|MadSci Network: Evolution|
These are really great questions, and they make me wonder what the original debate was about. Unfortunately, the concept of a 'species' as a defineable unit is a faulty one, which stems from the days when people saw the universe as inherently orderly, meaningful, and easily classified. Today, we recognize the world as a disorderly place that generally defies attempts at classification and organization. This results in such questions as the ones you pose.
In order to answer your second and third questions, we have to answer the very first question and define what a 'species' is. The definition of a species that you suggested, "Two individuals (of opposite sexes) can be said to be of the same species only if they are able to reproduce naturally, and their offspring are not sterile," really does not define a species. It tells us when two organisms are part of the same species.
It might be better to think about this on the level of populations rather than individual organisms, because it is populations which evolve, and not organisms. A wolf will never evolve into a dog, but some time in the distant past, a population of wolves was domesticated and evolved into a population of dogs.
Let's stipulate that the two breeds of dogs in the example that you gave are unable to "reproduce naturally" because one is much larger than the other. However, there are enough dog breeds of intermediate size to permit what we call 'gene flow' to occur between the large and small breeds. What I mean by 'gene flow' is the transfer of genes from one population to another, and in this example, we are treating breeds of dogs as if they were distinct populations. It is 'gene flow' between populations that keeps them from becoming distinct species.
So, we can answer the first question by incorporating the idea of gene flow into the definition of a species. Let's define a species as the set of all individuals between which there can be gene flow. The requirement of gene flow implies that reproduction is taking place and that the offspring are viable.
So, even though the Chihuaua and the Irish Wolf Hound cannot mate directly, some of the Irish Wolf Hound genes can 'flow' to an offspring of the Chihuaua through a series of matings with dogs of intermediate size, and therefore all dogs can be said to be part of the same species, and dog breeds are an example of micro-evolution but not macro-evolution.
Using the gene flow definition of a species, we can start to address your third question. When the degree of gene flow between two populations drops to "insignificant" levels, they can be said to be separate species. This is a tricky proposition however, because of the difficulty in defining what a "significant" amount of gene flow is. You might not be able to observe gene flow. For example, there are some very interesting examples of populations that live in close proximity to one another, and which do not appear to interbreed, but which are still defined as parts of a single species.
These populations are part of what is usually called a 'ring species' because the two non-interbreeding populations are connected to each other via gene flow through a number of intermediate populations. Geographically, these populations form a large loop or ring with the two non-interbreeding populations next to each other at one point in the ring.
If you lined up all of the dog breeds in the world by size, and then evenly spaced them in that order around a large circle, so that the Chihuaua (which we are defining as the smallest breed in the world for the purposes of demonstration) was standing next to the Irish Wolf Hound (which is simlarly the largest), you would see that gene flow would have to proceed all the way around the circle in order to connect these two breeds genetically. That is what happens geographically with ring species. The two populations at the non-interbreeding 'end' of the ring cannot be said to be part of separate species until enough of the intervening populations go extinct to prevent gene flow between the end populations.
When considering the genetic changes we can use to define species, the answers are unclear. Another way to say this is to say that there are many types of genetic barriers to interbreeding between species.
Many plants and some animal species come about as a result of a change in chromosome copy number that is called 'polyploidy'. Instead of having two copies of each chromosome, the new species all have four, or six, or eight, or even more. This change in chromosome copy number (-ploidy) makes it very difficult to reproduce with an organism that does not have the -ploidy number. However, the genes on those chromosomes might be the same, and in the same order, between the two species, and if you were to compare a chromosome from each species, they might seem to be identical.
On the other hand, some organisms maintain genetic barriers that basically result from a change in the overall number of pairs of chromosomes. Humans have 23 pairs of chromosomes. Chimpanzees, bonobos, gorillas and orangutans all have 24 pairs of chromosomes. Since these ape species are our closest primate relatives, we can conclude that the common ancestor of humans, chimpanzees, bonobos, gorillas and orangutans had 24 pairs chromosomes, and that two pairs of chromosomes became fused at some point during the evolution of the human branch of the ape tree. As with the problems of differing -ploidy values, differing numbers of chromosome pairs make reproduction difficult.
Like humans and the other apes, horses and donkeys had a different number of chromosomes. A mating between individuals of these two species results in a sterile mule, and there can be no gene flow between the two species.
Despite the difference in the number of pairs of chromosomes between humans and chimpanzees, we repeatedly hear reports that our two species are anywhere from 95-99% identical genetically, implying that that 1-4% difference is the genetic barrier between our species. However, the difference in chromosome pair number was probably a large part of the barrier between the species. It is hard to say what a mating between humans and the various ape species would produce, for obvious reasons, and this brings up a final point, that many of the barriers to speciation are not purely genetic, and may be behavioral.
Another discussion of the meaning of a species can be found in the MAD Sci archives, and descriptions of some real world ring species can be found here and here.
Try the links in the MadSci Library for more information on Evolution.