Table 2.

Average regression parameters derived from a posterior distribution of multiple PGLS analyses conducted in BayesTraits in the form ‘baculum parameter ∼testes mass + body mass'. λ, lambda (indicates degree to which traits are evolving as predicted given the underlying tree topology); α, intercept; r², coefficient of determination; β, slope; p, significance of the model; CI, 95% confidence intervals. Significant (p < 0.05) values indicated by an asterisk. Baculum length in mm. Body mass in kg. Testes mass in g. All data is log₁₀-transformed prior to analysis.

trait	n	λ	α ± SE	r2	predictor	β ± SE	p	β CI
baculum length	55	0.93	1.30 ± 0.15	0.43	testes mass	0.15 ± 0.09	0.042*	0.01–0.30
					body mass	0.16 ± 0.08	0.032*	0.02–0.30
baculum complexity	55	0.95	−0.60 ± 0.24	0.06	testes mass	0.09 ± 0.13	0.191	−0.11–0.33
					body mass	−0.21 ± 0.12	0.039*	−0.42–0.02