In population genetics stability of ecosystems is defined via the persistence of certain key species in time (GREGORIUS 1996). The persistence of a given species is determined by the adaptability of the populations within the species concerning changing environment in time and space. Within populations as the units of adaptation the genetic system determines the adaptive potential and is object to evolution (DARLINGTON 1939, GREGORIUS 1995). Population genetics analyses the genetic structure of populations by assessing allele and genotype frequencies at certain gene loci. Here, the adaptive potential is estimated in terms of allelic diversity (HATTEMER et al. 1993). On the other hand the analysis of phenotypes of quantitative adaptive traits is governed by genetic and environmental components. In quantitative genetics it is shown that adaptation includes both levels of variation, the structural level of al-lele and genotype frequencies and the regulatory level by which phenotypic traits are varied due to environment influences. The regulatory level is also governed genetically, i.e. by regulatory genes. These genes can vary, causing varying regulatory effects on the expressed phenotype. So far, models of population genetics and of quantitative genetics do not include the regulatory process as caused by the regulatory background of structural genes. This, however is necessary if we want to understand adaptation at the structural and at the regulatory level. Therefore so far there is a lack of explanation while trying to estimate the adaptive potential in real populations (MITTON 1995). Furthermore traits closely related to fitness tend to have rather low heritabilities (HARTL and CLARK 1989) and molecular gene markers often are poor predictors of the population differentiation of quantitative adaptive traits (KARHU et al 1996). In contrast to the allelic adaptive potential which is easily determined in the case of special assessable loci, e.g. isozyme loci, by counting alleles, the regulative adaptive potential can be measured for a special trait, mostly in provenance trials, but there is no general model yet how to integrate the understanding of the dynamics of the genetic system and the role of the regulatory domain.