Delaying tillering nitrogen topdressing until the midtillering phase improves nitrogen use efficiency and reduces ammonia emission via rice canopy recapture

Tillering nitrogen (N) topdressing is commonly employed to enhance primary-tiller number and achieve a high grain yield for wetland rice. Unfortunately, farmers typically apply N fertiliser shortly at tillering initiation following basal fertilisation, and substantial soil ammonia (NH3) volatilisation occurs due to the low N demand of small plants. Herein, we hypothesised that delaying topdressing N until the midtillering phase could increase canopy recapture of soil NH3 emission and better synchronise N supply with crop N demand, leading to reduced NH3 losses and increased fertiliser N use efficiency. A field experiment was conducted to investigate the effect of conventional tillering N topdressing (CT) at the early tillering stage (2 weeks after basal N application) and delayed tillering N topdressing (DT) until the midtillering phase (∼4 weeks after basal N application) on NH3 emission from soil and the soil-plant system, as well as N uptake, N use efficiency and yield. The 2-year field results showed that DT did not hinder effective tiller production or above-ground dry biomass compared with CT, but improved plant N uptake and total N accumulation after midtillering, leading to 18–27 % higher agronomic efficiency. Correspondingly, 8.9–15 % increases in grain yield were observed for DT compared with CT, mainly attributable to increased panicle number per area and filled grain number per panicle. Following tillering N application, NH3 loss from the soil-plant system under CT accounted for 22 % of top-dressed N, showing only a slight decline compared with that of soil emission. For DT, NH3 loss from the soil-plant system was reduced to 6 %, even though 15 % soil NH3 emission was observed, indicating a more than two-fold increase for in-canopy NH3 recapture. Correspondingly, cumulative net NH3 losses for the entire rice season were decreased by 31% under DT relative to CT. These results provide evidence that delaying tillering N topdressing until the midtillering phase when rice canopy formation is enhanced, can better synchronise fertiliser N supply with crop demand and increase in-canopy NH3 recapture. A large discrepancy remains between NH3 emission from soil and the soil-plant system at midtillering and panicle initiation stages, highlighting the overestimation by the soil dynamic chamber method when quantifying NH3 losses due to overlooking the contribution of canopy recapture of soil-emitted NH3.