Table 1. Generalized linear mixed models for (a) the binary trainability response variable and (b) the learning level variable in the acute experiment (experiment 1) using all exposed bees that took part in the task (n = 142 bees).

	Fixed effects	Parameterestimate	SE	Z value	P value
(a) Trainability	Intercept (Control)	−8.53	3.34	−2.56	0.011
	Treatment (250 ppb)	−1.63	0.53	−3.05	0.002
	Treatment (10 ppb)	−0.71	0.51	−1.40	0.164
	Treatment (2.4 ppb)	0.07	0.53	0.13	0.895
	Bee size	1.86	0.67	2.78	0.005
(b) Learning level	Intercept (Control)	−1.87	1.06	−1.77	0.077
	Treatment (250 ppb)	−0.79	0.17	−4.70	<0.0001
	Treatment (10 ppb)	−0.38	0.15	−2.63	0.009
	Treatment (2.4 ppb)	0.07	0.13	0.53	0.599
	Bee size	0.55	0.20	2.70	0.007

Parameter estimates are calculated with reference to the control group. Colony and week of PER testing were included as random effects. Significant p values (p < 0.05) are highlighted in bold.