Abstract: The effects of phosphorus and nitrogen fertilizers on yield and its components have been widely studied in wheat (Triticum aestivum L.), and it is generally understand that phosphorus fertilizer has the increase potential of spike number and 1000-kernel weight, whereas, nitrogen fertilizer may promote the spike number and grain number per spike but reduce the 1000-kernel weight. However, these conclusions were obtained in studies with high inputs of phosphorus and nitrogen. To gain insights into the effect of fertilizer input level on the yield response of wheat in dryland, we planted nine winter wheat varieties in the field where no fertilizer applied for six years. Compared with the control (zero input of fertilizer), dry matter and kernel yield increased by 17% and 20% in the treatment of low fertilizer, and by 35% and 40% in the treatment of high fertilizer, respectively. In the low-fertilizer treatment, spike number and kernel number per spike were 11% and 8% greater than those in the control, respectively. In the high-fertilizer treatment, larger values of the two traits were observed (25% and 10%, respectively). The increments of dry matter, spike number, and kernel number per spike were greater in high-yielding varieties than in low-yield ones, showing a stronger response of high-yielding variety to the input of fertilizer. However, there was no effect of fertilizer rate on 1000-kernel weight. The results implied that yield, dry matter, spike number, and kernel number per spike of different varieties had diverse responses to the level of fertilizer input under dryland condition. High-yielding varieties were characterized with high levels of dry matter, spike number, and kernel number per spike, and their sensitive responses to fertilizer rates contributed greatly to higher yields when more fertilizer was applied.

Key words: Winter wheat, Fertilizer rate, Grain yield, Agronomic trait