This article reports on the theoretical output of a design research study, which concerns the design of a learning environment for helping students challenge the ‘balance of nature’-idea and reach an up-to-date understanding about ecosystems’ contingency. Our focus is set on whether it is feasible to articulate an empirically tested theory of teaching/learning about contingency in nature while designing our LE. The study included an exploratory phase and three research cycles. The participants were first year educational sciences students who collaboratively (...) explored computer models that simulated ecosystems’ response to changes, in order to understand the underlying contingency. In the EP, we defined learning objectives/design criteria that informed the LE’s first version. This was driven by the global/overall question of how ecosystems may function, which was explored through inter-connected local/partial questions with the aid of scaffold questions embedded in worksheets linked to the computer models. Drawing upon the RC1 results, we introduced two-version models for each change that the peer groups had to explore and assigned the first version to half and the second to the other half. These changes made it through RC2-RC3 and account for key features of the study’s theoretical output, the ‘Bifurcated Domino Path Approach’. The RC3 results show that it works effectively, and thus, we suggest that the design of LEs addressing contingency in ecosystem’s response can integrate bifurcated domino paths with several forks, where some peer groups explore one version of the target phenomenon and others an alternative one and meeting points, where they all share their different, contingency-indicating conclusions. (shrink)