THE IMPORTANCE OF CONSIDERING PUNCTUATED CHANGE WHEN ASSESSING MACROEVOLUTIONARY SCENARIOS IN THE CONTEXT OF PHYLOGENIES

Most current methods of phylogenetic analysis assume that molecular or morphological character change is continuous and gradual over time. For morphological characters, this typically reflects an expectation of αt changes where α is the rate of change per character in a matrix, and t is the branch duration. This violates how many paleontologists think of morphological change in the fossil record, i.e., in punctuated pulses associated with speciation. To date, the possibility of punctuated morphological change associated with speciation is not well-explored in a Bayesian phylogenetic context.

One reason punctuated models have not been well-explored is that there are many possible ways to do in modern phylogenetic analysis. Here, we examine Skyline Fossilized Birth-Death models, which allow us to assess shifting origination and extinction rates over time given a set of divergence dates on a tree. A Punctuated Equilibrium model expects ελt changes over time t, where ε is the probability of change per speciation event and λ is the origination rate being varied by the Skyline model. Thus, Skyline models + punctuated change predict that rates of anatomical change per million-years increase or decrease as origination rates increase or decrease, even if rates of anatomical change per speciation event (ε) are static. This could have major implications for how we interpret macroevolutionary scenarios that might seem unrelated to speciation modes. For example, “Early Bursts” and post-extinction rebounds often simultaneously increase both anatomical disparity and clade richness. Explanations focusing on ecologic drivers do not necessarily predict elevated ε, whereas explanations focusing on shifting constraints do predict elevated ε. However, if we assess rates using only continuous time models, then both scenarios predict an elevated α and thus are not distinguishable. Thus, accounting for speciation modes often is necessary when testing hypotheses generated from very different types of macroevolutionary theory. Here, we will present a Bayesian framework for contrasting punctuated and continuous models of character change using Skyline FBD models and relaxed morphological clocks, and emphasize that phylogenetic and macroevolutionary studies should accommodate uncertainty in speciation modes.