N uptake and agronomic efficiency for rice production under the dry-seeded delayed-flood system in Italy

Dry seeding with delayed flooding at the 3-4 leaf stage has become more and more widespread in Italy since the mid-1980s. Some conditions are needed to dry-seed rice such as good weather, a dry and drying field at sowing, a leveled soil surface for further water management, and well-prepared soil (plowed, leveled, fertilized, and harrowed) to place the seed at no more than 2-3 cm depth. Some of these conditions (i.e., fields that dry fast at sowing to support tractors moving on them, fine soil preparation) are found more easily in the area of light soils located south of Milan and in Novara Province and northeast Pavia Province. Although several studies were conducted to pinpoint the correct strategy for N management in dry-seeded delayed-flood rice, little information comes from experiments conducted in Italy. The present field experiment aimed at evaluating N uptake by the plant, leaf N content, grain yield, and agronomic efficiency in typical conditions for dry-seeded rice in Italy. Two Italian varieties with different growth patterns were grown. Four preplant N applications (0, 60, 120, and 180 kg N ha¯¹) and three split applications at 120 kg N ha¯¹ were considered. Split N was applied at sowing, at preflooding, and at panicle differentiation. A split-plot design with three replications was adopted. The study was carried out at two sites in Pavia Province in 1996 and 1997. Grain yield was higher with split N applications, mainly with fertilizer topdressed at midtillering. High preplant-only N application led to good grain yield but at a higher N rate. N accumulation in the biomass was significantly affected by N application and ranged from 90 to 230 kg N ha¯¹. Agronomic efficiency (AE) varied from 5.0 to 38.5 kg grain kg¯¹ N depending on location, amount of N, and application timing. Higher AE corresponded to N applied in splits at sowing and preflooding or at preflooding and panicle differentiation, with some differences between years. Leaf N content at 52 and 85 days, after sowing in the best-yielding treatments was always high. Thus, farmers' leaf N may be a useful parameter for optimizing grain yield.