Improving Nitrogen Fertilizer Management in Subsurface Drip-Irrigated Cotton

An irrigated cotton study was established in Maricopa, Arizona, on a Casa Grande sandy loam. The objectives of the study were: (1) Compare lint yields and nutrient use efficiency (NUE) with soil test-based nitrogen (N) fertilizer management with canopy reflectance-based urea-ammonium nitrate (UAN)-N management approach in subsurface drip irrigated cotton; (2) Compare lint yields and NUE for full and deficit irrigation in subsurface drip irrigated (SDI) cotton; and (3) Construct N balances for subsurface drip irrigated cotton, i.e. quantify total N uptake, N recovery efficiency, nitrate leaching, and denitrification losses. Pre-plant soil nitrate-N (0 to 36 in.) averaged 51 lb/A. Our soil test-based N rate was 154 lb/A (225 lb N/A target minus 51 lb N pre-plant minus 20 lb N/A from irrigation water). We used the same soil test N rate for both the 100 and 75% ET irrigation level, in order to make these treatment comparisons strictly for water level.

Nitrogen deficiency in this study appeared rapidly in several vegetation indices as significant differences in N-fertilized plots vs. zero-N, on day 150, 15 days after the start of fertigation. Several vegetation indices for the reflectance-based N treatment fell significantly below soil test N plots, 29 days after fertigation commenced (N rates for reflectance were initially 50% of soil test target of 154 lb N/A). The last two weeks of the fertigation period, UAN injection rates were the same between the two treatments. Final reflectance-based N rate was 112 lb/A, a significant 42 lb/A less than the soil test treatment. In early August, first open boll biomass samples were taken. Biomass was high at 12,000 lb/A for soil test N and 100% irrigation, and 9,200 lb/A for soil test N at 70% water. Nitrogen and water effects on canopy height related very well to normalized difference vegetation index (NDVI). There was no yield reduction with reflectance-based management compared to soil test N, although biomass was less in the former. A significant savings of 42 lb/A less fertilizer N was achieved with reflectance-based N management. However, the lint yields of 1,600 lb/A were lower than the target of 2,000 lb/A. It is not clear why under SDI we are not obtaining the 1,800 lb/A lint yields we observed with furrow irrigation and overhead sprinkler in earlier studies. The high recovery efficiency of fertigated N of 90 to 92% in the soil test-based, 100% irrigation treatment was not unexpected but is a significant result that solidifies the hypothesis that NUE is very high in fertigated drip systems.