Objective  To optimize N×S interactions at differentiated habitats for maximal yield and fertilizer use efficiency of fragrant rice.

  Methods   A split-plot experiment was conducted in 2018 at two typically different fragrant rice production areas, Habit 1 at Jiangmen, Guangdong and Habit 2 at Chenxi, Hunan. The main lots were designed for varied applications of N and S, while the sublots for planting of two cultivars of high-quality fragrant rice. The N application rates included 0 (N₁), 90 kg·hm⁻² (N₂), and 180 kg·hm⁻² (N₃), and the S at 0 (S₁), 45 kg·hm⁻² (S₂), and 90 kg·hm⁻² (S₃). A total of 9 treatments of combined N×S applications, i.e., N₁S₁ (control), N₁S₂, N₁S₃, N₂S₁, N₂S₂, N₂S₃, N₃S₁, N₃S₂, and N₃S₃, with 3 replicates on the randomly distributed sublots were implemented.

  Results   (1) Because Habit 2 had a wider range of daily temperatures, cooler night temperature, longer growing period for the rice, and greater ability of the red paddy soil to retain water and fertilizer than Habit 1, the rice grown on Habit 2 had significant higher yield, number of productive spikes, and total grain number per panicle (P＜0.01). (2) N₃S₃, N₃S₂, N₂S₂, N₂S₁, N₂S₃, and N₃S₁ significantly increased the rice yield and N₃S₂, N₂S₂, N₂S₁, N₂S₃, and N₃S₁ significantly raised the number of productive ears per unit area, seed setting rate, and 1 000-grain weight of the rice but significantly reduced the total grain number per panicle (P＜0.01). (3) At Habit 1, N₂S₃ significantly increased the yield and N use efficiency of the rice variety, Xiangwanxian No.13. At Habitat 2, the yield and N and S use efficiency of Xiangwanxian No.17 were significantly increased by N₃S₂ with the increases ranking as N₃S₂＞N₃S₃＞N₂S₁＞N₂S₂＞N₃S₁＞N₂S₃ (P＜0.01), which were significantly higher than those for Xiangwanxian No.13. (4) The N use efficiency of Xiangwanxian No.13 at Habit 1 increased and then decreased parabolically with increasing S. The S use efficiency of rice plants reached a peak of 24.44 kg·kg⁻¹ under N₃S₂ and declined at a high S application level, such as N₂S₃ and N₃S₃. On the other hand, the N use efficiency of Xiangwanxian No.17 in Habit 1 increased with increasing S under N₂, but so did N₃S₃. On the S use efficiency of the rice plants, the parabolic effect of N was also observed with N₂S₃ yielding the maximum at 4.85 kg·kg⁻¹. (5) The N utilization by Xiangwanxian No.13 at Habit 2 improved with S application, but the S use efficiency peaked at 19.11 kg·kg⁻¹ under N₂S₂. Whereas Xiangwanxian No.17 showed a decreased N use efficiency upon increasing S, as well as the S use efficiency decreased from 69.27 kg·kg⁻¹ under N₃S₂ to that under N₃S₃.

  Conclusion   To varying degrees habitat, cultivar, and N×S interaction all affected the yield, yield components as well as the N and S use efficiency of the fragrant rice. N₂S₃ significantly increased the yield and N use efficiency of Xiangwanxian No.13, and N₃S₂ of Xiangwanxian No.17.