Chuckle. That experiment compared Europeans ("Caucasians") and Asians. It had absolutely nothing whatsoever to do with eye color. For whatever bizarre reason, the authors simply defined "Caucasians" to be light-eyed (!) and Asians to be dark-eyed, then pretended that the difference in melatonin suppression between the two groups had something to do with eye color. The authors only grudgingly admit that the difference might also be due to "ethnicity"--i.e., any of the many other average genetic differences between Europeans and Asians.
Shockingly, the authors did not even bother to tell us just where these "Caucasians" and Asians were really from. Scandinavia or Spain? Kamchatka or Cambodia?
Given the fact that the authors tried to ascribe the observed difference to eye color, it is utterly inexcusable that they didn't even bother to tell us the subjects' eye color! ("Ten healthy Caucasian males with blue, green, or light brown irises (light-eyed Caucasians)...")
That paper does bring up a point that we have discussed in this forum once or twice: The eye color and hair color that Europeans call brown, Africans and Asians call light; and the eye color and hair color that Africans and Asians call brown, Europeans call black. This crossover shows in the word brunet(te), which can mean either dark brown or black.