Stability for wheat genotypes had been compared in Central Zone of the country as per the BLUP and BLUE of yield values. Measures based on ranks of BLUP for 2015–16 i.e. Sis identified G1, G5, G7, G6. Corrected yield measures CSis pointed towards G1, G2, G5, G6. Values of NPi(s) identified G1, G2, G7. Overall similarity among non-parametric measures tested by Kendall’s coefficient of concordance. Positive correlations of Sis, CSis & NPi(s) observed with other measures. Biplot analysis exhibited cluster of CV with CCV, Si1, S2,’ Si4, S5, Si7 , S3, Si6, NPi(2), NPi(3) & NPi(4). Based on BLUE’s of genotypes, Sis found G1, G7, G4, G5 while CSis identified G5, G4, G2 as opposed to G7, G1, G4 genotypes as by values NPi(s). Positive and negative correlations exhibited by Sis, CSis & NPi(s) with the measures. Biplot analysis observed large cluster comprised of Yield with GAI, NPi(2), NP(3), NP(4), Si1, S2, Si4Si3, Si5, Si6 measures. Second year of study (2017-18) as per BLUP’s seen, Sis settled for G8, G7, G2 genotypes’ While NPi(s) settled for G1, G2, G8, G5. Negative correlations of yield had been observed with MR, CV, Med, NPi(2), NPi(3), NPi(4) while positive with GAI, CMR, CSi6. Measure CV expressed affinity with NPi(2), NPi(3) & NP(4), SD, Si3, Si6, Si1, S2, Si4, Si5, Si7 in Biplot analysis.- Measures Sis as per BLUE’s pointed towards G2, G5,G8, G7 whereas CSis settled for G6, G8, G7. . Wheat genotypes G8, G2, G7, G5 favoured by least values of NPi(s). Positive correlation Sis, CSis & NPi(s) with others. Large cluster of CCV, CSD, NP(1), Si1, Si2, Si4, CSi1, CSi2, CSi3, CSi4, CSi5, CSi6, CSi7 and Z1measures.
Stability of wheat genotypes had been compared as per the BLUP and BLUE of yield values for Central Zone of the country.
Kendall’s coefficient of concordance resulted an overall similarity among rank-based measures based on BLUP for first year. Positive correlations of rank-based measures based on BLUP viz. Sis, CSis & NPi(s) with other measures.
While both type of correlations observed among measures as per BLUE of genotypes.
First two PCA’s based on BLUP’s account for more of G*E sum of squares as compared to corresponding components by BLUE.
Significant genotype-by-environment (G*E) interaction had been reported in large number of multi environmental studies (Pour
The three defined objectives for current study was (1) analyse stability performance by nonparametric measures based on BLUP and BLUE of wheat genotypes yield (2) differentiate the performance pattern of wheat genotypes as per BLUP and BLUE values and (3) study the relationships, similarities and dissimilarities among non-parametric measures of stability.
Seven promising wheat genotypes were evaluated in research field trials at 14 centers of All India Coordinated Research Project on Wheat across zone during 2015–16 and a set of nine genotypes at twelve centers for 2017–18 cropping season. Randomized block designs had been laid out in field evaluation of genotypes with four replications. Recommended agronomical practices as per zone had followed in total to harvest the good wheat yield of genotypes. Parentage details and environmental conditions were reflected in
Non parametric measures for stability analysis proposed by Thennarasu (1995) as NRi(1), NRi(2), NRi(3) and NRi(4) based on ranks of corrected means of genotypes. In the formulas,
Significance of Si(1) and Si(2) non parametric measures had been explored by Nassar and Huehn (1987). Z1 and Z2 values were calculated for each genotype, based on the ranks of adjusted data and then sum of i.e. Z1 sum and Z2 sum are distributed as x2. Degree of similarity among measures had assessed by estimating correlation coefficients while considering genotypes ranking. Spearman’s rank correlation values among pairs (Piephoand Lotito 1992) estimated as follows :
where
Mean wheat yield observed, G5 as the highest yielding with 58q/ha followed by G2 and G4, though remarkable yield differences were observed among the genotypes (
Seven nonparametric measures based on original grain yield of genotypes (Si1, Si2, Si3, Si4, Si5, Si6 and Si7) indicated that (G1, G7, G6), (G1, G6, G5), (G1, G7, G6), (G1, G6, G5), (G1, G6, G5), (G1, G7, G6), (G1, G6, G5) as sets of genotypes respectively. According to corrected yield (
Calculated value of Kendall’s coefficient of concordance (W=0.40) and for the significance of W value the magnitude of x2 = 72.3 statistics was less than table of x2 (0.01, 290) = 135.8, which resulted an overall similarity among non-parametric measures (Vaezi
Spearman’s rank correlation analysis as per BLUP’s of genotypes yield had been reflected in
The loadings of measures based on first two significant principal components were reflected in
Average yield of genotypes showed G5 as of highest yield with 58 q/ha followed by G2 and G6, genotypes (
Si1, Si2, Si3, Si4, Si5, Si6 and Si7 indicated that (G7, G1, G5), (G7, G4, G1), (G7, G1, G4), (G7, G4, G1), (G7, G5, G4), (G7, G1, G4), (G7, G1, G4) as sets of desirable genotypes as per respective measure. G2, G5 & G1 by CMR values, G5, G4 & G2 by CSD and G5, G2 & G4 were the stable as per CCV. Median favoured G2, G6, G1 genotypes. CSi1, CSi2, CSi3, CSi4, CSi5, CSi6, CSi7 identified stable genotypes (G5, G4, G2) and G7, G3 were of unstable type as per these nonparametric measures. NPi(1) considered G5, G4 and G2 as desirable genotypes. Genotypes G7, G1 and G2 had expressed the lower values of NPi(2), while as per lower values of NPi(3) & NPi(4), G7, G1, G4 and lastly by Z1 and Z2 values selected G1, G2, G6 as suitable as well as G7 & G3 would be of unsuitable performance.
Calculated value of W (0.23) and for its significance x2 = 42.5 statistics was less than table of x2 (0.01, 290) = 135.8, which resulted an overall similarity among non-parametric measures. Values of Z1 sum = 18.9 and Z2 sum = 23.9 (
Yield has expressed highly significant positive correlation with GAI, CSD, CCV, CSi1, CSi2, CSi3, CSi4, CSi5, CSi6, CSi7, NPi(2) along with negative correlation with other measures (
High yield achieved by G3 followed by G6, G7 wheat genotypes, whereas GAI selected G3, G6, G7 genotypes, large values of mean ranks selected G2, G5, G8 more over the consistent yield of G8, G7, G2 expressed by least values of standard deviation (
Average mean of ranks as per corrected yield values selected G7, G5, G6 and corrected standard deviation observed suitability of G8, G1, G7 genotypes. Coefficient of variation as per corrected yield values exhibited G7, G5, G8 while median values for G5, G6, G7 and G8, G1, G7 by CSi1 & CS2, CSi3 pointed for G8, G7, G5 & as per CS4 wheat genotypes G8 G1 G7, G1 G8 G7 by criterion of CSi5 & CSi6 settled for G7 G6 G5 and lastly by values of CSi7 genotypes G8, G5, G6 (
Concordance coefficient W=0.46 and for its significance x2 = 110.8 statistic was less than table of x2 (0.05, 290) = 124.3 (135.8), which resulted an overall similarity among non-parametric measures. Values of Z1 sum = 14.4 and Z2 sum = 18.6 (
Spearman’s rank correlation analysis observed highly significant negative correlations of yield with MR, CV, Med, NPi(2), NPi(3), NPi(4) and positive with GAI, CMR, CSi6 (
Biplot analysis of nonparametric measures based on BLUP’s of genotypes (2015-16)
Biplot analysis of nonparametric measures based on BLUE’s of genotypes (2015-16)
Biplot analysis of nonparametric measures based on BLUP’s of genotypes (2017-18)
Biplot analysis of nonparametric measures based on BLUE’s of genotypes (2017-18)
Measure |
Component PC1 |
Component PC2 |
---|---|---|
Yield |
0.1316 |
0.2552 |
GAI |
0.1315 |
0.2553 |
MR |
-0.0953 |
-0.2753 |
SD |
-0.2150 |
0.1450 |
CV |
-0.0285 |
0.2921 |
Med |
-0.0550 |
-0.2895 |
Si1 |
-0.1238 |
0.2345 |
Si2 |
-0.2163 |
0.1461 |
Si3 |
-0.1061 |
0.2738 |
Si4 |
-0.2150 |
0.1450 |
Si5 |
-0.1878 |
0.1624 |
Si6 |
-0.0509 |
0.2926 |
Si7 |
-0.2019 |
0.0885 |
CMR |
0.0954 |
-0.0615 |
CSD |
-0.2521 |
-0.0609 |
CCV |
-0.2411 |
0.0067 |
CMed |
0.0835 |
-0.0558 |
CSi1 |
-0.2547 |
-0.0460 |
CSi2 |
-0.2506 |
-0.0692 |
CSi3 |
-0.2549 |
-0.0250 |
CSi4 |
-0.2521 |
-0.0609 |
CSi5 |
-0.2456 |
-0.0935 |
CSi6 |
-0.2396 |
-0.0263 |
CSi7 |
-0.2471 |
-0.0119 |
NPi(1) |
-0.2444 |
-0.0998 |
NPi(2) |
-0.0930 |
0.2580 |
NPi(3) |
-0.0498 |
0.2781 |
NPi(4) |
-0.0464 |
0.2799 |
Z1 |
-0.1700 |
-0.1536 |
Z2 |
-0.1827 |
-0.1560 |
% variance |
49.50 |
36.38 |
Five clusters among 30 rank-based measures had been expressed Biplot analysis and the loadings of measures were shown in
Measure |
Component PC1 |
Component PC2 |
---|---|---|
Yield |
-0.2322 |
-0.0864 |
GAI |
-0.2307 |
-0.0944 |
MR |
0.2263 |
0.0825 |
SD |
-0.1891 |
-0.1642 |
CV |
-0.2231 |
-0.1055 |
Med |
0.2000 |
0.1107 |
Si1 |
-0.1874 |
-0.1704 |
Si2 |
-0.1810 |
-0.1687 |
Si3 |
-0.2191 |
-0.1378 |
Si4 |
-0.1891 |
-0.1642 |
Si5 |
-0.1611 |
-0.1768 |
Si6 |
-0.2189 |
-0.1358 |
Si7 |
-0.2076 |
-0.1255 |
CMR |
-0.0884 |
0.2577 |
CSD |
0.1900 |
-0.2045 |
CCV |
0.1622 |
-0.2522 |
CMed |
-0.0898 |
0.1423 |
CSi1 |
0.1948 |
-0.1928 |
CSi2 |
0.1924 |
-0.2033 |
CSi3 |
0.1741 |
-0.2377 |
CSi4 |
0.1900 |
-0.2045 |
CSi5 |
0.1639 |
-0.2426 |
CSi6 |
0.1430 |
-0.2744 |
CSi7 |
0.2131 |
-0.1479 |
NPi(1) |
0.1791 |
-0.2124 |
NPi(2) |
-0.1178 |
-0.2542 |
NPi(3) |
-0.1375 |
-0.2429 |
NPi(4) |
-0.1432 |
-0.2323 |
Z1 |
0.1649 |
-0.0487 |
Z2 |
0.1498 |
-0.0866 |
% variance |
56.08 |
28.66 |
Higher average yield had expressed by G3, G6, G7 wheat genotypes, whereas G3, G6, G7 for possessed higher adaptability index values, mean of ranks selected G2, G9, G8 more over the consistent yield of G2, G8, G5 expressed by least values of standard deviation (
CMR selected G7, G6, G5 and CSD observed suitability of G7, G8, G6 genotypes. CCV exhibited genotypes G7, G6, G8 while median values for G9, G7, G6 and genotypes G7, G8, G6 by CSi1, CS2 , CSi3 CSi4 measures whereas as per least values of CSi5 genotypes were G8, G7, G6 & for CSi6 values G7, G8, G6 while G7,G6, G8 by CSi7 (
W=0.42 and x2 =101.2 statistic was less than table of x2 (0.05, 290) = 124.3 (135.8), which resulted an overall similarity among non-parametric measures. Values of Z1 sum = 8.08 and Z2 sum = 9.9 (
Majority of highly significant negative correlations of yield had evident from table. At the same time yield expressed positive values observed with GAI, CMR, CCV, CMed and CS7 (
Measure |
Component PC1 |
Component PC2 |
---|---|---|
Yield |
0.0266 |
0.3596 |
GAI |
0.0179 |
0.3601 |
MR |
-0.0285 |
-0.3472 |
SD |
0.2233 |
0.0219 |
CV |
0.1705 |
0.2377 |
Med |
-0.0282 |
-0.3441 |
Si1 |
0.2237 |
0.0035 |
Si2 |
0.2259 |
0.0100 |
Si3 |
0.2036 |
0.1525 |
Si4 |
0.2233 |
0.0219 |
Si5 |
0.2221 |
0.0109 |
Si6 |
0.1802 |
0.2147 |
Si7 |
0.2194 |
0.0379 |
CMR |
0.0327 |
0.1105 |
CSD |
0.2171 |
-0.1007 |
CCV |
0.1958 |
-0.1435 |
CMed |
0.0366 |
0.1042 |
CSi1 |
0.2170 |
-0.0945 |
CSi2 |
0.2160 |
-0.1098 |
CSi3 |
0.2108 |
-0.1317 |
CSi4 |
0.2171 |
-0.1007 |
CSi5 |
0.2115 |
-0.0916 |
CSi6 |
0.2068 |
-0.1391 |
CSi7 |
0.2002 |
-0.1009 |
NPi(1) |
0.2068 |
-0.1078 |
NPi(2) |
0.1794 |
0.2126 |
NPi(3) |
0.1682 |
0.2408 |
NPi(4) |
0.1674 |
0.2431 |
Z1 |
0.1804 |
-0.1675 |
Z2 |
0.1773 |
-0.1776 |
% variance |
63.64 |
24.60 |
Yield |
-0.0767 |
-0.3088 |
---|---|---|
GAI |
-0.0763 |
-0.3077 |
MR |
0.0621 |
0.3231 |
SD |
-0.2290 |
-0.0348 |
CV |
-0.1739 |
-0.2297 |
Med |
0.0749 |
0.2930 |
Si1 |
-0.2287 |
-0.0239 |
Si2 |
-0.2305 |
-0.0318 |
Si3 |
-0.2084 |
-0.1527 |
Si4 |
-0.2290 |
-0.0348 |
Si5 |
-0.2304 |
-0.0176 |
Si6 |
-0.1874 |
-0.2038 |
Si7 |
-0.2163 |
-0.0575 |
CMR |
0.0423 |
-0.2524 |
CSD |
-0.2084 |
0.1627 |
CCV |
-0.1747 |
0.2142 |
CMed |
-0.0391 |
-0.0712 |
CSi1 |
-0.2073 |
0.1611 |
CSi2 |
-0.2096 |
0.1594 |
CSi3 |
-0.1947 |
0.1901 |
CSi4 |
-0.2084 |
0.1627 |
CSi5 |
-0.2159 |
0.1282 |
CSi6 |
-0.1943 |
0.1805 |
CSi7 |
-0.1797 |
0.2128 |
NPi(1) |
-0.2201 |
0.1038 |
NPi(2) |
-0.1855 |
-0.1984 |
NPi(3) |
-0.1834 |
-0.2070 |
NPi(4) |
-0.1781 |
-0.2151 |
Z1 |
-0.1707 |
0.0508 |
Z2 |
-0.1425 |
0.0561 |
% variance |
58.08 |
29.15 |
Values of the loadings for measures as per first two significant principal components axes (PCA) were shown in
BLUP’s of wheat genotypes provide more valid estimates of yield in multi environment trials and more variations accounted by first two significant principal components of nonparametric measures. More affinity among measures had reflected by a smaller number of clusters in biplot analysis based on BLUP's. Association of Sis, CSis, NPi(s) with other measures is independent of ranks as per BLUP or BLUE of genotypes. Positive and direct relationships exhibited by these measures with other nonparametric measures.
Authors sincerely acknowledge the training by Dr. J. Crossa and financial support by Dr. A.K. Joshi & Dr. R.P. Singh CIMMYT, Mexico along with hard work of the staff at coordinating centers of AICW & BIP project to carry out the field evaluation and data recording.