Table 1.

Summary of the final GLMM explaining total performance for comparison of the experimental and ancestral population in time. (a) Fecundity assessed on bean: the final model included time, the origin of the mite (experiment or control) and their interaction. (b) Fecundity assessed on cucumber: mites from the experiment performed significantly better. (c) Fecundity assessed on pepper: no difference between control and experimental populations was found, only time had a significant negative effect.

		estimate	s.e.	Z-value	Pr (>|z|)
(a) bean	(intercept)	47.5620	4.2964	11.070	< 2 × 10−16***
	four months	−7.1388	3.8591	−1.850	0.0643
	six months	−4.2116	3.8049	−1.107	0.2683
	eight months	−0.8783	3.8049	−0.231	0.8174
	10 months	−18.4544	3.9770	−4.640	3.48 × 10−16***
	Experiment	−3.5087	3.3864	−1.036	0.3001
	four months : exp.	10.9829	4.1670	2.636	0.0084**
	six months : exp.	4.1284	4.1055	1.006	0.3146
	eight months : exp.	0.1652	4.0926	0.040	0.9678
	10 months : exp.	8.8113	4.2952	2.051	0.0402*
(b) cucumber	(intercept)	28.869	1.211	23.846	<2 × 10−16***
	experiment	2.889	1.136	2.544	0.011*
(c) pepper	(intercept)	1.9886	0.1101	18.063	< 2 × 10−16***
	four months	−0.8767	0.1926	−4.552	5.32 × 10−6***
	six months	−1.4587	0.2074	−7.035	1.99 × 10−12***
	eight months	−0.6534	0.1529	−4.274	1.92 × 10−5***
	10 months	−1.5919	0.2145	−7.421	1.17 × 10−13***