Peez

I must differ with you here. There is only one "level" of evolution, but selection can occur (in principle) at a number of levels. Not really. First, we need to be clear abou what "natural selection" is. The generally accepted definition is that individuals with particular traits tend to have greater reproductive success than others in that population. Note that this does not automatically involve genes, in fact it is quite possible to have strong natural selection without any change in allele (gene) frequency. Natural selection acts on the phenotype (observed trait), not the genotype. IF natural selection results in a change in allele frequency, then evolution has occurred.

Darwin imagined natural selection acting on individuals in a population, and that is still how most evolutionary biologists see it. Such selection favours individuals who maximize their reproductive rate regardless of other considerations. Some have suggested that natural selection might also act on populations in a species, so-called "group selection". Such selection would favour groups that grow and spread fastest, even at the expence of individual reproductive success. It has been shown that such selection might work under very restrictive conditions, but that generally selection acting on individuals will swamp out any selection on populations. Dawkins is looking the other way, at selection on alleles in an individual. In this case alleles that maximize their replication rate will be favoured, even at the expense of the reproductive success of the individual. There is evidence that this occurs in some cases, but it does not add anything to selection on individuals in the great majority of cases. Please explain, what is a "higher level" of evolution? Remember that natural selection acts on the phenotype, not the genotype. Only the phenotype can be selected, but only if the genes have been affected does evolution occur. I agree, the whole "hopeful monster" comparison with PE is a creationist ploy (one which I just ran across in another forum, as it happens).

Peez

Peez

To all,

There seems to be a kerfuffle about the inheritance of phenotypes. Perhaps some of the confusion is over what biologists call maternal effects. In addition to genes, mothers do pass on other material goods. The zygote (fertilized egg) is almost entirely a cell from the mother (in most animals), and the things present in that cell certainly influence the traits of the individual. In addition, the environment that the cells experiences in critical early development is determined largely by the mother. These maternal effects are typically minor compared to genetic effects, though there are exceptions, but in any event are not part of evolution. Although they can sometimes persist (watered down) more than one generation, it is genes that are carried through many generations, and it is the change in gene frequencies that is evolution.

Meanwhile, natural selection can act on variation that is not genetically determined. For example, if a lucky mother happens to find a lot of food during her pregnancy, perhaps her offspring will be a little healthier. Natural selection favours the healthier individuals, but there is no genetic basis for the greater health and so there is no evolution.

I hope this helps.

Peez

Oolon Colluphid

I hope ( ) that nothing I’ve said is at variance with this .

IOW, the Environmentally Added Healthiness is not passed on. Only the stuff in the genes. Of course, genes for being healthy in that environment and for taking advantage of getting healthy opportunities will prosper, at the expense of those with less good alleles for that sort of thing, so telling how much is genes and how much is environment can be difficult.

Erm... right, Peez?

Of course. As always .

Cheers, Oolon

[ October 16, 2002: Message edited by: Oolon Colluphid ]</p>

Peez

Hi Oolon!
It is actually a bit more complicated than that. In fact the ‘non-genetic healthiness' is passed on, at least for a generation or two. The mother was, just by chance, healthier than others when she produced this offspring. The offspring has a better start, and so is healthier than average even though it has not inherited any ‘healthy' genes. We do not call this "inheritance", rather "maternal effects". Such effects might be carried over to yet another generation, but will tend to be dampened out by various environmental effects.

Thus, beyond one or two generations, it is only genes that are passed on. Certainly evolution is only concerned with genes. Of course the other point that you raise is very important: how much variation in a character is due to genetic variation and how much is due to environmental variation, but that is a whole other ball of wax.

Peez

pz

No, no, you aren't allowed to be incredulous. Just accept what I say.
I'm not talking about Lamarckian inheritance here.

Phenotypes definitely are passed on! If they weren't, you could never get selection, although it is true that you could still get evolution -- it would all be by drift, though.
Dawkins always stresses this when he's called on the narrowness of his interpretations. Unfortunately, he then goes right back to talking about genes, genes, genes, and genes.
That is one definition of evolution, yes. It's a pretty good definition. It is not a complete definition, however...I prefer "evolution is a process that results in heritable changes in a population spread over many generations". It's more open about what a heritable change is, and doesn't restrict it to just allelic changes.

You are incorrect to claim that only the genetic component makes it to the next generation. I happen to have inherited a rather complex set of proteins, cofactors, membranes, and carbohydrates from my mom, in addition to a bit of DNA. She also made a substantial investment in my assembly for about 9 months, and then spent many years refining me afterwards. Were you shortchanged?
Errm, no.

An egg cell and a skin cell and a neuron and a hepatocyte and a fibroblast all contain precisely the same genes. What makes them different is the history of interactions with the environment that have led to differences in the distribution of epigenetic factors.

And eggs come from eggs come from eggs come from eggs. At no point in their history have they ever come from a collection of naked DNA.

Uh-oh. You picked the wrong example. You might want to look up centrosomal inheritance -- it's not done genetically, but instead, an existing centrosome seems to act like a seed crystal to generate a duplicate of itself. And yes, the pattern of non-genetic information tucked into my egg did get passed on to me, and was unfolded into a rather complex multicellular creature in cooperation with the genetic information I was given. I wouldn't be here without all of it.

Also, your argument from longevity doesn't work. My point is that there has been continuity of the cytoplasm and membrane as well as DNA, and that those structures also represent a substantial amount of extremely specific information. Recall from the standard models of the origin of life that DNA is a relative johnny-come-lately: before that there was the RNA world, and before that something even murkier (autocatalytic sets, perhaps). The activity in the cytoplasm/membrane/environment has been central to the history of life far longer than the details of a relatively static data repository.
Oh, dear. Think it through.

Say I have designed a new data storage format with amazing speed and density. I hand you a small silvery sphere, and tell you the information is all there -- including instructions on how to build your own CompactSphere(tm) player.

What do you do now? Do the instructions on the CompactSphere(tm) help you in any way?

The uterus is an interesting example, by the way. Do you know where the completely autonomous instructions for the mammalian uterus are located in the genome? Do you think the egg cell is a negligible component, compared to the egg nucleus?

If that is actually Dawkins' rebuttal to the analogy, then he has also failed completely in understanding it.
Yes! All of them -- genes, cytoplasm, and environment -- change over time. The observation that environments had changed radically over the course of the history of the earth was actually the earliest scientific impetus that led to the theory of evolution.
I am rather familiar with Dawkins work, including _The Extended Phenotype_. I'm afraid that you and he are both rather blinkered on this subject, to the extent that you don't even notice your biases anymore. I prescribe a healthy dose of Lewontin. You seem fairly knowledgeable otherwise, so you might also try Oyama's _The Ontogeny of Information_, although I recommend at least some background in the concepts of developmental biology if you hope to be able to absorb that one.
Do you really think of Down syndrome as a genetically heritable disease? There are instances where it is, but how do you think the overwhelming majority of cases arise?

It's also interesting and rather revealing that you are focusing on pathologies here. Genes aren't diseases. Neither are non-genetic heritable factors, or environmental influences. It's much more enlightening to think about stuff that is more fundamental and that we take for granted, such as the fact that we have two eyes, or that we have a dorsal side. Can you identify the "gene" for bilateral symmetry, or the one that stakes out dorsal? When do they act, and how are their actions restricted, without any contribution from the environment or cytoplasm?

We're all sitting here with a happy look on our face, then, and it's all good.

I'm confused about something, though. If, as you claim, both you and he have a solid appreciation of the significance of extra-genetic contributions, what is there that I have to set straight?

Peez

Evolutionary biologists do restrict "evolution" to changes in allele frequency. Any change that does not have a genetic basis is not considered evolution. You are correct that there are "maternal effects", which are in a sense some degree of phenotype being passed on to the next generation. Note that the set of proteins, cofactors, membranes, and carbohydrates (among other things) which you get from your mother (almost entirely) are just one set. Unless you inherit the genetic code for making those proteins, etc., there are quickly diluted among trillions of cells. Not only that, but many proteins, etc., have limited lifespans, and must be replaced. The bottom line is that there are maternal effects, but virtually no proteins, etc., that a woman gets from her mother without the genes for them will be passed on to her offspring. Maternal effects fade quickly, and are weak or absent after two generations. Maternal effects are certainly not evolution as biologists understand it. I think that I understand your point, and I agree that we start out as a zygote which is almost entirely made up of componants from the mother (and this is important), but I think that you are taking this a step too far. Virtually none of the componants of your mother's cell came from her mother, except by way of DNA. Half of her DNA came from her mother, including code for proteins, etc., involved in the production of plasma membrane, ribosomes, etc. Membranes do not prelicate themselves, as far as I know. Ribosomes do not either. The genetic code carries the information for producing these structures.

Peez

Doubting Didymus

PZ, I am having a lot of trouble with many things you are saying here. You seem to need to sort out many of my biological science lecturers and textbooks, as they are at complete odds with a few of your points.

There is no mechanism for passing on a phenotype. The genotype is passed on via the DNA, but where are the additional developmental factors that make up a phenotype encoded? Please, just to set me at ease, explain what you know about the difference between phenotype and genotype. I am sure you know, but I want to hear you say it.

I had a little cringe when I read this. Unless I am missing something big, then dawkins is saying that genes + environment maketh the man, but only genetic benifits can be inherited. Environmental benifits to not pass on, so they can not evolve over time. So dawkins goes back to talking about the only thing that can be perpetually inherited.

I am fine with that definition, but I still consider DNA to be the only thing I can inherit from my parents. I had not heard of centrosome replication. The important question in my mind is whether the new centrosome can inherit features from its parent centrosome. I doubt that this is the case. Would a mutant centrosome pass its mutation on?

Are you sure they were inherited? that is, are your membranes exact copies of hers? Do mutations in your mothers cell membranes pass on to you? If not, then they can not evolve.

Cytoplasm, membranes, and the environment can not pass on features to the next generation. There may be continuity, but there is no inheritance, unless everything I have been taught on the subject is wrong.

Unless egg cell components are inheriting the features of parent egg cell components, then yes, the rest of the egg is negligible. NOT in the sense that it is unneccesary, but in the sense that any other egg cell would do the job just as well.

I sincerely hope you are not equating non heritable change with biological evolution. The environment does not evolve (unless you use the standard dictionary definition instead of the biological term).

I can hardly believe I heard that. Are you aware of how much you sound like Behe? Of course you need a complete cell for genes to work, but the point is that only genes can pass on features to the next generation. Cell membranes and the environment do not evolve over the generations, anymore than spot fires evolve from parent fires.

pz

Yes. Look up cortical inheritance in ciliates.

You really are missing the entire point, however. The issue is not phenotype/genotype (well, in part it is...), but epigenesis. I am well aware that a great many people have immense difficulty wrapping their minds around the concept -- including Sidney Brenner, the recent Nobelist (although he got better, and recanted some of his rasher comments on the topic) -- but I'm afraid it is true. The genome is not even close to being a complete description of the organism.

If you are going to resort to insult, I recommend that you avoid obviously fallacious comparisons that do nothing but reveal your total incomprehension of my position. I know Behe's work painfully well. It has absolutely no relationship to what I've been saying.

It's also not a good idea to use an insult as a way to get out of answering a simple question.

Doubting Didymus

Done, apparently this is "one of the great exceptions to the rule that genes provide the only vehicles for inheritance from one generation to another." It does not seem a very powerful evolutionary force, though it does clearly constitute an exception to the normal rule.

What does this have to do with centrosomes?

Anything specific about epigenesis you wanted to point out? Elucidate if you will.

So? With a few minor exceptions, genes are the only thing that can be inherited. Unless there are more exceptions that are of enough note to warrant rejecting the focus on genes.

The fallacy is yours, I'm afraid. You said "When do they act, and how are their actions restricted, without any contribution from the environment or cytoplasm?". This is a strawman argument. No one is saying that genes can work on their own, if you take away the cell. My point is that, while the cell is important, it does not inherit information from its parent cell.

The reference to behe was not a comparison of your respective arguments, just pointing out that you were saying that 'genes don't work if you take away the rest of the cell', which just reminded me of irreducible complexity arguments.

No insult was intended, nor was I implying that your arguments are anything like his.

Your question was: what can genes do without the rest of the cell and the environment? The answer: nothing, of course. They dry out and blow away. Happy? But its a strawman. The fact so far remains that only things that are heritable can evolve. In cells, the only heritable thing I know of is genes. Centrosomes may also be heritable, which I am open to, but I am yet to see anything that suggests that they can pass on features.

RufusAtticus

You got that right. Molecular geneticists think along lies of "one gene, one phenotype." This is of course great for identifying genes that follow that rule. However, it doesn't work when scaling up to whole genomes. There is currently an initiative to determine the function of 25,000 Arabidopsis genes by 2010 (<a href="http://www.nsf.gov/pubs/2001/nsf01162/nsf01162.html" target="_blank">Arabidopsis 2010</a>.) The problem is that no one, except maybe Rodney Mauricio, studies Arabidopsis thaliana in nature. With out the component of the ecosystem, it is impossible to figure out what most genes do.

The nature vs. nurture argument is moot. It's nature and nurture.