Evidence networks: lessons from the redwood tree