Self-reconfigurable robots are expected to exhibit various interesting abilities, such as adaptivity and fault tolerance. These remarkable abilities originate from the fact that their mechanical systems intrinsically possess very large degrees of freedom. This, however, causes a serious problem, i.e., controllability. To overcome this, autonomous decentralized control is expected to play a crucial role, as widely observed in living organisms. However, much is still not understood about how such decentralized control can be achieved. This is mainly because the logic connecting local behaviors to global behaviors is still not understood. In this study, we particularly focus on a very primitive living organism, slime mold (physarum polycepharum), since it is believed to employ a fully decentralized control based on coupled biochemical oscillators. We modeled a decentralized control algorithm based on coupled nonlinear oscillators and then implement this into a two-dimensional modular robot consisting of incompressible fluid (i.e., protoplasm) covered with an outer skin composed of a network of passive and realtime tunable springs. Preliminary simulation results showed that this modular robot exhibits significantly supple locomotion similar to amoeboid locomotion and that the exploitation of the "long-distant interaction" stemming from "the law of conservation of protoplasmic mass" performs some of the "computation" that the controller would otherwise have to carry out. As a consequence, adaptive amoeboid locomotion emerges without the need for any centralized control system. The results obtained are also expected to shed new light on how control and mechanical systems with large degrees of freedom should be coupled.