Effect of water deficit and varieties on the morphological parameters of the cotton plant
The results of the analysis of variance showed that the effect of the water regime was significant on the parameters VBB and RV, and insignificant on the parameters VB, FB, FBB, PH, RD, MRL, and SRN (Table 3). The variety effect was significant on the parameters VB, FB, FBB, PH, RD, and MRL and not significant on the remaining parameters. The interaction was significant only for the parameters MRL and SRN.
Discrimination of means showed that water deficit reduced VBB (-49.56%) and RV (-26.34%) compared to ETM (Table 4).
On the varietal level, we noted that the deficit decreased the VBB for all varieties (-77% to -14%) except for Nazili where we observed an increase of 29%. The results also showed an increase in root volume in the varieties AJ275B (+56%) and BRS 286 (+15%), and a decrease in root volume in the other varieties (-60% to -10%).
Discrimination of means showed that under the ETM regime, the number of secondary roots of Deltapine 5690 was the same as those of FK64, AF 401-13, STONE 907, X148, and STAM 190, but higher than the remaining varieties. Under deficit conditions, the value of the SRN of the variety BRS 286 was identical to those of the varieties AJ 275B and STONE 907, but higher than those of the remaining varieties. These results also showed that the deficit significantly increased the number of secondary roots in BRS 286 (89%) and reduced the number in AF 401-13 (-44%) and Deltapine 5690 (-44%).
Regarding the main root length, the results showed that under the ETM regime, the MRL of variety FK64 was identical to those of varieties AJ275B, Deltapine SL Frego, BRS 286, AF 401-13, and X148, but, higher than the MRL of the other remaining varieties. Under deficit conditions, the MRL of variety X148 was higher than those of the other varieties except for Nazili and BRS 286 which were identical to it. The results also showed that the water deficit caused a significant reduction in the main root length of the Deltapine SL Frégo and FK 64 varieties and an increase in X148, Nazili, and NTALL88 (>15%). Root volume increased significantly in AJ 275B and BRS 286 (> 15%).
Effect of water deficit and varieties on agronomic parameters of cotton
The results of the analysis of variance showed that the effect of the water regime was significant on SCP, WB, AB, and RB parameters and not significant on NSB and FBOD parameters (Table 5). The variety effect was significant on NSB and FBOD parameters and not significant on the remaining parameters. The interaction was non-significant for all parameters.
The discrimination of the means showed that the water deficit decreased SCP (-23%), AB (-29%), and RB (-41%), and increased the WB compared to the ETM on the other hand (table 6).
The results showed that the deficit decreased the SCP in all varieties (-39% to -11%) except BRS 286 for which we noted an increase of +3%.
For the variables AB and RB, the deficit caused a reduction in all varieties. However, this reduction was small to nil in BRS 286.
For the PMC, we noted that the deficit led to an increase in all varieties except Nazili where a reduction of -15% was observed.
Effect of water deficit and varieties on biochemical parameters of cotton
The results of the analysis of variance showed that the effect of the water regime was significant on the parameters ChlT, Chla, and Chlb, and not significant on the parameters PT, PRO, and MDA (Table 7). The variety effect was significant on all parameters except PRO. The interaction was non-significant for all parameters.
Discrimination of the means showed that the water deficit decreases the chlorophyll contents ChlT (-76%), Chla (-79%), and (-70%) compared to the ETM regime (Table 8).
The results showed that the deficit reduced the chlorophyll content in all varieties, but the reduction was smaller in STONE 907 than in the others.
Grouping of varieties according to their similarity
Choice of dimensions to be considered
Following the principle of KAISER (1960), the first two dimensions Dim1 and Dim2 of the AFM are those whose eigenvalues are greater than 1. They summarize about 54.81% of the observations. But to extract more information spread over the other dimensions, Dimensions 3 and 4 were also considered in the interpretation of the data, they summarize 25.89% of the initial information. The first four dimensions sum up 80.7% of the initial information.
Correlation between groups of variables and the main dimensions of the AFM
Three groups of variables were studied. These were agronomic, morphological, and biochemical variables. The results showed that the group of agronomic variables was correlated with dimension1 (ctr= 43.21%, r= 0.95) and dimension 2 (ctr= 29.72%, r= 0.78). The group of morphological variables was correlated with the dimensions Dim1 (ctr= 21.46%, r=0.90), Dim2 (ctr= 57.41%, r=0.90), Dim3 (ctr= 35.00%, r=0.66) and Dim4 (ctr= 57.41%, r=0.89). The group of biochemical variables was correlated with the dimensions Dim1(ctr= 35.33%, r=0.87) and Dim3 (ctr= 55.03%, r=0.82).
Relationship between the different groups of variables
The graph of the groups of variables (Figure 1) shows that in both planes (Dim1-Dim2 and Dim3-Dim4), the groups of variables have been moved away from each other, so they are not related. Therefore, they do not describe the different individuals studied in the same way.
Relationships between individual variables and grouping of individuals
The first dimension was significantly correlated with the variables (i) biochemical: Chla (0.88, p < 0.001), ChlT (0.87, p < 0.001), Chlb (0.82, p < 0.01), (ii) agronomical: UCS (0.74, p < 0.01), PMC (0.74, p < 0.01), PCG (0.72, p < 0.01), D1CAPS (0.70, p < 0.05) and (iii) morphological: BF (-0.58, p < 0.05) (Figure 2). The second dimension was correlated with the the variables (i) morphological: HR (0.83, p < 0.001), BV (0.75, p < 0.01) and CBF (-0.69, p < 0.05) and (ii) agronomic: D1FL (0.77, p < 0.01).
The projection of individuals in the Dim1-Dim2 plane, shows that (i) the variety STONE 907 is characterized by high chlorophyll contents, (ii) the variety Nazili is characterized by a low value of the variables WB and SNB and a high value of FBB (iii) variety X148 is characterized by the high value of PH and VB variables, (iv) variety NTAL88 is characterized by the high value of PH, VB and FFAD variables, (v) varieties Deltapine 5690 and Deltapine SL Frego are characterized by low values of PH, FFAD, and VB variables.
The third dimension was significantly correlated with the variables MDA (-0.81, p < 0.05), PRO (-0.69, p < 0.05) and FB (0.68, p < 0.05) and, the fourth dimension was correlated with the variable SRN (0.73, p < 0.01) (Figure 3).
The projection of individuals in the Dim3-Dim4 plane shows that (i) variety STONE 907 (ctr=25.91%) is characterized by low MDA and PRO content, (ii) variety NTAL 88 is characterized by high MDA and PRO contents, (iii) variety BRS 286 is characterized by high values of PRO and SRN variables and (iv) varieties Deltapine SL Frego and Deltapine 5690 are characterized by low values of SRN.
The hierarchical ascending classification made it possible to group the varieties into 7 classes (Figure 4). Each class is characterized by variables with values higher or lower than those of the general average at the 5% threshold.
- Class 1: it was characterized by the variety with a high value for the variables PRO (758.30 ± 78.1 mg/g MF) and SRN (41.67 ± 5.71 jas). This class consists of the variety BRS 286.
- Class 2: is characterized by the variety with a high value for the variables Chlb (0.16 ± 0.03 mg/g MF), ChlT (0.39 ± 0.08 mg/g MF) and Chla (0.23 ± 0.05 mg/g MF). This class consists of the variety STONE 907.
- Class 3: is characterized by varieties with a low value for the variables PH (73.50 cm ± 1.17 cm) and VB (3.16 ± 0.16 jas). It consists of the varieties Deltapine 5690 and Deltatpine SL Frego ,
- Class 4: is characterized by the variety with a high value of FBB (11.00 ± 1.57) and a low value of the variables WB (2.95 g ± 0.6 g) and SNB (20.33 ± 3.72). It consists of the variety Nazili ,
- Class 5: This class is characterized by varieties whose values do not differ significantly from the mean. This class is composed of the varieties STAM 129A, STAM 190, AF 401-13, FK 64, and AJ 275B.
- Class 6: is characterized by a variety with a high MRL value (53.33 ± 7.40 cm). It consists of the variety X 148.
- Class 7: is characterized by the variety with high values for the variables MDA (3.77 ± 0.32 mg/g MF) and FFAD (53.00 ± 1.72 jas). This class is composed of the variety NTAL 88.