Deficit irrigation for enhancing sustainable water use: Comparison of cotton nitrogen uptake and prediction of lint yield in a multivariate autoregressive state-space model

Deficit irrigation is an optimization strategy for achieving sustainability of irrigated crop production. A field-study of cotton (Gossypium hirsutum L.) response to a limited water supply was conducted in an Alfisol in the southern High Plains of Texas. The objectives were to investigate cotton N uptake, canopy temperature, plant spectral index and lint yield variation under deficit irrigation and to provide information for enhancing sustainability of the water resources and Alfisols in the semi-arid environment. The experimental treatments were two deficit-irrigation levels at 50% and 75% of cotton evapotranspiration (ET). Plant and soil variables were measured 15 m apart along the center-pivot irrigation circles. The results show that cotton plants under the 50%-ET deficit irrigation level were 21% more water stressed (P < 0.05) based on the reflectance water index ratio. The 50%-ET irrigation resulted in a 25% lint yield loss with a 33% water saving compared to the higher irrigation level (75%-ET). Plant reflectance, canopy temperature, total N uptake and lint yield were correlated with normalized difference vegetative index (NDVI), soil water content (SWC), soil NO₃-N concentrations and elevation (−0.69 < r < 0.72, P < 0.05, respectively). Future cotton lint yield is weighted on NDVI and water variation, quantified in a multivariate autoregressive state-space model. Increases in plant reflectance in the water band are signs of early plant water stress. Compared to the 12-year regional cotton lint yield obtained with full irrigation, the 75%-ET deficit irrigation would be agronomical, economical efficient in Alfisoils with only 7.8% of lint-yield loss from water stress but 25% of water saving for sustainable water use.