Table 2.

Aphid survival and fecundity.

Aphids	Phenolic class	χ2	DF	P > χ2	Parameter estimates full model		Preferred model
Survival of founders	Condensed tannins	3.32	3	0.34	Intercept	4.44 (0.69)***	Intercept
					GTmean	0.03 (0.14)n.s
					Induced	0.42 (0.52)n.s
					G*I	− 0.30 (0.20)n.s
	Total phenolics	0.56	3	0.91	Intercept	5.34 (1.46)***	Intercept
					GTmean	− 0.12 (0.19)n.s
					Induced	0.10 (0.19)n.s
					G*I	− 0.03 (0.05)n.s
	Anthocyanins	0.65	1	0.42	Intercept	2.19 (3.60)n.s	Intercept
					Preformed	0.50 (0.63)n.s
Fecundity	Condensed tannins	65.62	3	< 0.0001	Intercept	9.62 (0.84)***	9.66 ± 0.84<0.0001
					GTmean	− 1.04 (0.14)***	− 1.07 ± 0.12<0.0001
					Induced	0.42 (0.52)n.s	0.57 ± 0.220.017
		65.50	2	< 0.0001	G*I	− 0.30 (0.20)n.s
	Total phenolics	64.62	3	< 0.0001	Intercept	15.21 (1.50)***	14.59 ± 1.41<0.0001
					GTmean	− 0.90 (0.14)***	− 0.88 ± 0.14<0.0001
		64.31	2	< 0.0001	Induced	0.22 (0.18)n.s	− 0.15 ± 0.060.018
					G*I	− 0.02 (0.04)n.s
	Anthocyanins	3.60	1	0.058	Intercept	6.07 (2.21)**	Intercept
					Preformed	− 0.02 (0.02)n.s

Aphid survival and fecundity. GLM models with Poisson distribution function and identity link estimating the general effect of phenolic classes on aphid survival and reproduction across experimental set-ups. The full model includes genotype mean of the phenolic class (GTmean), induced value of the same per sample, and the interaction. Preferred model indicates model after backward removal of insignificant effects. Model summary given in bold. Suggesting a lack of effect on founder survival but effects of both CTs and PTs on fecundity. Moreover, anthocyanins measured non-destructively before the start of the experiment explained neither survival nor fecundity of the aphids.

n.s. not significant.

Model significances***P < 0.0001;**P < 0.01;*P < 0.5.