JEFFERY, J.E.; BININDA-EMONDS, O.R.P.; COATES, M.I.; RICHARDSON, M.K.; Institute of Evolutionary and Ecological Sciences, University of Leiden, NL; Technical University of Munich, Freising-Weihenstephan, Germany; Department of Organismal Biology & Anatomy, University of Chicago, USA: New methods for determining heterochronic change

Developmental sequence (�event�) heterochrony is a widespread evolutionary phenomenon, although it has always proved difficult to study within a cladistic framework. This is partly because there is no �universal clock� with which to compare timing differences between species. To overcome this problem, the �event-pairing� method was developed in the 1990s. An �event-pair� is a statement, in the form of a tri-state character, about the relative timing of two developmental events. Event-pair statements are made for every possible pair of events and together these encode the developmental sequence. The numeric scores of event-pairs can easily be compared between species using standard cladistic methods. When optimised on a phylogenetic tree, the synapomorphies at each node take the form of shared event-pair score changes. However, when taken individually, each synapomorphy can only give information about a change in the relative timing of two events. Further analysis is necessary before the underlying changes in the developmental sequence can be identified. To this end, we have developed an exhaustive method which will always identify the smallest set (or sets) of events whose movements could have given rise to the observed pattern of event-pair score changes. Time-limits mean that this method can only be applied to medium-sized data sets. However, we have developed a related heuristic method, which is slower but more accurate than established heuristic methods. Identifying specific instances of heterochrony allows a broader study of the possible mechanisms and selective pressures which might have given rise to the changes.