Register | Sign In


Understanding through Discussion


EvC Forum active members: 65 (9162 total)
7 online now:
Newest Member: popoi
Post Volume: Total: 915,815 Year: 3,072/9,624 Month: 917/1,588 Week: 100/223 Day: 11/17 Hour: 0/0


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   How many generations does speciation take?
custard
Inactive Member


Message 1 of 52 (188812)
02-26-2005 9:52 PM


I'm not sure how to phrase this question, so I'll just come out and ask it.
Assuming the following:
1- Common decent from a single species
2- Rate/frequency of mutation is constant for all species
3- No evolutionary stasis/stagnation - i.e. a species does not remain unchanged for 60 million years
4- No cataclysmic events wiping out significant numbers of species.
To draw a clear, unequivocal line between species I want to use the following, restrictive definition of speciation (from talk origins), which as Pink Sasquatch has pointed out is NOT the currently accepted version of BSC:
"... that stage of evolutionary progress at which the once actually or potentially interbreeding array of forms becomes segregated into two or more separate arrays which are physiologically incapable of interbreeding." (Dobzhansky 1937)
It is important to note that this is a highly restrictive definition of species. It emphasizes experimental approaches and ignores what goes on in nature.
How many generations of offspring would have to be produced to achieve a similar level of speciation we see today - say 50 million species- from a single (Alpha) species over 3.5 billion years?
This question arises as a result of several things I have read recently: Jay Gould's Structure of Evolutionary Theory, some stuff on neo-darwinism, Haldane's dilemma, and from some things I've seen in ongoing threads that keep using analogies such as "imagine evolution is a walk down the street from house A to house G and the fossil record is a series of snapshots..." implying a gradual rate of mutation that leads to eventual speciation.
I realize this question might seem very open ended, but I'm just looking for a hypothetical mathmatical model that I can use as a starting point to help me understand how frequently mutations need to occur, and how frequently speciation needs to occur, for a Common Decent model of evolution to work.
I couldn't find anything this specific, in this forum, but if this has been explored in other threads, or if you know of links that would be helpful, please point the way.
Thanks.
This message has been edited by custard to actually add the definition of speciation I keep referring to , 02-26-2005 21:57 AM
This message has been edited by custard, 03-03-2005 17:13 AM

Replies to this message:
 Message 3 by pink sasquatch, posted 02-27-2005 5:14 PM custard has replied
 Message 33 by Brad McFall, posted 03-03-2005 5:06 PM custard has replied

  
AdminJar
Inactive Member


Message 2 of 52 (188954)
02-27-2005 3:22 PM


Thread moved here from the Proposed New Topics forum.

  
pink sasquatch
Member (Idle past 6022 days)
Posts: 1567
Joined: 06-10-2004


(1)
Message 3 of 52 (188983)
02-27-2005 5:14 PM
Reply to: Message 1 by custard
02-26-2005 9:52 PM


another, difficult assumption needed...
Hey custard,
Even if you make the assumption that rate of mutation is constant for all species, there is still another variable - speciation can occur with a single mutation, or may require countless mutations. Environment also plays a large role in speciation events. In other words, speciation cannot be distilled down to a simple equation of 'x' mutations = 1 speciation.
Perhaps someone else might have an idea about approximating an 'average' number of mutations per speciation, but it would likely be pure speculation, unless perhaps it is limited to a single species group.

This message is a reply to:
 Message 1 by custard, posted 02-26-2005 9:52 PM custard has replied

Replies to this message:
 Message 4 by custard, posted 02-27-2005 6:17 PM pink sasquatch has not replied

  
custard
Inactive Member


Message 4 of 52 (188993)
02-27-2005 6:17 PM
Reply to: Message 3 by pink sasquatch
02-27-2005 5:14 PM


Re: another, difficult assumption needed...
Great point. And let me restate that this exercise is just speculation - I'm not trying to 'prove' anything; just looking at Haldane's dilemma got me thinking and I thought playing around with a similar equation would help me better understand some things (certainly the complexity involved) about evolution from a more mathmatical point of view than I am used to seeing.
I was hoping someone could help devise some of these variables, as the one you pointed out, or to point to similar work others had done, in order to come up with some values that would fall within parameters that seem likely (yeah, subjective I know).
I was looking at some of the Haldane stuff where he calculates that it takes approx 300 generations for new genes to be 'fixed' in the population.
I'm sure this has been discussed here, and I was hoping for some help coming up with figures, or a range of figures, that people on this forum would feel comfortable with and see how the numbers play out.
The point of this is really just a sort of hands on learning exercise for me; and I thought this might be the best place to do it because perhaps some others will calloborate, or at least offer insight and criticism.
So back to your point, perhaps the 'frequency rate' I'm looking for is a combination between the number of generations it would take for new genes/mutations to 'stick' in the population AND the number of mutations necessary to create BCS type speciation (org A can't produce viable offspring with org B).
As you said, this could theoretically occur in a single generation, but is there a mean number/rate we could find acceptable that we could use in this equation?
Maybe it would be easier to stick with just one order, like Coleoptera (beetles). It's sufficiently large, I think, to get a sense of scale of the time involved.
Also, I'm not adverse to working backwards to develop some of these variables. Say using the oldest known Coleoptera fossil as a starting point for the timeline, and using 2005 as the end point.
Then extrapolating a 'mutation rate' and see how that compares to other orders?
This message has been edited by custard, 02-27-2005 18:31 AM

This message is a reply to:
 Message 3 by pink sasquatch, posted 02-27-2005 5:14 PM pink sasquatch has not replied

Replies to this message:
 Message 5 by NosyNed, posted 02-27-2005 6:26 PM custard has replied

  
NosyNed
Member
Posts: 8996
From: Canada
Joined: 04-04-2003


Message 5 of 52 (188995)
02-27-2005 6:26 PM
Reply to: Message 4 by custard
02-27-2005 6:17 PM


Speciation events
So back to your point, perhaps the 'frequency rate' I'm looking for is a combination between the number of generations it would take for new genes/mutations to 'stick' in the population AND the number of mutations necessary to create BCS type speciation (org A can't produce viable offspring with org B).
It is clearly the case that you can not assign any such numbers. A speciation event can be "instantaneous", exceedingly gradual or anywhere in between. It can be a single mutation or whole bunch of them. It can be due to the possibility of a viable hybrid at the genetic level, the fact that one population doesn't happen to like the "looks" of another (including different bird songs) or who knows what factor.
Another thing that might be true (but I'm guessing again) is the you don't have a "population" until you have a speciation event. Until that happens you have gene flow all over.
I was looking at some of the Haldane stuff where he calculates that it takes approx 300 generations for new genes to be 'fixed' in the population.
I don't know pop. genetics or the math here but that strikes me as a nonsense number right off the top. It would seem to me that the time taken for fixation would depend on too many things for there to be any such single number. The population size is one obvious one, the degree that a change is selected for (how beneficial is it), how it relates to other genes and so on. It stikes me that no such single number would say anything meaningful at all. Perhaps a geneticist can help here.

This message is a reply to:
 Message 4 by custard, posted 02-27-2005 6:17 PM custard has replied

Replies to this message:
 Message 6 by custard, posted 02-27-2005 6:49 PM NosyNed has replied

  
custard
Inactive Member


Message 6 of 52 (189001)
02-27-2005 6:49 PM
Reply to: Message 5 by NosyNed
02-27-2005 6:26 PM


Re: Speciation events
ned writes:
I don't know pop. genetics or the math here but that strikes me as a nonsense number right off the top. It would seem to me that the time taken for fixation would depend on too many things for there to be any such single number. The population size is one obvious one, the degree that a change is selected for (how beneficial is it), how it relates to other genes and so on. It stikes me that no such single number would say anything meaningful at all. Perhaps a geneticist can help here.
Good points. And I'm certainly not trying to devise some sort of 'golden ratio' that I will apply to evolution and say "see! It does/doesn't work!"
I'm just trying to learn more about pop genetics and achieve a better understanding of what really needs to happen to achieve a level of biodiversity similar to what we see on earth today; because the explanation "evolution = mutation + NS + time" sounds great, but it is almost to the point of being a platitude.
Trying to comprehend exactly what COULD occur or what would HAVE to occur over 4 billion years without really digging into it, for me, is like trying to comprehend just 'how much' one million trillion dollars will buy. Sure it will buy a lot, but 'a lot' is still an abstraction at that point.
ned writes:
Another thing that might be true (but I'm guessing again) is the you don't have a "population" until you have a speciation event. Until that happens you have gene flow all over.
Great point. Obviously there is going to be a lot of guessing involved here. That's why I am soliciting your, and other folks' help, in making or reviewing some of my guesses.

This message is a reply to:
 Message 5 by NosyNed, posted 02-27-2005 6:26 PM NosyNed has replied

Replies to this message:
 Message 7 by NosyNed, posted 02-27-2005 7:01 PM custard has replied

  
NosyNed
Member
Posts: 8996
From: Canada
Joined: 04-04-2003


Message 7 of 52 (189004)
02-27-2005 7:01 PM
Reply to: Message 6 by custard
02-27-2005 6:49 PM


Some numbers.
Well, you should be able to do some playing with numbers yourself.
One thing that we might want to know is just how many "tries" at a different phenotype have there been.
For starters, how many different individual multicellular (to keep the number only astronomical) organisms have lived on the planet in the last 600 million years?
How many were born (but many were a bad try and died immediately or were still born)?
Each of these is a more input material for selection to work on? How many have there been?

This message is a reply to:
 Message 6 by custard, posted 02-27-2005 6:49 PM custard has replied

Replies to this message:
 Message 8 by custard, posted 02-27-2005 7:40 PM NosyNed has not replied

  
custard
Inactive Member


Message 8 of 52 (189011)
02-27-2005 7:40 PM
Reply to: Message 7 by NosyNed
02-27-2005 7:01 PM


Re: Some numbers.
For starters, how many different individual multicellular (to keep the number only astronomical) organisms have lived on the planet in the last 600 million years?
Sure thing. You want fries with that?
Let me try this with beetles first. I'll post what I come up with and y'all can shoot it full of holes.
**side note: You know, if anything, this thread should be some indication to creationists of how much scrutiny scientists undergo when they present something to their peers. This is just an online forum, yet no one here is about to let me get away with anything - which is hardly a WHIFF of the type of scrutiny an actual scientist would get from his peers.**
This message has been edited by custard, 02-27-2005 19:40 AM

This message is a reply to:
 Message 7 by NosyNed, posted 02-27-2005 7:01 PM NosyNed has not replied

Replies to this message:
 Message 9 by RAZD, posted 02-27-2005 8:29 PM custard has replied

  
RAZD
Member (Idle past 1405 days)
Posts: 20714
From: the other end of the sidewalk
Joined: 03-14-2004


Message 9 of 52 (189027)
02-27-2005 8:29 PM
Reply to: Message 8 by custard
02-27-2005 7:40 PM


Re: Some numbers.
what you need is some kind of control on the numbers.
if you can take a group where all (Z) species are known for the last (X) years and during that time there have been (Y) speciation events
then you can say for this group you get Y/X/Z speciation events per year per species
you could also graph it against time and work out a standard distribution of events over time
but if you run the same parameters on another group you will get different results
and this doesn't even address if the species are in momentary stasis or a period of rapid adaptation.
I wish you well, but I just see way too many variables to have any kind of meaningful answer.

we are limited in our ability to understand
by our ability to understand
RebelAAmerican.Zen[Deist
{{{Buddha walks off laughing with joy}}}

This message is a reply to:
 Message 8 by custard, posted 02-27-2005 7:40 PM custard has replied

Replies to this message:
 Message 10 by custard, posted 02-27-2005 9:46 PM RAZD has replied

  
custard
Inactive Member


Message 10 of 52 (189038)
02-27-2005 9:46 PM
Reply to: Message 9 by RAZD
02-27-2005 8:29 PM


Re: Some numbers.
RAZD writes:
if you can take a group where all (Z) species are known for the last (X) years and during that time there have been (Y) speciation events
then you can say for this group you get Y/X/Z speciation events per year per species
OK, here's what I came up with so far:
Using my Coleoptera example I found that the earliest known Coleoptera fossil was about 265 million years old.
Currently there are at least 350,000 species of Coleoptera.
Using the fecundity rate of the western corn rootworm beetle (for no other reason than it is convenient), the fecundity rate of my Coleoptera are, on average, 1 new generation per year.
Not accounting for extinction at this point the average number of new species that HAS to occur per year to reach 350K species is:
0.001320755 or ~ 1 new species every 757 years.
350,000 species/ 265 mil years
Now that I write this I already see two major problems:
1-Obviously there have been more than 350,000 Coleoptera species in the last 265 mil years, so to estimate the total number of Coleoptera I need to use a species extinction rate. I've decided to go with .01% extinctions/century as that falls within the estimated range of natural extinction rates .001% - 1% per century.
2- And this is where I am really blowing it I think, using my 1 generation/year fecundity rate after achieving my first speciation event 757 years in, won't I be dealing with two populations so now my fecundity rate is really 1 generation/year/species? Does that mean that I have a fecundity rate of 2 generations per year (one from pop A and one from pop A1)?
Freaking exponentials.
AS for RAZD's equation X = 265 million years. Z= 350,000. But I'm not sure what you mean be "speciation event." What is the difference between Z and Y?

This message is a reply to:
 Message 9 by RAZD, posted 02-27-2005 8:29 PM RAZD has replied

Replies to this message:
 Message 11 by NosyNed, posted 02-27-2005 10:20 PM custard has not replied
 Message 12 by RAZD, posted 02-28-2005 7:25 AM custard has not replied
 Message 13 by mick, posted 03-02-2005 4:09 PM custard has replied

  
NosyNed
Member
Posts: 8996
From: Canada
Joined: 04-04-2003


Message 11 of 52 (189046)
02-27-2005 10:20 PM
Reply to: Message 10 by custard
02-27-2005 9:46 PM


It's a start.
But you're right there is some more tinkering that needs to be done.
Clearly there have been one hell of a lot more the 350k species. An extra input assumption you need is the average duration of a species. This could be a bit tricky even to define over time since one will frequently blend into another. But why not pick a number as a starting point. Call it 1 million years (seems high for beetles but I've read the 5,000,000 is the average for larger animals).
Remember that life is a bush of species. So once you have a new twig (species) you get a new opportunity for a branch. I think that means that if you get to 10,000 species you want to ask how many of them wil l produce a speciation event per year.
I would say that the rate of a new species arising from one species every 757 years seems high (though critters like these seem to be able to speciate in decades) However, as noted above, we aren't dealing with one species once we get going.
Let's look from the current diversity forward , how many new species may arise from 350 K species in a set number of years? How many have to speciate to maintain diversity if we all 350K of them will be gone in a million years? We need 3 new species a year don't we. But this one has to arise out of 350,000 species.
Continueing to think out loud: That suggests a speciation event from one in 100,000 species each year will do it. That seems to work out to an event from 1 species each 100 kyr period which is plenty long enough right?
This is fun! But it is going to take a lot of reworking to arrive at anything which makes any sense at all.

This message is a reply to:
 Message 10 by custard, posted 02-27-2005 9:46 PM custard has not replied

  
RAZD
Member (Idle past 1405 days)
Posts: 20714
From: the other end of the sidewalk
Joined: 03-14-2004


Message 12 of 52 (189123)
02-28-2005 7:25 AM
Reply to: Message 10 by custard
02-27-2005 9:46 PM


Re: Some numbers.
Y is new species, Z is number of existing species, so Y/Z is ratio of new species.
I think you need a smaller first box to sort out the variables.
take something in the last 40 years where the parameters are known?

we are limited in our ability to understand
by our ability to understand
RebelAAmerican.Zen[Deist
{{{Buddha walks off laughing with joy}}}

This message is a reply to:
 Message 10 by custard, posted 02-27-2005 9:46 PM custard has not replied

  
mick
Member (Idle past 4986 days)
Posts: 913
Joined: 02-17-2005


Message 13 of 52 (189683)
03-02-2005 4:09 PM
Reply to: Message 10 by custard
02-27-2005 9:46 PM


Re: Some numbers.
Hi custard,
This is just a quick thought. You may be looking at the problem the wrong way round - trying to forward-simulate the evolutionary process using assumed fecundity to reach a desired level of speciosity. An alternative way of doing it is to a) use observed speciosity; b) use an inferred timescale for the evolutionary process using some form of the molecular clock; c) from this data calculate an inferred speciation and extinction rate. Along with a and b, you can also use an observed extinction rate if the species you're interested in leaves fossils behind it. The method is described in "inferring speciation rates from phylogenies" by Nee (http://homepages.ed.ac.uk/snee/evolution.pdf)
Regards,
mick

This message is a reply to:
 Message 10 by custard, posted 02-27-2005 9:46 PM custard has replied

Replies to this message:
 Message 14 by custard, posted 03-03-2005 2:09 PM mick has replied

  
custard
Inactive Member


Message 14 of 52 (189801)
03-03-2005 2:09 PM
Reply to: Message 13 by mick
03-02-2005 4:09 PM


Re: Some numbers.
Awesome. Thanks for the suggestion and the link Mick. I'll take a look.
You are absolutely right that some of the problems with my initial approach are that I've backed myself into a model where I'm using too many hypothetical variables.
RAZD suggested using a more restrictive model, say something that has been observed over 40 years, but to the best of my knowledge, there have been no observed speciation events that meet the criteria of the BSC (biological species concept):
quote:
"... that stage of evolutionary progress at which the once actually or potentially interbreeding array of forms becomes segregated into two or more separate arrays which are physiologically incapable of interbreeding."
And, frankly, I'm starting to wonder where the evidence for this type of speciation can be found. Looking at wolf-like canids alone I have to ask WHY haven't we seen this type of speciation (BSC) despite evidence that humans have segregated canid sub-species populations and bred them selectively for what, 10,000 years or so?
Yet wolves, coyotes, and pugs can still interbreed and produce viable offspring. The typical evo argument "well enough time hasn't passed" is starting to ring hollow. How much time is 'enough time?'
What is the minimum number of generations necessary for BSC speciation? 100,000? 10 million?
It's not that I doubt evolution occurs, we certainly went from populations of single celled organisms to our current state of biodiversity, but I'm really beginning to question HOW it occurs; specifically, the common explanations of how it occurs which I encounter in this forum daily. These explanations are beginning to seem simplistic to me and smack of ignorance of the actual process, and, yes, even 'faith' in a process that isn't nearly as well understood as one would think reading some of these threads.

This message is a reply to:
 Message 13 by mick, posted 03-02-2005 4:09 PM mick has replied

Replies to this message:
 Message 15 by Loudmouth, posted 03-03-2005 2:32 PM custard has replied
 Message 16 by pink sasquatch, posted 03-03-2005 2:59 PM custard has replied
 Message 47 by mick, posted 03-09-2005 9:18 PM custard has not replied

  
Loudmouth
Inactive Member


Message 15 of 52 (189808)
03-03-2005 2:32 PM
Reply to: Message 14 by custard
03-03-2005 2:09 PM


Re: Some numbers.
quote:
It's not that I doubt evolution occurs, we certainly went from populations of single celled organisms to our current state of biodiversity, but I'm really beginning to question HOW it occurs; specifically, the common explanations of how it occurs which I encounter in this forum daily. These explanations are beginning to seem simplistic to me and smack of ignorance of the actual process, and, yes, even 'faith' in a process that isn't nearly as well understood as one would think reading some of these threads.
The conditions for each instance of speciation is different. Using a really poor analogy that probably only makes sense to me, speciation is an analog process instead of a digital process (I bet McFall get's this one though). Perhaps meteorology is a better analogy. We know the forces that go into cloud formation, hurricanes, etc. However, we can't predict long term patterns with any specificity because the forces in motion are not amenable to modeling. It comes down to the randomness of mutations, the randomness of environmental changes, and the inherent chaos that our universe is in. Explanations are simplistic because the specifics are always different. The effects of speciation are consistent, but the causes of speciation are always complicated and inconsistent.
All of this came to me when I was trying to think of an answer for the OP. I think it is impossible to calculate speciation rates with any accuracy. Any model hits the old wall of "the map is not the territory". Every model is going to be insufficient, at least in my opinion. I think it is a topic worthy of consideration and debate, but in the end I think it will be a Fool's Errand. And as we all know, I am always right.

This message is a reply to:
 Message 14 by custard, posted 03-03-2005 2:09 PM custard has replied

Replies to this message:
 Message 17 by custard, posted 03-03-2005 3:13 PM Loudmouth has replied
 Message 51 by jatin hira, posted 10-05-2012 10:01 AM Loudmouth has not replied
 Message 52 by Kusum Yadav, posted 12-10-2012 11:18 PM Loudmouth has not replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024