Spectral Signature-Based Water Stress Characterization and Prediction of Wheat Yield under Varied Irrigation and Plant Bio-regulator Management Practices

Abstract

Canopy reflectance based spectral indices help in effective irrigation scheduling of wheat for optimization of yield in water-scarce regions. A field experiment for two consecutive years (2013 to 2015) was conducted to evaluate the responses of wheat crop to exogenous application of plant bio-regulators (PBRs) in the water-scarce Deccan region of India (Baramati, Pune, Maharashtra). We predicted grain and biomass yields of wheat using water stress-sensitive spectral indices under varied water regimes. The water regimes were seven levels of irrigation water (equaling to 1.00, 0.85, 0.70, 0.55, 0.40, 0.25 and 0.10 times of cumulative open pan evaporation, CPE) and applied using a line-source sprinkler system). There were five PBRs, viz. thiourea, salicylic acid, potassium nitrate, gibberellin and ortho−silicic, with concentration 10 mM, 10 μM, 15 g L−1, 25 ppm and 8 ppm, respectively, applied at various growth stages, namely crown root initiation, flag leaf and seed milking stages. Water stress indices were computed from spectral reflectance pattern recorded at different crop growth stages using ASD FieldSpec-4 Spectroradiometer (350–2500 nm). The PBRs significantly influenced the canopy reflectance pattern and maintained superior values of water stress indices over the control (without PBRs) by stabilizing leaf pigments and water contents, controlling the stomatal opening and better water use. Among the five PBRs, thiourea and salicylic acid mitigated water stress better and improved overall grain yield (4.6–17.5%) and total biomass (3.6–15.3%). There was no significant (p < 0.05) variation in both yields (grains and biomass) up to IW: CPE 0.70, indicating that irrigation scheduling at 0.70 IW: CPE could be a better option rather than full irrigation in water-scarce areas. At flowering and milking stages, all spectral indices were correlated significantly with the wheat yields. Thus, these stages could be considered as more water-sensitive stages during entire wheat growth period. Regression models based on WI and NWI-2 accounted for 92% and 78% variation in the observed yields for grain and biomass, respectively, with minimum root-mean square error. Hence, to predict the grain and biomass yields of wheat, regression models based on WI and NWI-2 at milking stage can be used successfully.