
In biology, a species is one of the basic units of biological classification and a taxonomic rank. A species is often defined as a group of organisms capable of interbreeding and producing fertile offspring. While in many cases this definition is adequate, more precise or differing measures are often used, such as similarity of DNA, morphology or ecological niche.
For our purposes, the second sentence is probably enough. We could get picky and point out that, say, fish, living isolated in a lake somewhere are actually NOT capable of interbreeding with other fish in another isolated lake.
Another important thing to remember about the identification of a 'species' is that this is a purely human concept used for classification and description so that scientists can talk about this or that 'species' and know which biological form they are discussing. Nothing actually happens to the group of organisms being described when human taxonomists decide that this or that group should now be regarded, not as a variety or sub-species, but as a separate species in its own right.
When Creationists claim that evolution can only occur at the 'micro' level, i.e., within a species, but not at the 'macro' level, they, to be charitable, have not grasped this essential piece of basic biology: that classification is a human concept; nature does not read the rule book and is not bound by our conventions.
Now to dispel a common misconception about speciation. Speciation is a result of evolution but it is not the purpose of it. Evolution has no purpose and no pre-determined goals. The scientific Theory of Evolution (TOE) explains speciation but it does not require it to occur and nothing ‘intends’ a new species to arise.
So why does speciation occur at all?

Eventually these difference may build up in each population so that, if ever they DO come back into contact again the two populations’ genetic make-up will be such that they physically can’t interbreed successfully to produce fertile offspring, even if they can still successfully mate. This is the case with donkeys and horses, lions and tigers and many species of plant.
In this model speciation is merely a passive, incidental result of gradual evolution. There is another model however, in which speciation is itself driven by evolutionary selections. (I’ll use European finches to illustrate this but I could equally have chosen almost anything else; insects, reptiles, plants, fish or frogs, etc.)
Consider Europe either side of the last ice age. Northern Europe, the Alps to the north of Italy and the Pyrenees between Spain and France were all heavily glaciated, driving many species south into Spain, Italy and the Balkans and effectively isolating them there with impassable ice sheets.
Now take a species of finch, adapted to live in Northern Europe with a generalised bill for eating a variety of seeds. This would have been pushed south to form several isolated populations. Each would have evolved and adapted to best use the evolving and changing plant population.
One species in, say, Spain, may have evolved slender bills for picking seeds from thistles and other wild flowers, the more successful ones passing these bills on to their offspring. The other population in, say, Italy, may have evolved stouter bills for cracking harder seeds, also passing these on to their descendants. The two populations would be diversifying according to local conditions.
Now move on to the end of the ice age when the Alps and Pyrenees became free enough of ice for the finches to return, together with their food plants, into an increasingly temperate Europe as the ice retreated:
Suppose these finches had not been isolated for long enough to make interbreeding impossible. What type of bill would their offspring inherit? They would probably inherit an intermediate bill, but an intermediate bill which was no use for either of the favoured food plants of its parents. To all intents and purposes they would be handicapped and incapable of feeding or capable of feeding only with difficulty.
These would be rapidly removed from the gene-pool. Interbreeding would be hugely wasteful as the result of all that effort would be a lost brood, or, at best, a brood of individuals with a greatly reduced chance of themselves producing offspring. Anything which favoured non-interbreeding between the two forms would now be highly favourable. Changes in plumage combined with display mating rituals, territorial and mating song, and, especially, female sex selection would all be favoured.
Gene-pool isolation would be reinforced now, not by geography but by any other means available. A process of speciation which began casually and incidentally in, and because of, isolation, would now be accelerated paradoxically by a lack of the very isolation which initiated it.
And so we have lots of different finches in Europe, each with its own plumage, song, mating rituals and food plants, many of which are actually STILL capable of interbreeding successfully, and do so in captivity, but which rarely do in the wild.
Speciation has occurred because it was in the ‘interests’ of both gene-pools to speciate. An incidental yet inevitable result of evolution and an undirected, yet highly directional, process of natural selection acting as though it were driven by the needs of genes to replicate through time.