Kronholm, Ilkka (2010) Genetics of local adaptation in Arabidopsis thaliana - seed dormancy as a case study. PhD thesis, Universität zu Köln.
Local adaptation occurs when natural selection favours different phenotypes in different locations. Here, I studied the genetics of adaptation using local adaptation for seed dormancy in Arabidopsis thaliana as a model system. I asked, is there local adaptation for seed dormancy and what environmental factors drive it? What is the genetic basis of adaptation and what is the molecular nature of adaptive changes? To answer these questions I conducted a population genetic study, comparing neutral markers, candidate genes and traits. Some QTL-mapping experiments were also performed. The results obtained indicate that there is local adaptation in seed dormancy and this is mediated by the amount of precipitation received during the summer months. Local adaptation seems to occur at a regional geographic scale. Based on genetic mapping and studies, the large effect gene DOG1 is mainly responsible for adaptation, together with several other loci with minor effects. A population genetic study of DOG1 revealed that there is a signature of local selection on DOG1. Several functional alleles of DOG1 are segregating in natural populations. Mutations that increase or decrease dormancy seem to have occurred several times independently. This likely happens because of a low migration rate, new mutations occur in separate populations but they cannot migrate efficiently to other populations and thus no single mutation becomes fixed. The molecular basis of adaptive changes could not be determined, yet some candidate mutations for functional changes were identified. In addition, some of the results raised concerns about the proper way to estimate genetic differentiation. Therefore, the statistical properties of some estimators of genetic differentiation were studied using computer simulations. An estimator that takes mutation model into account can be used to compare different types of markers.
Actions (login required)