Improving Nitrogen Fertilizer Use Efficiency in Subsurface Drip‐Irrigated Cotton in the Desert Southwest

Abstract

Declining water availability in the American Southwest continues to generate interest in efficient subsurface drip irrigation (SDI) for cotton (Gossypium hirsutum L.) production. Fertigating urea ammonium nitrate (UAN) at low rates with high frequency is an important advantage of SDI. However, N fertilizer management guidelines specific to SDI cotton are lacking. A 3‐yr study was conducted on a Casa Grande sandy loam soil in Maricopa, AZ, to test a pre‐plant soil profile NO3 test algorithm and a canopy reflectance approach to manage in‐season N fertilizer for SDI cotton. Treatments included soil test‐based N management, reflectance‐based N management, and zero‐N at 100% evapotranspiration irrigation replacement. A second irrigation level of 70% evapotranspiration replacement included just the soil test‐based N and zero‐N treatments. The five treatments were replicated three times. Soil test–based N treatments received from 172 to 224 kg N ha−1, and reflectance‐based N amounts were 112 to 158 kg N ha−1. Nitrogen recovery efficiency (RE) of UAN‐N was high, with 24 fertigations during 6 wk between first square and mid bloom ranging from 58 to 93%. The isotope dilution method estimated similar RE in 2017. Residual post‐harvest soil NO3–N was notable only with 70% irrigation. Lint and seed yields were significantly reduced with the 70% irrigation treatment compared with 100% irrigation. The key result of this study is that reflectance‐based N management saved 17 to 112 kg N ha−1 without reducing lint yields compared with the soil test–based N treatment.

Which active optical sensor vegetation index is best for nitrogen assessment in irrigated cotton?

“…This information is contained for NDVIR, NDVIA, and NDRE in Figure 1 and 2, but Table 1 allows for easily comparing all 12 VIs. In addition, Table 1 shows that lint and seed yields for all five site‐years were significantly reduced in zero‐N plots vs. soil test plots (Bronson et al., 2017b, 2019). This indicates that N deficiencies in zero‐N plots remained until harvest.…”

“…Canopy reflectance was measured weekly using the Crop Circle ACS‐470 AOS (Holland Scientific Inc., Lincoln, NE), starting about 2 wk after plant emergence and ending about mid‐bloom in two N management irrigated cotton studies in Maricopa, AZ (33.067 N, 111.97 W and 360 m above sea level) on a Casa Grande sandy loam (fine‐loamy, mixed, superactive, hyperthermic, Typic Natrargid); an overhead sprinkler irrigation (OSI) study, 2014–2015 (Bronson et al., 2017b), and a subsurface drip irrigation (SDI) study, 2016–2018 (Bronson et al., 2019). In the OSI study, there were eight N treatments replicated four times, described in detail in Bronson et al.…”

“…In the SDI study there were five N treatments (two of these at a deficit 70% irrigation level) replicated three times as described in Bronson et al. (2019). The experimental design was a randomized complete block in each study.…”

“…If NDVIA from reflectance plot means fell significantly below the plot mean of NDVIA of the soil test‐based N treatments, then the N rate was increased to match the soil test rate (designated “reflectance‐based N‐1″ in Bronson et al., 2017b). In the SDI study, we used NDRE instead of NDVIA to base the same reflectance plot N treatment approach on (Bronson et al., 2019).…”

“…Leaves, stems, seed, and burrs were analyzed for N content on the Leco CN analyzer. Lint and seed yields were sampled by a cotton picker on the center two rows on a 6‐m length, centered on the DGPS points (Bronson et al., 2017b, 2019).…”