Table 1.

Comparison of the latency to initiate a flight response

Factors	Estimate ± SE	t value	p value
Intercept	− 0.09 ± 0.414	− 0.23	0.822
C ➔ (NPP, NPS, PP, PS)	− 0.002 ± 0.004	− 0.59	0.556
(NPP, NPS) ➔ (PP, PS)	− 0.03 ± 0.008	− 3.71	< 0.001
NPP ➔ NPS	− 0.007 ± 0.011	− 0.63	0.529
PP ➔ PS	0.004 ± 0.011	0.38	0.708
Size	0.190 ± 0.198	0.96	0.339
Sex	− 0.003 ± 0.026	− 0.12	0.901

All raw values were negative. To achieve a normally distributed error structure, data were transformed to positive values by reversing their sign to the opposite and subsequently square root transformed. Note that this results in estimates having opposite signs as well. Intercept estimates represent the grand mean of all treatments. Orthogonal comparisons of the treatments are displayed. The arrows indicate the direction of comparison within the contrast. The estimate value always refers to the treatment left of the arrow. If treatments are combined in parentheses, mean values of these treatments are used in the comparisons. The reference category for factor sex is ‘females’; N = 20 test fish in 100 trials, p-values < 0.05 are highlighted in bold

C control, NPP herbivore picture, NPS herbivore smell, PP predator picture, PS predator smell