Abstract: As a main component of carotenoid in tobacco leaves, lutein not only plays an important role in protection against light stress, but also produces many important aroma compounds in tobacco including megastigmatrien, β-damascenone, 3-hydroxy-α-ionone, when lutein degraded. A pot experiment with the design of quadric circum-regression was performed to investigate the effects of N, P, K fertilizers and watering on lutein in tobacco leaves. The zero level of four factors was designed as 4, 6, 12 g and 70%, and the intervals as 2, 3, 6 g and 15%, respectively. The results from interaction test indicated that the first order interactions, the second order interaction of (i) N, P and water, (ii) P, K and water, the third order interaction were all significant, and interaction between N and P, N and K, P and K, P and soil water, K and soil water content was synergistic. But there existed a critical value in the interaction of N and water, when soil water content was lower than the critical value, the interaction between them was synergistic, otherwise it was antagonistic. The second order interaction analysis showed that K and soil water content were more important than P for increasing lutein content in tobacco leaves. It suggested that the proper ratio of N, P, K fertilizers, and watering could increase the lutein content. More P and optimal soil relative water content would increase the lutein content when nitrogen was deficient. More N and K and reasonable watering could increase the lutein content at the lower P level. More N, P, and K could compensate the deficiency of relative soil water content. The results showed that the lutein content accumulated significantly at the high N (6 g pot^-1), high K (18 g pot^-1 K₂O), moderate P with the ratio of N:P₂O₅:K₂O at 1:1.3:3.1, as well as optimal watering maintaining the relative soil water content of 74.3%–79.5%.

Key words: Tobacco (Nicotiana tabacum L.), Lutein, Water