Computational protocol: Suchian Feeding Success at the Interface of Ontogeny and Macroevolution

Similar protocols

Protocol publication

[…] To identify shifts in evolutionary rates, we modeled the changes in RAP length relative to body size throughout the evolution of Suchia. All calculations were performed in R, using the packages “evomap,” “phytools,” “ape,” “diversitree,” and “geiger” (respectively: ; ; ; ; ). This first involved assessing the fit of standard models for continuous phenotypic character change and identifying the presence and probability of shifts in the adaptive landscape of RAP length and bite forces. These preliminary analyses are described in detail in the Supplementary Material. According to the best-fit model of character change (Supplementary Table S1), we derived ancestral character states and evolutionary rates for both RAP and HW, using an adaptive peak, multiple variance Brownian motion (mvBM) model (adaptive peak: ; mvBM: ). This model allows variable rate estimation on individual branches, which renders it well suited for modeling evolution of traits that are subject to multiple selective pressures (; ; ; ; ). The flexibility offered by a model that infers variable evolutionary rates helps mitigate the inaccurate tracing of character history inherent in methods that rely on a single evolutionary model (e.g., directional, Brownian motion or single-optimum Ornstein-Uhlenbeck; ; ; ; ; ). The mvBM model uses a Bayesian Markov chain Monte Carlo resampling protocol (10,000,000 iterations) to estimate ancestral states at internal nodes and to quantify rates of morphological change along each branch (, ). For complete details of evolutionary model fitting and selection, see Supplementary Material.Following rate and character value estimations, we conducted reduced major axis regression of RAP length and HW evolutionary rates (). The residuals of these rates were then plotted against each other to create a plot of relative evolutionary rate space (; ). This rate space represents tradeoffs encapsulated by six possible evolutionary scenarios (). Disproportionately higher bite forces (when compared to body size) can be achieved under three scenarios: (1) accelerated increase (AI), whereby the RAP increases faster than body size; (2) decelerated decrease, whereby the RAP decreases slower than body size; and (3) complete separation of the traits, such that pure RAP elongation is achieved. Conversely, a relatively lower bite forces can be achieved under inverse scenarios—a pattern not identified for direct ancestors of extant suchians. These relative changes for each trait are then mapped onto each branch in the phylogeny (; ). For this study, only AI is considered, as this scenario represents the most rapid, positive bite-force changes (i.e., the scenario that most closely mirrors bite-force allometry throughout development). FIG. 1 FIG. 2 […]

Pipeline specifications

Software tools Phytools, APE, Diversitree, GEIGER
Application Phylogenetics
Organisms Homo sapiens