Concluding remarks

Abstract

The variation in time and space of a natural animal population is determined by many factors. There are systematic effects caused by historical events, migration and selection. There are random effects of mutation and genetic drift. One of the problems in interpreting evidence for selection is therefore making allowance for other factors. Little or nothing will usually be known about historical events. Migration patterns and effective population sizes will be known only roughly. Making an allowance may therefore be very difficult, if not impossible. It seems inevitable that evidence for selection will tend to be somewhat ambiguous, with alternative non-selective explanations being readily available. This is particularly true with electrophoretic data, for which a usable generally accepted model of random drift is not yet available.