Results of continuous-time Markov models of discrete trait evolution for the trait eggshell type for 32 squamate species.

Shown are models of discrete trait evolution that assume different character states models (ER = equal rates, ARD = all-rates-different, meristic = stepwise transition between countable traits) and different lambda (λ) values (star-like tree with λ = 0; original tree with λ = 1). Additionally presented are the model likelihoods based on the Akaike Information Criterion (AIC) (AICc = sample-size-corrected AIC). Models with an untransformed phylogenetic tree (original tree) provided the best AIC values for the different models of character states (ER, ARD, meristric). Only for these models the transition rates (qnm with subscripts for eggshell types; 1 = shell-less, 2 = parchment-shelled, 3 = rigid-shelled) from one eggshell type to another are shown. *The λ value was additionally estimated and optimized by the fitDiscrete function (see main text). In the case of the meristic trait model, a direct transition from shell-less to rigid-shelled and vice versa is impossible (q13 or q31 = 0).Results of continuous-time Markov models of discrete trait evolution for the trait eggshell type for 32 squamate species.