All species are transitional

We're using different definitions of arbitrary. When I used the term arbitrary to describe the choice of criteria by which species boundaries are drawn, I did not mean random or capricious. I meant that the criteria are a matter of opinion and are open to debate. No matter how well a given dividing point can be substantiated with reason and evidence, it is still arbitrary because reason and evidence can be offered for other dividing points. It is the rare circumstance when criteria are completely objective. Your preference for "totally isolated" as the criteria is your opinion (one shared, by the way, by almost no one in a position of influence within biology).I think you were assuming that if we chose "totally isolated" as our criteria that it would greatly reduce the ambiguity of classification of organisms into species categories, but you'd be wrong. This isn't the main point so I won't dwell on why this is so, but we can spend some time on it if you like. Whether or not any of the analogies work for you, species change is a continuum. This must be so because the unit of change can be no larger than the difference between parent and child, which is very tiny. For this reason, all species change occurs in very, very tiny steps that only accumulate into large changes over the course of many generations.--Percy

Criteria? Suppose we decide to divide life forms up according to color. So we would have red lobsters being the same species as red foxes, and brown bears being the same species as brown wasps. Now that would be arbitrary.So there's a sense in which either definition--either morphological change or gene pool isolation--is not arbitrary. Both relate directly to the causes or effects of speciation. Morphological change causes isolation and gene pool isolation leads to more differences.It's not about criteria; it's about which generation or individual we are going to pick--say, out of 40 fossils along an evolutionary branch--to represent a different species. If we use the morphological approach, we have arbitrariness as regards which we are going to pick. If we use the gene pool isolation approach, we have definite moments of isolation--which could be dated if we had a time machine--as illustrated in our example. If it can be dated it's not seamless, and we have what creationists would call a "speciation event" (the death of the last slightly speckled Eutherian).Therefore, in order for the theory of seamlessness to work, we must use the morphological definition of "species."This message has been edited by robinrohan, 10-27-2005 11:51 PM

But what if the original population of non-speckled was still able to produce slightly-speckled. As soon as this slightly-speckled offspring appears again the heavily-speckled population is no longer a species any more. Would creationists call this an "un-speciation event"?As NosyNed would say "a bit more odd"As Razd would say "Line ... no line ... line ... "As Schrodinger's cat would say "If I eat this one does the species exist again or not"

I've been thinking about RAZD's finches, and I believe we can say that that's a case of mistaken identity. What we thought were two different species turns out to be one. But "on-and-off" is not the same thing as "seamless." However, if we use the morphological definition, it doesn't matter if we call them variants of the same species or different species. I'm sure these finches are very similar.

Agreed, "on and off" is not seemless but "now it is a species now it's not" still seems a bit odd.The point about the finches, I assume, is that, they are defined as different species because they do not mate rather then they can not mate.
It would be very inconvenient to change species boundaries every time that we see 2 species mate that do not normally mate, and back again when they stop mating again when conditions stabilise.

I would think the definition ought to be that they cannot mate. If it's a mere matter of them not mating, then 2 groups of black bears, one in Vermont and one in Oregon, would be different species.

That is from Vermont Fish & WildlifeWhat you describe is actually very close to what biologists find. For example, the Georgia Department of Natural Resources is currently running a gentic study of Red Drum. Tagging studies have shown that the Red Drum population is local. They have a specific territory and stay within that territory. The genetic results are expected to be available in 2009 or 2010 and it will be interesting to see if infact, there are genetic differences between the different populations of Red Drum.

---

Aslan is not a Tame Lion

Mutations are required if the change in pigmentation is to be heritable. But those mutations don't have to change the genetic basis of the pigmentation system. The mutations can just be the bog-standard SNPs that exist in any population, we're not talking about major mutations or the origin of new enzymes or anything like that.Let me try to put it a different way.You have a system that lays pigment out on the surface of an animal. This system probably involves a large number of components, many different hormones, hormone receptors, enzymes, and these are all linked together in a number of feedback mechanisms that ensures the pigment is laid out in some specified manner consistently, across generations. Many of these components are not specific to pigmentation, they may be things like growth hormone that are involved in many different processes at different points in the life cycle of the animal. The system as a whole also depends on things like temperature and dietary intake of the animal (so for example if you don't eat carrots you might not be able to make orange spots, they might come out brown instead).What I want to suggest is that once you have a generic pigmentation system operating in an animal, you can get a huge variety of pigmentation patterns without adding any new parts to the pigmentation system, without evolving new components or developing new enzymes or anything like that; the mutations we're talking about are just the standard polymorphisms that exist in a population, single nucleotide changes etc.Now consider how a human being would design such a system, for laying out speckled patterns on a surface. I've used a real example of a human-designed system "photoshop", I just downloaded the demo. Photoshop contains an algorithm that "lays out pigment" on the computer screen in the shape of pseudo-random vertical stripes (It's found in the Filters>Render>Fibers menu). The algorithm takes two input parameters ("variance" and "strength") and generates the pattern for you. Here's some examples of the patterns you can get from this single algorithm by varying the parameters (variance increases from left to right, strength from top to bottom). It's clear that this single algorithm is capable of producing anything from clouds (top left) to blotches (top right), fine even stripes (bottom left) to speckles (bottom right). It's just one algorithm.How does this correspond to the biological system for pigment layout. I would say that the algorithm corresponds to the genetic basis of the trait - the collection of enzymes, hormones, pigments etc. and their interactions that underly pigment layout on the surface of the animal.The input parameters correspond to any pertinent variable involved in the pigment layout system, ranging from environmental conditions (i.e. diet quality) to properties of the system itself (i.e. solubility of an enzyme, decay rate of a hormone, reaction rate of an enzyme and its substrate, the time required to create one pigment molecule, etc).Once you have the generic "algorithm" in place, then you can get all sorts of patterns, (clouds, speckles, stripes) just by varying the parameters. Very slight, subtle changes in enzyme solubility or hormone half life can tweak the pattern in quite an extreme manner without necessitating any change to the genetic basis of the trait. The genetic basis of the trait (i.e. the algorithm) can be static, but can still generate a variety of patterns.Well, this is all hypothetical. Are there any examples of this in nature? Here are some photos of conus shells. The kind of mechanism I've proposed above seems quite a parsimonious explanation of this variability. It seems unlikely to me that major mutations (changing the genetic basis of the trait) were involved in each an every speciation event. It seems more likely that the pattern is tweaked by natural selection acting on small subtle mutations that exist naturally in populations.

Good stuff, Mick. I think I even understand it to some extent. Would one of those "algorithms" you are talking about be responsible for the variety of coat patterns in a litter of puppies?

But how, in reality, do you distinguish this - by taking every species that there is and force them to mate?

You observe them. If they are adjacent to each other geographically, and they don't mate for a long time, then you assume they can't. If later on they start mating, like those finches, then you just say,"We thought they were different species; come to find out they're not."This is if you want to use the "gene pool isolation" definition. But like I said, there are problems with that definition: not arbitrary enough. It would be better to classify them using some arbitrary system about how different they are physically. If we did that, we wouldn't have this problem with the finches.This message has been edited by robinrohan, 10-28-2005 01:18 PM

Certainly. Remember that the algorithm is just the genetic basis of the coloration pattern - comprising enzymes, hormones, helper proteins, blah blah blah, and their interactions.I think it's interesting that in cats there is a "colour density" gene (and this is also present in rabbits, known as the "colour dilution" gene, and you can do a google search for it) which corresponds exactly to "saturation" or "strength" in image editing software. The existence of a "colour density" gene does imply, in my opinion, that there is a processing system encompassed in the genetic basis of the trait. Because "dilution" is not a concept that is embodied in genetic material.So there is this complex system, which has emergent properties like "saturation" that are based in the physiological properties of proteins (their solubility or whatever).But what I would emphasize is that massive changes to pigmentation (or any other trait) don't require massive changes to the genetic basis of that trait. An albino, for example, does not use a different genetic system to determine its colour than does a normal animal. And an albino can give birth to normal-coloured animals, showing that the genetic basis is retained.In cats and rabbits you also get coloration determined by temperature, which just shows that the pigmentation system has a level of complexity greater than that embodied in the genetic material. (Of course that is not to say the system as a whole is not the result of natural selection. The complexity and lability of the system is enabled by the genetic basis of the trait, that's all I'm saying).Here are some links on the genetics of rabbit coloration, cat coloration, dog coloration.mick

That, I think, answers my question. I really appreciate your thorough comments on the subject. I didn't pick it all up, but I think I got the gist.

No we would not. That is the whole point of this recent part of the discussion.As has been noted we use multiple definitions of species. E.O. Wilson (Diversity of Life) makes it clear that no definition is totally satisfying. We are trying to attach category labels to things which refuse to always fit into nice, neat niches.Generally, the biological species definition based on gene pool isolation works fine. Obviously, a whale and an elephant are in isolated gene pools at any one moment in time. It works fine. However, as we've seen there are many examples where speciation is happening but is only "complete" to varying degrees.When we lay thing out in time as well, that blurring becomes the norm rather than the exception.Species (one of the mutiple definition) works well enough to allow it to be used as part of a method of managing the information about life. To fuss about it at the most detailed level is a waste of time. There is NO "perfect" definition. Biologists have been trying for one for decades. It ain't gonna happen.This message has been edited by NosyNed, 10-28-2005 06:09 PM

Notice that I highlighted in red the part where I said that by arbitrary I did not mean random or capricious. I said it was a matter of opinion, explaining that different reasoned and evidence-based arguments could be advanced for different dividing lines between species. I don't think anyone would ever seriously suggest color as the criteria, and it clearly fits under the heading of random and capricious. No one is saying that either one is arbitrary. The point being made is that whatever criteria you develop, whether based upon morphology or genetics or a combination or something else, they are open to discussion and debate. It isn't possible to develop any hard and fast objective criteria that would remove ambiguity. That's why I asked the rhetorical questions at the beginning. When do foothills become mountains is the same type of question as when do two populations become sufficiently different to be considered different species. How much different is enough different? We can get into more detail if you don't believe such questions do not have firm objective answers.And the whole reason for this is that species change occurs in tiny immeasurable steps. It simply isn't possible to point to a step and say, "Before this step it was X, after this step it was Y."--Percy