The diversity of mosses with species-specific N-fixing microbiomes hinted that this mutualism was pretty old, but I never expected to see fossils showing the association this clearly