TABLE 1.

Results of analysis of deviance (df: degrees of freedom; χ²: Chi-square value; LR χ²: likelihood ratio Chi-square value; P: P value) from several models examining the effect of treatment on subsequent pine weevil damage in experiment 1.

	Debarked area			Number of feeding scars			Girdling rate
Source of variance	df	χ2	P	df	χ2	P	df	LR χ2	P
Treatment	6	11.32	0.08	6	4.75	0.58	6	6.40	0.38
Provenance	1	6.38	0.01	1	26.28	<0.01	1	5.30	0.02
Height	1	2.23	0.15	1	0.41	0.52	1	0.10	0.75
Treatment × Provenance	6 × 1	8.23	0.22	6 × 1	20.04	<0.01	6 × 1	5.92	0.43

More specifically, these models examined the effect of different plant defense induction treatments [Large stem window damage (WinL), small stem window damage (WinS), needle-piercing damage to the stem bark (P), root bark damage (RD), previous weevil feeding damage (WF), 10 mM MeJA (MeJA), and undamaged seedlings as controls (C)] on levels of damage (area debarked, mm²), number of feeding scars, and girdling rate by pine weevils (H. abietis) in Scots pine (P. sylvestris) seedlings for experiment 1 (Insect feeding tests were conducted at one time point, 12 days post-treatment). The models included the fixed variables: treatment, provenance (Hade, Gotthardsberg), their interaction, and seedling height (cm, measured a day before feeding test) as a continuous covariate. Significant effects (P < 0.05) are in bold.