Raw data and supervised statistical multivariate analyses of the physiological and biochemical response of three wheat genotypes to mild water deficit under elevated CO2 and high temperatures [Dataset]

Comprises 6 files: - ‘Figure S1_Bendou et al.docx’ comprises one single page: ‘Figure S1. Structure of the raw LEYQ data set and multivariate analyses applied to the data set using the ade4 package.’ - ‘TS1_TS5_Supplementary material_Bendou et al.xlsx’ comprises 5 sheets: ‘Table S1. Raw data set.’ ‘Table S2. Extreme outliers of the raw data set.’ ‘Table S3. Shapiro tests.’ ‘Table S4. Levene tests.’ ‘Table S5. Selection of tests (two-way, one-way Anova or T test) to be performed (if needed) according to the presence or absence of extreme outliers and the compliance with the assumptions of normal distribution and homocedasticity.’ -‘TS6_TS7_Supplementary material_Bendou et al.xlsx’ comprises 2 sheets: ‘Table S6. List of variables showing statistical significance after performing, firstly, two-way ANOVA and then one-way ANOVA or T test if there was no statistical significance for an interaction effect between factors.’ ‘Table S7. Summary of post-hoc tests.’ -‘TS8_TS10_Supplementary material_Bendou et al.xlsx’ comprises 3 sheets ‘Table S8. Dudi output elements of the BCA.’ ‘Table S9. Results of the BCA Monte-Carlo test (randtest in ade4).’ ‘Table S10. BCA column normed scores of the variables of the four meaningful blocks. -‘TS11_TS12_Supplementary material_Bendou et al.xlsx’ comprises 2 sheets ‘Table S11. Result of the COIA Monte-Carlo test (randtest in ade4).’ ‘Table S12. COIA loadings for the variables of the blocks of Yield and Quality.’ -‘TS13_TS15_Supplementary material_Bendou et al.xlsx’ comprises 3 sheets ‘Table S13. MBPLS regression using the grain yield as the dependent data block and flag leaves and ears as the explanatory blocks.’ ‘Table S14. Projection of the dependent variables on the LCs and loadings of the explanatory variables in the MBPLS regression.’ ‘Table S15. Determination of the cumulated variable importance index (vipc) to obtain the variables of the explanatory blocks (i.e., leaves and ears) that best explained the dependent block (i.e., grain yield) in the MBPLS regression.

The effect of mild water deficit starting at the vegetative state on the physiological and biochemical response of wheat was investigated in a pot experiment. Wheat plants were grown in a controlled warming environment, in which the atmospheric CO2 concentration and temperatures were set to match those projected by the end of this century. Three wheat genotypes of different polyploidy level were selected: two bread (hexaploidy) genotypes named Gazul and HTWSN 41 respectively, and a durum (tetraploid) genotype named Regallo. The set of the potted plants of each genotype were split into two groups when they reached stage 14 within the decimal Zadoks growth scale. One was maintained under well water conditions, at field capacity, and another followed mild water deficit imposed at 65% of field capacity. At ear emergence, foliar pigments and leaf photosynthesis were measured. Additionally, flag leaves and ears from a subset of each group per genotype and water treatment were harvested to measure morphophysiological, water-related and biochemical traits. The remaining potted plants were left until maturity. The above- and belowground parts of the plants were harvested and the grains separated from the ears to quantify both yield- and quality-related traits. A total of 81 continuous measures were collected from the analyses performed on flag leaves and ears at ear emergence, and grains and total biomass at maturity. The variables were divided into four blocks named Leaf, Ear, Yield and Quality to performed supervised statistical multivariate analysis with the ade4 package. The statistical analysis of the results showed that three genotypes experienced reductions in leaf nitrogen balance and the number of productive tillers, though leaf photosynthesis and water use efficiency were rather stable. Ear water content and antioxidant status largely remained unchanged. Grain yield per plant decreased in the three genotypes, while grain yield per ear and grain biomass displayed reduced variability. Both leaves and ears equally contribute to grain yield. Under water deficit conditions, bread wheat genotypes showed a significant increase in nitrogen content, whereas Regallo maintained higher stable nitrogen content and better grain quality except for iron content. From a methodological viewpoint, the use of several functions maintaining the structure of an object of the subclass dudi (duality diagram) within the ade4 package open a new predictive statistical methodology in the screening of wheat with relevant organ and grain traits

This work was supported by Ministerio de Ciencia, Innovación y Universidades [grant numbers PID2019-107154RB-100 and PID2023-148311OB-I00, MICIU/AEI/10.13039/501100011033 and FEDER, UE] and the Junta de Castilla y León [grant number CSI260P20]. The Project ‘CLU-2019–05—IRNASA/CSIC Unit of Excellence’ funded by the Junta de Castilla y León and co-financed by the European Union (ERDF ‘Europe drives our growth’), the CSIC Interdisciplinary Thematic Platform (PTI) Optimization of Agricultural and Forestry Systems (PTI-AGRO4FOOD) and the CSIC scientific network WheatNet are also acknowledged.

No