There are two evolutionary trends in genome organization among microbes: towards either amplification or reduction. Which evolutionary scenario overcomes depends on environmental conditions and the complexity of gene networks determining phenotypic traits such as metabolic features of cells. In this simulation study, we have shown that the habitats characterized by nutrient gradients allow spatial subdivision of evolutionary trends depending on the distance to the nutrient source. We have considered interrelations between cell motility, metabolic complexity of dominant populations and ecological features of developing communities and have shown that the distribution of local dominant ecogroups follows clear patterns in chemotaxis-on and -off cases. Chemotaxis was shown to be a factor impeding introduction of new forms and decreasing total biomass of the community. Our simulations have shown that ecological patterns of self-organization of microbial communities cause sustainable different strategies underlying antagonistic evolutionary scenarios.