Teaching modeling in computer science is complicated. Many factors contribute, and are related in diverse ways. We regard some combinations as more successful than others, but we also value diversity, and we struggle to elucidate the relationships and our value structure. Similar remarks apply to the study of biological ecosystems. This contribution views teaching modeling as an ecosystem: we push the analogy as far as we can, with the intention of provoking readers' thought in unaccustomed directions, which may perhaps lead to interesting advances. We identify the factors that contribute to the ecosystem and establish some relationships between the constituent factors. We discuss some of the factors that need further work and improvement. Finally, we discuss what it should mean for the ecosystem of the teaching of modeling to be healthy.