The post-WIMP (Windows, Icons, Menu and Pointer) user interfaces are bringing new interaction modalities and the use of new input and output devices. Many of these new interfaces are not yet mature and issues related with a clear definition of an application context and technological requirements are still under investigation. The study of the relationship of the properties of interaction devices and their influence on the performance of 3D tasks (navigation, manipulation, and selection) is an important factor to identify adequate setups for carrying out these tasks. However, in a broader context, each task can be decomposed into sub-tasks which technological demands can be a challenge, since it requires multiple interaction environments, as well as transitions between them. These transitions are relevant because they can either orient or disorient the user during the overall task execution. Thus, this thesis aims to propose a technological setup (a set of interaction devices) to integrate the advantages of different functional environments. The transitional interactions between them are explored by investigating the design properties (dimensional congruency and properties of continuity) during the 3D annotation task execution. In order to achieve such a goal, a semi-immersive environment composed of 3 functional environments was developed. That composition includes a set of devices enabling the tasks of text edition (2D), navigation and manipulation (3D). Assuming that 3D annotation forces transitions between the integrated functional spaces, an exploratory study was conducted with users to investigate the behavior of interaction during these transitions by identifying the influence of the design properties discussed in this work.