Tautology and Natural Selection

My view is that fitness isn't reproductive success -- someone else
said it was and I disagreed.I veiw it that survival is more important than reprodution,
but that both clearly play a part.All I am saying is that there can be no meanignful concept of
fitness without it being a function of survival (it may also
be a function of reproductive output -- at the same time --
a function of both).

quote:

---

Again I re-iterate, an engineer has an explicit a priori goal or desired outcome. Natural selection and evolution do not..so how are they comparable.

---

I think it's perfectly reasonable to talk about reproductive success as the 'goal' of natural selection, more specifically a self optimizing goal. Maybe you can say I'm falling into the "Modeler's Fallacy."

quote:

---

And natural selection is not optimizing reproductive success...how is it that when a trait becomes more frequent it is "optimized"...what is optimal?

---

The trait is only considered optimal because it confers reproductive success upon the organism that has it. Some traits might not confer any fitness (or negligable fitness) and get passed along also.

quote:

---

Not a single biological system is optimal. Reproductive success can be the difference between producing 10 offpring or 1 in a small population..is this optimized if the related group or species the next mountain range over has a population size of millions? Which one is optimized? How has natural selection optimized reproductive success? Some phenotypes are more abundant because of advantages in transmitting their heritable material from one generation to the next...not because they are the optimized reproducers....

---

I think it can only be considered optimal in relation to its direct environment, i.e its optimized in relation to the organism that produced 1 offspring. As for the organisms over the mountain range, you can't make the comparison because they are not competing.To clear things up I'm using 'optimal' to loosely. I'm thinking about optimal in terms of a a progression from 0 optimality to infinite optimality.JustinC

Would it be wrong to say the purpose of an eagles wing is to aid in flying and then to say an eagle has no purpose? Or purpose should never ever be used? Just random questions.JustinC

Hi Justin

quote:

---

I think it's perfectly reasonable to talk about reproductive success as the 'goal' of natural selection, more specifically a self optimizing goal. Maybe you can say I'm falling into the "Modeler's Fallacy."

---

Now you have switched from engineering to "goal" which compounds the problem. You don't have the environment pushing individuals to produce...those that happen to produce more offspring in a generation have higher relative fitness because of some quality that has nothing to do with refining the process of reproduction...those indiviudals and whatever trait that conferred the benefit may in the next generation be a detriment and some very rare variant may come to predominate...how is this goal oriented?..goal implies a priori knowledge i.e. pre-determined outcome just like engineering does.

quote:

---

The trait is only considered optimal because it confers reproductive success upon the organism that has it. Some traits might not confer any fitness (or negligable fitness) and get passed along also.

---

So what if there are 20 variants that are at equal frequency but are highly selected for such that all other variants but the 20 are quickly eliminated..which of the 20 is optimal?

quote:

---

I think it can only be considered optimal in relation to its direct environment, i.e its optimized in relation to the organism that produced 1 offspring. As for the organisms over the mountain range, you can't make the comparison because they are not competing.

---

actually the individual the produced 10 offpring has a higher relative fitness. I should have better clarified that there can be gene flow between the two populations though not frequent...thus, they are still competing.How would an optimal of 0 to infinity be descriptive? That is why there are terms such as relative fitness because it does not imply that something is better "designed, engineered, reaching a goal"..it just means that for some reason there are more of one type of variant than another in a given generation...over time which variant or variant of a variant can alter drastically in just about any trajectory.cheers,
M

quote:

---

My view is that fitness isn't reproductive success -- someone else
said it was and I disagreed.
I veiw it that survival is more important than reprodution,
but that both clearly play a part.

​

All I am saying is that there can be no meanignful concept of
fitness without it being a function of survival (it may also
be a function of reproductive output -- at the same time --
a function of both).

---

When you say survival do you mean 'Individual Survival'? If so I'd have to disagree. 'IS' is only pertinent in context of reproduction. Why do you think salmon swim all the way up stream to reproduce, and then die? Because there is no 'selecting' for mere survival unless it is in context of reproductive success. I guess you could say the summum bonum of the process is reproductive success.

ok..I agree...if you are never born you will never reproduce..But to claim that survival is the most important thing misses the point...I gave an example that was a snapshot because I don't know what the selective pressures (if there are any) will be on the next generation. So of course I am measuring at time X. you are treating evolution and natural selection as if it is a constant....you observe a population that exists and see what the frequency is of different variants. At that moment you are not measuring future events.Again, what is the fitness of a group of variants that live 20 times longer than all others but have no offspring?

This is pretty sweet getting responses so quick.

quote:

---

Now you have switched from engineering to "goal" which compounds the problem. You don't have the environment pushing individuals to produce...those that happen to produce more offspring in a generation have higher relative fitness because of some quality that has nothing to do with refining the process of reproduction...

---

But it must have some relation to reproduction if it causes the organsims to be reproductively successful.

quote:

---

those indiviudals and whatever trait that conferred the benefit may in the next generation be a detriment and some very rare variant may come to predominate...how is this goal oriented?..goal implies a priori knowledge i.e. pre-determined outcome just like engineering does.

---

I know 'goal' is probably a bad word to use, but I would still say that a goal could be better designed systems for reproduction.

quote:

---

So what if there are 20 variants that are at equal frequency but are highly selected for such that all other variants but the 20 are quickly eliminated..which of the 20 is optimal?

---

Equal optimality.

quote:

---

actually the individual the produced 10 offpring has a higher relative fitness. I should have better clarified that there can be gene flow between the two populations though not frequent...thus, they are still competing.

---

If it's not frequent I'd say it is probably negligable.

quote:

---

How would an optimal of 0 to infinity be descriptive? That is why there are terms such as relative fitness because it does not imply that something is better "designed, engineered, reaching a goal"..it just means that for some reason there are more of one type of variant than another in a given generation...

---

Wierd, but I think terms like relative fitness are useless. But I guess that can be because I'm thinking about 0 to infinite fitness (optimality is to confusing).What does it mean to be relatively fit? To go back to the eye examples, you'd say that 1 percent vision is fit relative to 0 percent vision. How can you compare the two systems? Don't you have to have a reference to the environment, and the better designed (adapted if you like) will survive?[This message has been edited by JustinCy, 07-30-2003]

quote:

---

Tautologies don't make predictions. Definitions don't predict. They just identify. That is all that is going on when you say that 'survivors survive' -- you are pointing out the relevant subset of creatures. Why this subset? The reason is reproduction. Dead creatures don't reproduce and are hence irrelevant to evolution.

---

Survial in 'survival of the fittest' would be synonomous to reproductive success. If it is defined by the post hoc, then you would get 'survivors survive' or 'reproductive success of the reproductive successful'. How is that supposed to explain evolution? They're just definitions, as we agree.

quote:

---

What direction?

​

If you mean that 'survivors survive' doesn't tell us which adaptations lead to higher survival rates, then of course it doesn't. You have to examine the survivors to find out. I think you are trying to drag way too much out of the first step of the analysis.

---

I'm not sure I disagree with the above. Can you tell me where you think i do?

quote:

---

Well, if it is always true, it is a very good place to start. Then you ask, "Why?" "Is there a reason some survive and others do not?"

---

It's always true based on logical necessity. Yes, the 'why' question is what natural selection tries to explain. Again, if fitness is defined by the post hoc analysis, then it would explain nothing. No more than my ground breaking theory of "people that walk, walk."

quote:

---

One hundred critters are born. You tag them and release them. A year later, you find all the tags but only 25 are attached to living animals. This is not observation?

---

You observed that some animals survived. They are the survivors, by definition. You did not observe survivors surviving. If I stare at a blank wall I can assert that survivors survive without observation, because its a logical necessity. It's no more of an observation than saying, "x=y because y=x". Is that an observation?

quote:

---

I dare you to do science without definitions. Hell, I dare you to think about anything without starting with definitions. Why are you downplaying this? Consider. We want to study light. What is light? Well, you have to define it somehow, otherwise you can't move on to further study. "Light is what illuminates the Earth during the day." "Light is what candles emit." All very circular definitions really. It is unavoidable. You seem to object to this step.

---

I think you know exactly what I mean, and it's not that definitions aren't useful for science. They are needed to communicate.What I meant is that they are not scientific theories.[This message has been edited by JustinCy, 07-30-2003]

quote:

---

But it must have some relation to reproduction if it causes the organsims to be reproductively successful.

---

Why? That is only true if you think that every trait of an organism is directly involved with the reproductive system. Do you think that flight is directly involved in gonadal development for example?

quote:

---

I know 'goal' is probably a bad word to use, but I would still say that a goal could be better designed systems for reproduction.

---

Since it is an inappropriate word to use why use it? Fitness advantage, higher relative fitness, even higher frequency describe the same thing without adding a goal directed connotation.

quote:

---

Equal optimality.

---

That is why optimal is a useless term...equal optimality does not help...so if a new group of variants are produce in the next generation some of which are slightly better adapted are they more equal optimal..or hundreds of opitmal variants...optimal fails then to describe the relative advantage.

quote:

---

If it's not frequent I'd say it is probably negligable.

---

Unfortunately genetics is against you here...very small levels of migration/immigration i.e. gene flow can homogenize populations over long distances over long periods of time...it is hardly negligible.

quote:

---

Wierd, but I think terms like relative fitness are useless. But I guess that can be because I'm thinking about 0 to infinite fitness (optimality is to confusing).
What does it mean to be relatively fit? To go back to the eye examples, you'd say that 1 percent vision is fit relative to 0 percent vision. How can you compare the two systems? Don't you have to have a reference to the environment, and the better designed (adapted if you like) will survive?

---

Relative fitness means that an individual, species, allele produces more offspring, population members, copies of itself than another individual, species, allele...that simple. The cause of this relative advantage of passing this information from one generation to the next is what evolutionary biologists are interested in.Your vision example misses the point. The vision trait itself is not fit...if those with 1% vision leave more offspring behind than those with 0%, then the 1% are relatively more fit. If they leave equal number of offspring behind then they have equal relative fitness...which over time comes to predominate in the population will be dictated by chance and not by selection...any kind of refinements to the vision system as you are interested will only come when variants appear which then have a higher relative fitness to those that already exist for example a mutation occurs in the 1% vision system that allows for a new light wavelength to be distinguished and gives those individuals an advantage in locating food. They eat more, live longer, produce more offspring and voila..a few generations down the road, that variant is predominant...and it can turn around just like that...a new pollutant is added to the atmosphere and that wavelength of light is now longer visible..individuals of the original 1% vision group don't expend energy making the protein that allowed the new variant to see the now blocked wavelength..so the original variant a few generations later it is again at high frequency...no goals, no direction, no pre-determined path, not optimized...more like a never ending lottery.Comparing species with no vision system to those with a vision system is similar in concept but the parameters are very different as you are then comparing individuals that do not reproduce with one another...and it is much more difficult to determine the relative fitness of species i.e. predator species such as wolves will almost always have a smaller population density than their prey species...if the wolves increased their number dramatically it would actually be a disadvantage long term....but this is where an ecologist like Quetzal would have better input...

I completely agree with everything here, except that I’d rephrase almost always to always. I can’t think of any case where predators have a higher population density than prey. However, I think a discussion of trophic levels, energy webs, carrying capacity, ESS, arms races etc, will get way off topic.The key point for Justin to take away here is that relative fitness is a valid tool to describe both inter and intra population/species/etc dynamics.

But if some salmon have an adaptation that makes it
more likely for them to make it back to their
breeding pool, then traits from THOSE salmon will
become the norm.The individual selection (survival oriented) has an impact
on the subsequent generations of salmon.

They are well fit , but thet don't contribute to
subsequent generations.Fitness (survival oriented) has an impact in the
allelic frequencies, so does reproductive output.So long as both are represented in a definition of
'reproductive success' I would be happy -- neglecting EITHER
misses the point.My objection is to defining fitness soley as reproductive
success, when reproductive success is understood to be, basically, reproductive output.If that is not the case, and survival is a part of the equation
I don't have a problem.Natural selection is more straightforward to understand if
one views it as cycles of reproduction followed by survival
(it's also a reasonable description of most critters life
cycles).Any living, non-sterile organism can reproduce. The impact
of survival rates for varying trait sets will (in general)
be greater than the impact of rerpoductive output unless
some members of the population have orders of magnitude
more offspring -- even then survival rates could vary enough
to ensure that those that leave the most offspring don't become
the norm (because those offspring don't all survive to
breed).

quote:

---

Any living, non-sterile organism can reproduce. The impact
of survival rates for varying trait sets will (in general)
be greater than the impact of rerpoductive output unless
some members of the population have orders of magnitude
more offspring -- even then survival rates could vary enough
to ensure that those that leave the most offspring don't become
the norm (because those offspring don't all survive to
breed).

---

Ok..I finally see where the problem is and why we are talking past each other. Measurement of fitness comes at a specific time point...you can measure trait frequency at time X and that will tell you what trait etc. is the most predominant at time X. This tells you nothing about time Y. If I have ten kids and you have none at time X, I have a higher relative fitness than you...if at time Y all of my kids die and you have one kid then you have a higher relative fitness...you could measure relative fitness over a few generations (well in fruit flies or shorter lived organisms) as well to add a temporal component...but you cannot measure events that have not happened yet...maybe humans go extinct...then we ultimately have fitness = 0. Evolution is not static....however, taking a measurement is a mere snapshot in an ongoing process of unknown trajectory.

quote:

---

You can measure trait frequency at time X and that will tell you what trait etc. is the most predominant at time X.

---

Now THAT I can agree with

quote:

---

This tells you nothing about time Y. If I have ten kids and you have none at time X, I have a higher relative fitness than you...if at time Y all of my kids die and you have one kid then you have a higher relative fitness

---

This is the bit I'm still having problems with.If an individual lives for a long time {in relation to
others of the same type) due to heritable characteristics I would consider it 'fit' in a survival sense
whether it bred or not.But I do see that it also needs to be 'fit' in a reproductive
sense.'Reproductive output' isn't 'reproductive success' I suppose,
in which case we may well be in agreement after all.Fitness = Number_of_Offspring * Probability_of_Survival_to_BreedI think 'nature' agrees with me, as those creatures whose young
are most likely to die without issue have the most young.[Added by edit:: I'm not sure but I think I might be mixing
'levels' in my view of fitness][This message has been edited by Peter, 07-31-2003]

quote:

---

Why? That is only true if you think that every trait of an organism is directly involved with the reproductive system. Do you think that flight is directly involved in gonadal development for example?

---

We are referring to traits that will cause reproductive success if selection, right? I don't think I said every trait will. I don't think its involved in gonadal development, but it will more than likely have an affect on the organisms reproductive success (atleast at one time if the trait is completely homogonized).

quote:

---

Since it is an inappropriate word to use why use it? Fitness advantage, higher relative fitness, even higher frequency describe the same thing without adding a goal directed connotation.

---

Then maybe I'll just stop using it then, huh? What about that?

quote:

---

That is why optimal is a useless term...equal optimality does not help...so if a new group of variants are produce in the next generation some of which are slightly better adapted are they more equal optimal..or hundreds of opitmal variants...optimal fails then to describe the relative advantage

---

If they are better adapted, why would I say they are 'more equal optimal (fit)? Wouldn't I just say they are more fit?

quote:

---

Unfortunately genetics is against you here...very small levels of migration/immigration i.e. gene flow can homogenize populations over long distances over long periods of time...it is hardly negligible.

---

I understand it can homogonize a population, but would you say the populations are competing with each other to a noticable degree? Wouldn't the gene flow just introduce a new allele into the population which would affect competition in that population? I'm not speaking as if I know the answer, I'm just asking.

quote:

---

Relative fitness means that an individual, species, allele produces more offspring, population members, copies of itself than another individual, species, allele...that simple. The cause of this relative advantage of passing this information from one generation to the next is what evolutionary biologists are interested in.

---

The thing I don't like about relative fitness is that it is completely ad hoc. I mean if there is a mudslide that kills a member of a population, is it less relatively fit? If the cause is a trait or traits, wouldn't you be looking for what is better designed for reproductive success?

quote:

---

Your vision example misses the point. The vision trait itself is not fit...if those with 1% vision leave more offspring behind than those with 0%, then the 1% are relatively more fit.

---

This is what I don't like about relative fitness. Is something only more fit when compared to another in the population? Shouldn't fitness be determined by the environment, and the fittest will survive?

quote:

---

If they leave equal number of offspring behind then they have equal relative fitness...which over time comes to predominate in the population will be dictated by chance and not by selection...any kind of refinements to the vision system as you are interested will only come when variants appear which then have a higher relative fitness to those that already exist for example a mutation occurs in the 1% vision system that allows for a new light wavelength to be distinguished and gives those individuals an advantage in locating food. They eat more, live longer, produce more offspring and voila..a few generations down the road, that variant is predominant...and it can turn around just like that...a new pollutant is added to the atmosphere and that wavelength of light is now longer visible..individuals of the original 1% vision group don't expend energy making the protein that allowed the new variant to see the now blocked wavelength..so the original variant a few generations later it is again at high frequency...no goals, no direction, no pre-determined path, not optimized...more like a never ending lottery.

---

I understand in the long run there is alot of chance, but aren't we talking about natural selection, which won't factor in genetic drift?

quote:

---

Comparing species with no vision system to those with a vision system is similar in concept but the parameters are very different as you are then comparing individuals that do not reproduce with one another...and it is much more difficult to determine the relative fitness of species i.e. predator species such as wolves will almost always have a smaller population density than their prey species...if the wolves increased their number dramatically it would actually be a disadvantage long term....but this is where an ecologist like Quetzal would have better input...

---

I was thinking more along the lines of the first organism to have some sort of photorecepting cells, or the first single-celled organism to have a photorecepting path of pigments. In which case I would think that they would still be able to reproduce, although I can't be a hundred percent sure.JustinC