The theoretical underpinnings of contemporary ecological genetics are mainly borrowed from the literature on quantitative genetics and the breeding of crop plants and domesticated animals. A number of fundamental issues remain to be resolved before the mapping of quantitative genetic methods from the realm of plant and animal breeding to the less controlled world of natural populations can be accomplished. One of the most serious difficulties in applying quantitative genetic methods to natural populations is that one of the assumptions of this approach is that the members of the population under study are mating at random. This is an assumption that is unlikely to be fulfilled in most natural populations. When there are complex forms of genetic variation, such as interactions among genes at different loci, violation of the assumption of random mating may have profound effects on the potential for evolutionary changes to occur. This research will use Plantago lanceolata to examine the role of population subdivision and localized mating in influencing estimates of the genetic variation in a population, and the potential for selection to bring about evolutionary change. The basic approach will be to compare genetic variance components estimated in a population by using the resemblance among relatives when the parents are chosen based on their probability of mating in the nature with similar estimates of genetic variance components when the parents are chosen at random from the population at large. If estimates of the genetic variance components under these two mating systems are different it will indicate that the mating structure can have a strong influence on the potential for evolutionary change to occur. If they are different the estimates based on localized mating, which are expected to be larger, will give the best predictions of the response to selection.
