Fig. 7
Joint analyses on the effect of male sexually selected traits and female placentotrophy on speciation rate. Bayesian Markov chain Monte Carlo posterior distributions of speciation “main effects” are presented from joint analyses of male and female traits. Ten-parameter models, which include main effects of the two traits on speciation and extinction and an independent model of character evolution (i.e., four transition rate parameters, gain and loss of each trait), were implemented in diversitree using the “make.musse.multitrait” command. These analyses were used to investigate whether male sexually selected traits or placentation was a stronger predictor of speciation rate. lambdaC/D/O refers to the main effect of speciation rate for taxa with courtship, sexual dichromatism, or ornamentation, respectively, and lambdaP refers to the main effect of speciation rate for placentotrophic taxa. We analyzed all paired combinations of male and female traits. The 95% credibility intervals for the effect of each trait on the rate of speciation are indicated as horizontal colored bars above the x axis. Male traits are gray and female traits are orange. The distribution of parameter estimates strongly suggests that male sexually selected traits are associated with a higher speciation rate and female placentation is not
