. 2014 Mar 20;7(4):506–518. doi: 10.1111/eva.12154

Table 1.

Evaluation of alternative models for genetic parameters based on population size and isolation variables.

	Models with D.centre					Models with D.near
	N.pop	Area	D.centre	Area + D.centre	Area^D.centre*	N.pop	Area	D.near	Area + D.near	Area^D.near*
H_e
AIC_c	−91.9	−85.9	−99.6	−103.6	−115.6	−91.9	−85.9	−87.3	−89.8	−89.1
ΔAIC_c	23.8	29.7	16.0	12.0	0.0	0.0	5.9	4.5	2.1	2.7
w_i	0.0	0.0	0.0	0.0	1.0	0.6	0.0	0.1	0.2	0.1
F_IS
AIC_c	−63.2	−64.2	−63.0	−61.8	−79.0	−63.2	−64.2	−63.2	−62.1	−69.2
ΔAIC_c	15.7	14.8	16.0	17.2	0.0	6.0	5.0	6.0	7.1	0.0
w_i	0.0	0.0	0.0	0.0	1.0	0.0	0.1	0.0	0.0	0.8
F
AIC_c	−123.0	−119.5	−139.5	−137.1	−134.6	−123.0	−119.5	−121.9	−120.0	−117.0
ΔAIC_c	9.5	20.0	0.0	2.4	4.9	0.0	10.5	8.1	10.5	13.0
w_i	0.0	0.0	0.7	0.2	0.1	1.0	0.0	0.0	0.0	0.0
D_ST
AIC_c	−133.2	−127.4	−145.8	−143.4	−141.1	−133.6	−127.4	−133.4	−131.0	−129.0
ΔAIC_c	12.6	18.5	0.0	2.4	4.7	0.2	6.0	0.0	2.4	4.4
w_i	0.0	0.0	0.7	0.2	0.1	0.4	0.0	0.4	0.1	0.0

Genetic parameters: H_e = expected heterozygosity, inbreeding (F_IS), accumulated inbreeding (F) and differentiation (D_ST). Model selection repeated for two alternative measures of isolation: distance to the centroid of all populations (D.centre) and distance to the nearest population (D.near). N.pop = male territories, Area = patch size. Additive (+) and interactive (^*) effects were considered. The corrected Akaike information criterion (AIC_c), difference with the best model (ΔAIC_c) and the relative weight of evidence for each model (w_i) are reported (two equally good models when w_i < 0.5). Statistics of the most parsimonious model are highlighted in bold.