Exploring optimal fertigation strategies for orange production, using soil-crop modelling

<p>Water and nitrogen (N) are two key limiting factors in orange (Citrus sinensis) production. The amount and the timing of water and N application are critical, but optimal strategies have not yet been well established. This study presents an analysis of 47 fertigation strategies examined by a coupled soil-crop model. The strategies include 27 main scenarios with a factorial design of 3 irrigation levels (420, 520 and 640 mm, representing 80%, 100% and 120% of water demand), 3 N input levels (100, 200 and 300 kg ha^-1) and 3 main N split applications, and 20 additional scenarios testing alternative N split applications, extreme rainfall years (dry and wet) and different soil textures.The simulations showed that orange yields were strongly influenced by N input and N split applications, but not so much by irrigation. Increasing water and N input led to increased N losses (via leaching and denitrification), and there were significant positive interactions between water and N input with respect to N losses. On average, low N input (100 kg ha^-1) led to relatively low N losses (16 kg ha^-1) but resulted in low yield (33 t ha^-1, 25% yield reduction). High N input (300 kg ha^-1) produced a high yield (43 t ha^-1) but led to large N losses (104 kg ha^-1). Optimal N input (200 kg ha^-1) significantly reduced N losses (45 kg ha^-1) without yield reduction. Importantly, with optimal N input, improving N split applications significantly increased yield by 13% and reduced N losses by 40%, compared to sub-optimal N splits.Significant interactions between water inputs, N inputs and N split applications in yield and N losses indicate that the optimization of fertigation strategies must consider these three key variables simultaneously. Our results clearly show that over-optimal water and N inputs lead to large water and N losses. Reduced irrigation (80% of water demand) and N input equal to N demand (200 kg ha^-1) can significantly reduce N losses without yield reduction. The N split applications should be adjusted to the N demand of the crop during the growing season. Our study focused on a Mediterranean climate, but the methodology and results can be applied to other situations in the world.