Hi Jonathan,
TB gave a very good precis of what is meant by "mutation". I'll just add a few details on how they fit within evolutionary theory.
Speciation (and in fact evolution in general) is a two-step process. In the first step, variation arises in a population. Variation = some traits or characteristics of an individual organism that reflect a difference between it and all other organisms of the same type within a local population. There are basically two ways (simplifying the thing) as to HOW this variability comes about:
1. genetic recombination that occurs during gametogenesis (the formation of egg/sperm - meiosis) through crossing over, randomization during the anaphase of meiosis I, and random selection of gametes during fertilization (there's no way to predict which egg is fertilized by which one of millions of sperm).
2. mutation, which in its essence is simply an error that arises during some portion of the DNA copying process, etc. As TB noted, most errors are fixed when they occur. Some aren't, hence mutation. These unfixed errors are the source of novel genetic material. If they occur in a sperm/egg, then the mutation is inherited.
These two processes are the direct ways variation arises in a population. Another, indirect process, is called genetic drift, and is a random statistical process by which the frequency of a particular trait or suite of traits can change within a small population simply through the vagaries and vicissitudes of life. For the purposes of simplifying as far as possible this discussion, I will only consider the first two direct sources of variation.
First off, we need to clarify the relationship between these two sources of variation. Mutation creates the novel genetic material — the multiple alleles affecting a given characteristic — within the population’s gene pool. Major mutations are generally eliminated immediately because their deleterious effects preclude them being passed on. (I’m excluding gene doubling — polyploidy - in plants here, since that’s something of a special case.) Most of the non-lethal mutations are essentially neutral in a particular environment simply because they have no immediate effect on the organism’s development or survival. After all, there’s a second copy of a fully functioning gene around to take up the slack (avoiding dominance and polymorphism questions for simplicity). In this instance normal recombination during meiosis can throw up homozygous versions of the new genes. BTW: Crossing over is able to (rarely) produce novel sequences all on its own. But novel genetic material is primarily created through mutation. Recombination and the randomizing process of meiosis MAY permit the expression of the novel material. Otherwise it just sort of hitch-hikes its way down the generations, or may possibly be eliminated from the population by chance alone. In addition, polymorphisms can be created when the recessive has an net effect even though recessive (think sickle cell anemia).
The next step in the process is natural selection. Environmental factors (known collectively as selection pressures) act on those varied characteristics that have some impact on the organism’s individual survival. (environment in this context = all biotic and abiotic factors that have an influence on the organism in it’s particular ecosystem. These factors include both non-living elements such as terrain, climate, rainfall, etc, and living elements such as other species, food resources, and even members of the organism’s own population). These selection pressures in the organism’s environment tend to favor the survival of individuals with certain characteristics, primarily by weeding out those individuals who DON’T have those characteristics. Over evolutionary timescales, these gradual changes in individual populations can create genuinely new species (there are a number of ways to shortcut the process, but that's the gist.)
Hope this answers your question.
™ Version 4.2