TABLE 5.

Comparison between predicted and simulated genetic parameters for the neutral model

	t	E[M]		M		E[VA]		VA		E[VB]		VB
k = 1	20	0.815	(0.002)	0.815	(0.002)	0.023	(0.000)	0.024	(0.000)	0.035	(0.002)	0.033	(0.002)
	50	0.713	(0.003)	0.713	(0.003)	0.050	(0.001)	0.050	(0.001)	0.260	(0.014)	0.231	(0.012)
	100	0.613	(0.004)	0.613	(0.004)	0.051	(0.001)	0.051	(0.001)	0797	(0.041)	0.678	(0.036)
k = 2	20	0.818	(0.002)	0.819	(0.002)	0.023	(0.000)	0.023	(0.000)	0.037	(0.002)	0.033	(0.002)
	50	0.718	(0.003)	0.719	(0.003)	0.050	(0.001)	0.050	(0.001)	0.271	(0.017)	0.231	(0.015)
	100	0.619	(0.004)	0.622	(0.004)	0.051	(0.001)	0.051	(0.001)	0.830	(0.051)	0.716	(0.042)
k = 4	20	0.816	(0.001)	0.815	(0.001)	0.024	(0.000)	0.024	(0.000)	0.042	(0.003)	0.039	(0.003)
	50	0.712	(0.001)	0.713	(0.001)	0.053	(0.001)	0.053	(0.001)	0.320	(0.022)	0.287	(0.020)
	100	0.610	(0.002)	0.614	(0.002)	0.058	(0.001)	0.058	(0.001)	1.028	(0.066)	0.833	(0.055)

Comparison is shown of values predicted by the neutral analytical model for the mean E(M), additive variance E(V_A), and between-line variance E(V_B) after t consecutive bottlenecks of size N = 50 and those obtained by simulation (M, V_A, and V_B) in the D scenario (model I), for different types of gene action (k = 1, no epistasis; k = 2, moderate synergistic epistasis; k = 4, strong synergistic epistasis). Standard errors are in parentheses.