Representations such as drawings, graphs, and computer simulations, are central to learning and doing science. Furthermore, ongoing success in science learning requires students to build on the representations and associated practices that they are presumed to have learned throughout their schooling career. Without these practices, students have difficulty learning new representational forms and new scientific concepts that rely on specific representations. Students' representational practices are also heavily influenced by both their prior experiences with representing (Sherin, 2000) and the context in which they are creating or using a representation. To better understand how students of all ages can come to effectively use representations in science class, it therefore behooves us to first understand how students early in their school careers engage with representations, and how this is influenced by the activity context in which are engaging with the representation. This dissertation reports on the results of a 10-week design experiment that was implemented with two mixed-age classrooms of kindergarten and first grade students (N = 42). Students were videotaped engaging in creation of science representations in three different contexts--individual creation of representations such as drawings, participatory modeling, and participatory simulation. The concept of the primary interactive pathway is introduced in order to compare students' talk about their representations across these three contexts. Results show that the activity context influenced whether students discussed what to represent, how to represent it, and why their representations should be created in a particular way. In order to examine students' representations in the service of meaningful content learning, the curriculum involved teaching the students about how honeybees collect nectar from a complex systems perspective. Complex systems are increasingly important in science education. Pre- and post-tests in the form of structured interviews were analyzed using a Structure-Behavior-Function framework to identify shifts in student understanding from a complex-system perspective. These interview results were combined with analysis of video data of students' ongoing classroom activities. Results suggest that kindergarten and first grade students not only learned about how honeybees collect nectar, but were also able to do so at the level of aggregate behaviors, and functions, both of which have previously proven elusive for students. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]