Thermal imaging and multivariate techniques for characterizing and screening wheat genotypes under water stress condition

The present study was undertaken to understand the combined impact of thermal imaging and multivariate techniques to characterize and screen different wheat genotypes under water stress condition. Ten wheat genotypes namely HD-2987, HDR-77, PBW-343, PBW-175, HD-2967, HD-2781, HD-2985, HD-3043, C-306 and PBW-502 were grown under two irrigation conditions (I and Io corresponding to IW/CPE ratio of 1.0 and 0.4, respectively). The water stress conditions in wheat genotypes were characterized thermo-graphically using canopy temperature-based stress indices, namely crop water stress index (CWSI), stomatal conductance index (IG) and stomatal resistance index (CSI3). Simultaneously, plant biophysical parameters like normalised difference vegetation index (NDVI), leaf area index (LAI), transpiration, stomatal conductance, photosynthesis, were measured at peak vegetative stage along with yield and its attributes (panicle number, tiller number, 1000 seed weight, crop yield, total dry matter, harvest index) were observed. These parameters were analyzed using multivariate techniques namely principal component analysis (PCA), hierarchical cluster analysis (HCA) and discriminant analysis (DA) to develop a new index called normalized water stress tolerance index (NWSTI), to group wheat genotypes depending on their ability to tolerate water stress. Based on these multivariate analysis, wheat genotypes were classified as tolerant (C-306, HD-3043, HD-3987 HD-3985 and HD-2781), moderately tolerant (HDR-77, PBW-175 and PBW-502) and sensitive (HD-2967, PBW-343) to water stress with NWSTI score of >0.66, 0.33–0.66 and <0.33, respectively. Structural matrix from DA analysis showed that biophysical parameters like stomatal conductance, harvest index, photosynthesis, transpiration, NDVI and panicle number have significant (p < 0.001) role in discriminating wheat genotypes into three different groups of water stress tolerance. Distinct location of group centroid in the canonical distribution function further signifies (p < 0.001) the truthfulness of the HCA based dendrogram prepared using the normalized values of crop biophysical parameters. A significantly higher correlation (R² , p < 0.001) between NWSTI and the thermal image-based stress indices (that characterize the instantaneous crop condition for the ten wheat genotypes), prove the usefulness of NWSTI as a stress index. In future, the newly developed water stress index NWSTI, may play a potential role in the screening and selection of drought tolerant wheat genotypes.