Skip to main content
. 2012 Dec 5;7(12):e51087. doi: 10.1371/journal.pone.0051087

Table 3. F statistics and p-values for GLM analysis testing for differences in eyespot trait sizes across treatments using wing size as a covariate.

B. anynana J. coenia
Surface Wing Trait N (5E1) N (NS1) F p N (5E1) N (NS1) F p
Ventral Forewing M1 white NA NA NA NA 82 59 1.381 0.241
M1 black 40 48 0.250 0.618 82 59 0.657 0.419
M1 gold 40 48 0.490 0.486 82 59 1.360 0.245
Cu1 white 40 49 0.066 0.797 82 59 0.014 0.905
Cu1 black 40 49 0.115 0.735 82 59 5.455 0.020 *
Cu1 gold 40 49 0.097 0.757 82 59 4.152 0.031 *
Hindwing M1 inner NA NA NA NA 82 59 1.558 0.213
M1 gold NA NA NA NA 82 59 0.001 0.989
Cu1 black/inner 37 49 0.063 0.803 82 59 1.023 0.313
Cu1 gold 37 49 0.125 0.725 82 59 1.482 0.225
Dorsal Forewing M1 white NA NA NA NA 82 59 0.418 0.519
M1 black 40 46 0.049 0.826 82 59 0.612 0.431
M1 gold 40 47 0.469 0.496 82 59 0.788 0.376
Cu1 white 40 48 2.961 0.090 82 59 2.123 0.147
Cu1 black 37 41 0.071 0.790 82 59 5.308 0.022 *
Cu1 gold 37 41 0.386 0.537 82 59 5.337 0.0218 *
Hindwing M1 inner NA NA NA NA 82 59 0.276 0.600
M1 gold NA NA NA NA 82 59 0.259 0.611
Cu1 inner NA NA NA NA 82 59 0.094 0.797
Cu1 gold NA NA NA NA 82 59 0.318 0.573

Individuals were injected with either 5E1 antibody or NS1 control media as larvae. All significant differences correspond to smaller trait sizes in the 5E1-injected individuals. GLM analyses were performed with sex (always significant across traits with females usually displaying a larger trait size than males; not shown); interaction between line and sex (not significant in all analyses; not shown); and wing size as a covariate. *p<0.05.