The age of insects and the revival of the minimum age tree

Abstract

AbstractDeriving node calibrations from fossils remains by far the most common method for attaching an absolute timescale to a molecular tree. But this 'node dating' approach has a credibility problem: different studies using the same molecular data and even the same sets of fossils regularly arrive at drastically different age estimates. A major reason for these differences is well known: even well‐dated and firmly placed fossils can only provide a minimum age for a particular node. The time lag between that node and the fossil (or in other words: between the origin of a clade and its first appearance in the fossil record) is on the other hand notoriously difficult to quantify. Using recent studies on the age of the insect tree of life as an example, I discuss different rationales for choosing calibration priors based on fossils and argue that none of them is entirely convincing. I thus suggest shifting the objective of node dating analyses back to what can be achieved in a scientifically justifiable way: a dated tree that reflects the minimum ages of its constituent taxa. While this interpretation of time trees was followed widely in the early years of molecular dating, its conceptual advantages seem to have been suppressed over the promises of methods that allow integrating more complex node‐age constraints, even when those are difficult to justify. I argue that ‘minimum age trees’ should be revived for groups for which more sophisticated approaches, such as total‐evidence dating, are currently infeasible. Although minimum age trees do not provide mean estimates for the age of a group, they hold a lot of information in their own right and allow the testing of various hypotheses from the areas of biogeography and co‐evolution, while avoiding reliance on ad hoc assumptions about calibration densities.