. 2012 Apr 17;7(4):e34629. doi: 10.1371/journal.pone.0034629

Table 1. Values and statistics for Responsiveness, Latency (mean±s.e.m.), Foraging time (mean±s.e.m.), Empty flowers (mean±s.e.m.).

	≠ Colour ≠Brightness			≠Colour = Brightness		= Colour ≠Brightness
	yB	Yb	bY	YB	BY	Bb	bB	B
Responsiveness (%)	87.0	69.0	77.8	88.9	89.0	83.3	87.0	82.6
Latency time (s)	22±8	65±1	56±13	40±1	52±12	56±9	55±10	39±9
Foraging time (s)	180	180	180	180	180	180	180	180
Empty flowers	9.7±0.9^A	10.3±0.8^Aα	6.1±1.4^β	11.2±2.6^A†	9.7±0.8^†	6.6±1.2^a	10.1±1^b	6.3±1.3^a
Number of replicates (N)	23	29	18	25	27	24	23	24

Number of replicates indicates the number of moths that were exposed to each flower model (the base for the responsiveness percentages). The statistical tests are based on an α-level = 0.005 after a Bonferroni correction. Responsiveness (G-test): Gh = 5.52; p = 0.7; N = 193; Latency (ANOVA): F_{(7, 155)} = 2.64; p = 0.0135. Each comparison between models for the variable empty flowers is denoted by a superscript of a different type (A, a, α, and †). Statistically significant differences are denoted by different characters within each type (e.g. “a” and “b”, or “α” and “β”). Empty Flowers (Kruskal-Wallis tests): χ² = 35.6; p<0.0001. Comparison among models with a yellow cross (A): χ² = 2.35; p = 0.31. Comparisons among all models with only blue colour (a, b): χ² = 10.79; p = 0.0045; only Bb vs. bB: χ2 = 10.13; p = 0.0015. Comparisons among “inverted patterns” Yb vs. bY (α, β): χ2 = 11.6; p = 0.0007; YB vs. BY(†): χ² = 7.2; p = 0.0074.