A context-aware ubiquitous learning environment is an authentic learning environment with personalized digital supports. While showing the potential of applying such a learning environment, researchers have also indicated the challenges of providing adaptive and dynamic support to individual students. In this paper, a decision-tree-oriented mechanism is developed for that purpose, enabling digital guidance for students to observe and classify real-world objects in the learning activities of natural science courses. To show the effectiveness of the innovative approach, a context-aware ubiquitous learning environment was implemented for the "butterfly and ecology" unit of a fourth-grade natural science course in Taiwan, and 42 students participated in the learning activity. From the surveys and interviews that were conducted to understand the functionalities of the system and the learning effectiveness for the students, the results show that the system had a positive impact on students' learning, especially on the affective domain, including participation, motivation, and interaction. (Contains 6 tables and 4 figures.)