Objectives We studied the effects of urea fertilization methods on rice yield, nitrogen utilization efficiency, and the fate of nitrogen fertilizer in soil to propose economically and environmentally friendly N management measures.

Methods A paddy field experiment was conducted in the Taihu Lake region in 2019 and 2020, and a ¹⁵N-urea micro-plot experiment was installed in the same field plot in 2020. Urease inhibitor N-(n-butyl)thiophosphoric triamide (NBPT) and nitrification inhibitor methyl 3-(4-hydroxyphenyl) propionate (MHPP) were added at 1% of urea-N. We included six fertilizer treatments: no N application (CK); broadcasted urea N at 300 kg/hm² (conventional fertilization, CN); broadcasted urea N at 225 kg/hm² (RNB); 50% in deep and 50% in broadcasting of urea N at 225 kg/hm² (RND); broadcasted urea N at 225 kg/hm²+NBPT+MHPP (RNB+DI); 50% in deep and 50% in broadcasting of urea N at 225 kg/hm²+NBPT+MHPP (RND+DI). Broadcasted urea N was applied in three splits (4 : 3 : 3) as basal, at tillering and panicle initiation stages. Deep-placed nitrogen was applied in two splits (7 : 3) as basal, and at panicle initiation stage. Rice yield, N uptake, N utilization, absorptive capacity, soil residual and total ¹⁵N loss were quantified.

Results Compared with CN, RNB, RND, RNB+DI and RND+DI treatments recorded similar grain yield and straw biomass, except the lower straw biomass in RNB in 2020; However, all the four treatments achieved higher N use efficiency, agronomic efficiency and partial factor productivity of nitrogen fertilizer, and ¹⁵N tracing result also proved the significant reduction in ¹⁵N uptake and total N loss under the four treatments. Compared with RNB, RND enhanced apparent N recovery efficiency (NRE) and ¹⁵N use efficiency by 20.4%–27.3% and 28.2%, and reduced ¹⁵N loss by 34.6% (P<0.05). RNB+DI had no impact on NRE, but (P<0.05) increased ¹⁵N use efficiency by 11.6% and reduced ¹⁵N loss rate by 13.1%. RND+DI increased rice yield by 2.6%–4.3%, NRE by 23.4%, ¹⁵N use efficiency by 36.9%, and reduced ¹⁵N loss by 45.0%. Compared with RND, RNB+DI had no impact on rice yield and NRE but reduced the ¹⁵N loss rate.

Conclusions In the rice production of Taihu lake area, the measurement of reducing urea-N rate from 300 kg/hm² to 225 kg/hm², but increasing the basal fertilizer ratio from 40% to 70%, and changing total broadcasting to 50% by deep placement, will not decrease rice yield, but significantly increase nitrogen fertilizer use efficiency, decrease total N loss. Adding NBPT and MHPP in urea will further decrease the total N loss. We recommend that a higher proportion of urea should be allocated to the deep basal placement and a lower proportion to broadcasting, rather than adding N inhibitors, to increase rice yield and N use efficiency.