Table 2.

Results of statistical analyses on simulation results used as pseudo-empirical data when response variable is either extinction/persistence (binomial GAM and binomial GLM), time to extinction (Gaussian GAM and ordinary least-square regression, OLS) or mean phenotype or additive genetic variance at the end of simulation time for the populations that persisted (average of last 10 years) (Gaussian GAM and OLS). edf are the estimated degrees of freedom. Symbol (:) denotes interaction between terms. All terms with * are statistically significant at the 0.05 level. I report mean estimate and standard error of the regression coefficients.

	extinction			time to extinction
	GLM R2 = 0.73	GAM R2 = 0.62	edf	OLS R2 = 0.27	GAM R2 = 0.27	edf	OLS R2 = 0.75	GAM R2 = 0.76	edf	OLS R2 = 0.14	GAM R2 = 0.14	edf
intercept	−1.83(0.04)*	2.16(0.05)*		261.33(1.02)*	245.40(0.37)*		1.416(0.002)*	1.341(0.002)*		1.370(0.001)*	0.151(0.001)*
s	3.09(0.04)*		1.97*	−21.00(0.92)*		1.00*	0.139(0.002)*		1.99*	−0.010(0.001)*		1.00*
p(Ea)	0.41(0.02)*	0.40(0.02)*		−3.21(0.39)*	−3.43(0.31)*		−0.001(0.002)	−0.004(0.002)*		−0.001(0.001)	−0.000(0.001)*
βσ,Θ	1.63(0.03)*	1.78(0.03)*		−10.68(0.84)*	−15.90(0.36)*		−0.005(0.002)*	−0.008(0.002)*		−0.001(0.001)*	−0.001(0.001)*
βμ,Θ	1.10(0.02)*	1.03(0.02)*		3.68(0.81)*	−1.43(0.32)*		0.504(0.002)*	0.471(0.002)*		0.010(0.001)*	0.006(0.001)*
	−0.11(0.02)*		1.00*	−0.63(0.30)*		1.90*	0.047(0.002)*		1.76*	0.010(0.002)*		1.00*
s: βσ,Θ	0.31(0.05)*			−4.82(0.76)*			0.012(0.002)*			−0.001(0.001)
s: βμ,Θ	−0.08(0.04)			−4.66(0.71)*			0.060(0.002)*			0.003(0.001)*
βμ,Θ: βσ,Θ	−0.12(0.03)*			−1.61(0.41)*			0.006(0.002)*			0.001(0.001)
βσ,Θ: p(Ea)	−0.03(0.02)			−0.38(0.38)			0.001(0.002)			0.001(0.001)
βμ,Θ: p(Ea)	−0.06(0.02)*			0.03(0.39)			0.004(0.002)*			0.001(0.001)
βμ,Θ: βσ,Θ:p(Ea)	0.11(0.02)			−0.19(0.38)			−0.001(0.002)			−0.001(0.001)