallel, having split from the pseudosuchians.
But the anatomy of archaeopteryx and its dinosaur twins is so close except for the wishbone that convergence seems less likely than that the collar bone reappeared as a throwback in species that had lost it. If atavism is caused by failed suppressor genes then even though dinosaurs like deinonychus had lost their wishbones, the genetic information for wishbone production still existed and could be recalled. In archaeopteryx, it proved advantageous to do so. Powerful flight muscles anchored to the revived wishbones were essential to flight and the evolution of birds.
Having found itself with a new, or revived, macro-feature, the mutant organism, whether it be bird or bacterium, finds itself in a new evolutionary channel possibly leading to an undiscovered Shangri-La in the evolutionary landscape. With its novel characteristics, the mutant can evolve rapidly and, ultimately, disperse into the niches available to it.
In the evolutionary landscape lakes are stasis and torrents tumbling down steep hillsides are rapid evolution. Both are understandable in the ways outlined above, so, long periods of little change and short periods of rapid change can be explained provided that the conditions are suitable for each.
But could evolution be yet more highly directed? Could the development of the embryo be an initial mutational filter, a preselector especially valuable in selecting viable saltational changes?
Embryos spontaneously abort if a macromutation is unsuitable for embryological development. Subject to micromutations only, the embryo will survive until birth because small changes will not normally affect its viability. But macromutations very often are not viable for survival of the embryo. The windpipe might be missing or the heart punctured. The brain might be large but the neck too weak to support it, and so on. Development of the embryo is a filter for macromutations. Rejects are spontaneously aborted stopping the parent from wasting time and effort on unviable offspring.
Hence the only macromutations that see the light of day are those that permit survival of the embryo until birth. Many, if not most, of those will also die, but successful embryological development is the first stage of natural selection.
Evolution might be a vector quantity, having direction as well as magnitude. A species in stasis and subject to no particular selection pressure experiences mutations in all directions in evolutionary space and most are neutral or selected against. But once evolution towards a new equilibrium position begins to occur, mutations in one direction can be favored by selection—evolution can therefore be a vector quantity in evolutionary space complete with its own momentum.
Not all genetic mutations are equally likely. Mutation can be itself controlled to some extent by a gene. That is not to say that all mutations are thus controlled, random mutations will still occur, but a gene could tag another gene or mixture of genes for change when appropriate. Such a gene could control the extent to which a group of genes mutate thus effectively providing a mechanism for macromutation in the right conditions. It is a gene having mutagenic properties leading to saltation, a saltatory gene, a saltagen.
In stable conditions the saltagen, which itself mutates readily, would be switched off. Obviously it would have no effect on evolution. If it switched on, since the species is in stasis, the mutations would tend to be unfavorable and selected against.
If conditions changed, and the saltagen switched on some mutations would be favored and selected. The saltatory quantum would be one, one gene at a time would mutate, it would be in its x1 quantum state. In a subsequent generation it could mutate to the x2 state, triggering other genes to mutate or single genes to mutate more. In the offspring of those which survive, the saltagen would mutate again, perhaps to a x4 state, inducing more mutations in the main sequence genes which it influences. Further mutations to x8 etc could occur as long as the gene was transmitted to a successful later generation.
What are the conditions for activation of the saltagen? The environment must have changed sharply (in geological terms) so that the species is unstable in it. It is ill-fitted to the conditions in some way. The saltatory gene determines how much mutation occurs in the genes which it controls. As long as the next generation survives, the saltagens it carries will also survive and be able to switch to higher states. Each higher state will cause larger amounts of genetic change. Eventually it will cause macromutation of the organism.
Obviously a stage will be reached when the saltagen's mutational steps and the scale of genetic change are so large that the mutations induced are all damaging. Selection or embryonic preselection will then tend to eliminate the individuals with the highest level of saltagen and the escalating process will go into reverse eventually switching off the saltatory gene when stasis is again reached.
The gene then only switches on randomly, to test the environment, so to speak. This process allows a faster approach towards the evolutionary solution. It also might go some way towards accommodating the "creative evolution" of Bergson and Shaw, generally frowned upon by orthodox evolutionists.
All very fanciful, I hear sceptics say, but where is the evidence?