In the contexts of food or beverage production, biotechnology, and human health, microbial communities provide human societies with many types of benefits called microbial community functions. Although researchers have tried to optimize these microbial community functions, ecological and/or evolutionary dynamics can drive the communities away from the states where community functions are maximized. In this thesis, I theoretically investigate two design frameworks to optimize microbial community functions. One approach is to control the microbial dynamics by fluctuating environmental conditions (Chapter 2), and the other approach is to introduce hierarchical spatial networks (Chapter 4) so that we can restrict species interactions. I show algorithms to reveal the optimal control of the environmental conditions and the optimal allocations of microbes into a given hierarchical spatial structure, respectively. In addition, the environmental fluctuations and the hierarchical spatial structures can affect the fundamental aspects of ecology. When the environmental conditions become harsh, the intensity of demographic noise increases, which affects species diversity (Chapter 3). In a hierarchical spatial structure, stability of a downstream community depends on upstream species because the upstream communities can change the downstream environments by secreting or absorbing chemical compounds that flow downstream. In this case, positive interactions from upstream species to downstream species increase the stability of downstream communities (Chapter 5). These studies show the importance of environmental fluctuations and spatial structures in applied and fundamental microbial ecology and evolution.