Gene action and heritability in bi-parental crosses of sunflower Received for publication: 12 February, 2020. Accepted for publication: 21 July, 2020. Doi: 10.15446/agron.colomb.v38n3.84472 1 Oil Crops Research Department, Field Crops Research Institute, Agricultural Research Center, Giza (Egypt). * Corresponding author: mohamedtemraz1@yahoo.com Agronomía Colombiana 38(3), 305-315, 2020 ABSTRACT RESUMEN A field experiment was conducted in the contrasting locations of Kafr-El-Hamam/Sharkia and Al-Arish/North Sinai agricul- tural research stations of the Agricultural Research Center in Egypt. We estimated the additive and dominance genetic vari- ances as well as heritability in the broad and narrow senses for sunf lower yield and its attributes in the cross L350 × L355 using North Carolina Design-III. The magnitude of the additive vari- ances for all studied traits at both locations in proportion to the phenotypic variances was larger than the dominance variances. The average degree of dominance was greater than the unit, as also confirmed by the high narrow-sense heritability for most of the studied traits. The genetic improvement may be achieved through the selection of genotypes with larger head diameter and heavier seed weight which also showed moderate to high values of narrow-sense heritability. Thirty-two sunf lower backcrosses were grouped into eight distinct clusters/groups by canonical analysis. Se realizó un experimento de campo en lugares contrastantes de las estaciones de investigación agrícola Kafr-El-Hamam/Shar- kia y Al-Arish/Norte del Sinaí en el Centro de Investigación Agrícola en Egipto. Se estimaron la variación genética aditiva y dominante, así como la heredabilidad en sentido amplio y restringido para el rendimiento y atributos de girasol en el cruce L350 × L355 usando North Carolina Design-III. La magnitud de las variaciones aditivas para todos los rasgos estudiados en ambos lugares en proporción a las variaciones fenotípicas fue mayor que las dominantes, y el grado promedio de dominancia fue mayor que la unidad, como también lo confirma la alta heredabilidad de sentido estrecho para la mayoría de los rasgos estudiados. La mejora genética se puede lograr mediante la selección de genotipos con mayor diámetro de capítulo y mayor peso de semilla que también mostraron valores, moderados a altos, con rasgos de heredabilidad de sentido estrecho. Treinta y dos retro cruzamientos de girasol se agruparon en ocho clús- teres/grupos distintos mediante análisis canónico. Key words: North Carolina Design-III, average degree of domi- nance, selection criteria, cluster. Palabras clave: diseño de Carolina del Norte-III, grado pro- medio de dominancia, criterios de selección, agrupamiento. Gene action and heritability in bi-parental crosses of sunflower Acción genética y heredabilidad en cruces biparentales de girasol Mohamed Ali Abdelsatar1* and Tamer Hassan Ali Hassan1 Introduction The genetic improvement of sunf lower yield and its qual- ity mainly relies on considerable genetic variability in the population. Various breeding procedures such as North Carolina design III (NCD III) can be used for creating new genetic variability, which is obtained by selecting plants in the F2 generation with crossing to both inbred parents. NCD III also provides more information on the nature of gene action governing either additive or dominance inheri- tance of seed yield and its component traits in populations. However, it assumes the absence of epistasis as mentioned by Acquaah (2012). Hence, simple mating designs improve seed yield and its components. The additive gene action was more important than the non-additive one in contributing to the variability of the head diameter (Gvozdenović et al. 2005), 1000-seed weight and seed yield (Ortis et al. 2005), plant height, seed num- ber/head, 1000-seed weight, and seed yield (Karasu et al., 2010). Abd EL-Satar (2017) and Abdelsatar et al. (2020) also confirmed a predominant role of the additive gene action in the control of the days to 50% f lowering, days to physi- ological maturity, plant height, head diameter, number of green leaves/plant, and seed oil content. However, the seed weight/plant took an opposite trend. Thus, selection for improving these traits would be effective in early genera- tions. Nevertheless, the importance and predominance of non-additive or dominance effects was detected for plant height (Gvozdenović et al., 2005; Shankar et al., 2007), days to 50% f lowering (Shankar et al., 2007), and seed yield (Parameshwarappa et al., 2008; Karasu et al., 2010; Chandra et al., 2011). Similarly, Azad et al. (2016) con- cluded the predominance of the non-additive gene action http://dx.doi.org/10.15446/agron.colomb.v38n3.84472 306 Agron. Colomb. 38(3) 2020 in inheritance of all the studied traits except for days to maturity. Additionally, many of the yield-attributing traits in sunf lower are under the inf luence of non-additive gene action as reported by Shinde et al. (2016), Abd EL-Satar et al. (2017), Lakshman et al. (2019) and Ahmed et al. (2019). Consequently, according to the previous authors, the im- provement of these traits can be performed by using hybrid varieties. Seed weight/plant is considered a complex trait; hence, attention must be paid to correlation and path analy- ses at both phenotypic and genotypic levels for improving seed weight/plant. Vidhyavathi et al. (2005) concluded that plant height and head diameter were considered the best selection criteria for improving seed weight/plant. Simi- larly, 100-seed weight and head diameter had the highest direct and indirect inf luence on seed weight/plant as Abd EL-Satar et al. (2017) mentioned. Cluster analysis by D2 statistics is a powerful tool in mea- suring the divergence of tested genotypes on yield and its attributes. Thirumala et al. (2005) grouped 94 sunf lower genotypes into 10 clusters based on D2 values. Moreover, Ram et al. (2018) classified the thirty-two genotypes of sunf lower into six clusters using the Tocher method. Given the previous points, this research aimed to estimate additive and dominance genetic variance, and heritability in broad and narrow senses for sunf lower yield and its at- tributes. Additionally, the selection criteria for improving seed weight/plant were determined using correlation and path analyses at both phenotypic and genotypic levels, and the bi-parental crosses were classified using cluster analysis. Materials and methods Plant material The F1 cross (i.e., L350 x L355) was selected based on its yield performance in F1 evaluation, with the highest heterotic effects over mid- and better parents for further breeding (Abd EL-Satar et al., 2015); this population was used for generating bi-parental progenies. F1 seeds of the tested cross (L350 x L355) were sown on March 5, 2018 at the Al-Arish agricultural research station of the Agricultural Research Center, in Egypt, and F1 plants were selfed at the f lowering stage to produce the F2 seed genera- tion. F2 seeds were sown on June 5, 2018 along with their two parents at the Kafr-El-Hamam agricultural research station of the Agricultural Research Center, in Egypt. Sixteen plants were randomly selected from the F2 popula- tion and grouped in four sets. Each one of these plants was crossed as the male parent with each of the two parents as a female parent to produce 16 crosses of F2 x P1 and 16 of F2 x P2 as in the North Carolina Design III. Field evaluation trials The 32 bi-parental crosses, and the two parents, were hand- planted on adjacent plots without separators on June 5, 2019 by using a randomized complete block design with three replicates in the four sets at the two locations of Kafr-El- Hamam and Al-Arish agricultural research stations of the Agricultural Research Center in Egypt. Soil samples of each location were analyzed to determine their compositions and chemical properties following Jackson (1973), as shown in Table 1. Wheat was planted before in the winter season of 2018/2019 in both locations. Each replicate was grown in a plot consisting of 3 rows. Each row was 4 m long and 60 cm apart, with plants spaced at 30 cm within rows. The cultural practices were followed as recommended for sunf lower production in the region. FIGURE 1. Monthly temperature, rainfall, and relative humidity during the period of sunflower growth in the 2019 season at both locations. 25.91 3.81 63.40 27.42 0.91 60.88 0.70 65.99 2.02 26.18 60.86 27.95 29.58 5.45 39.34 30.57 0.29 39.53 0.54 48.11 0.22 27.73 40.62 30.64 B SeptemberAugustJulyJune 10.00 1.00 0.10 100.00 A Relative humidity (%)Rainfall (mm)Temperature °C Relative humidity (%)Rainfall (mm)Temperature °C Sharkia/Kafr-El-Hamam SeptemberAugustJulyJune 10.00 1.00 0.10 100.00 North Sinai/Al-Arish 307Abdelsatar and Hassan: Gene action and heritability in bi-parental crosses of sunflower Data collection At harvest, ten competitive plants were randomly selected from the first and third rows in each plot to measure plant height (cm), head diameter (cm), 100-seed weight (g), and seed weight/plant (g). The seed weight/plant was adjusted to a 15.5% seed moisture, whereas the days to physiological maturity were determined on a plot basis. Seed yield/m2 was recorded from the plants located in the second row as the middle one, which was 4 m long for a total of 2.4 m2 in each experimental plot. These values were then con- verted to yield ha-1. Seed oil content was determined after drying seeds at 70°C for 48 h, by the Soxhlet extraction technique, using diethyl ether (AOAC, 1990). Gas-liquid chromatography (Aglent 6890 GC, USA) was used for the determination and identification of fatty acids methyl esters such as oleic acid and linoleic acid following Zygadlo et al. (1994) methodology in the Central Laboratory of the Food Technology Research Institute, ARC, Egypt. Statistical analysis The analysis of North Carolina design III was performed as outlined by Comstock and Robinson (1948) to estimate different genetic components. The phenotypic and geno- typic correlation coefficient (Weber and Moorthy, 1952) and phenotypic and genotypic path analysis (Dewey and Lu, 1959) were determined. Additionally, the grouping and arrangement of genotypes by canonical analysis (Rao, 1952) based on D-square statistics (D2) developed by Mahalano- bis (1936) were carried out. Results and discussion Analysis of variance (ANOVA) Mean squares of North Carolina design III due to females/ sets and males/set (Tabs. 2-3) were highly significant for all the studied traits at both locations and their combined analysis. These results indicate that there is great diver- sity among families, and hence, it would be effective for improving these traits. Females x males/sets interaction mean squares at both locations and their combined analysis were highly significant for all the studied traits as shown in Tables 2 and 3, suggesting that some divergent dominance could play a role across the specific crosses and across the sets. The presence of highly significant effects of the loca- tions on all the studied traits (Tab. 3) showed that environ- mental factors inf luenced the phenotypic expression for all the studied traits. Table 3 shows that the mean squares due to the interaction between location and females, males, and females by males were highly significant for all the studied traits. However, the interaction between the location and males for plant height, head diameter, seed weight/plant, seed yield ha-1 and oleic acid did not reach a significant level. The significance of these traits due to the interaction the location with previous populations indicates that the location has sufficient environmental variability to lead to f luctuations in the ranking of these population compo- nents. Similar results were reported by Abd EL-Satar et al. (2015), Abd EL-Satar (2017) and Abdelsatar et al. (2020). Furthermore, female variances (Tab. 2) contributed to the largest relative proportion of variances compared to males and female x male for all the studied traits at the two loca- tions. These results indicate that these differences among females led to obtain great genetic diversity among their crosses with males, so they can be considered as desirable/ preferred in future sunf lower breeding programs. Mean performance The mean performance of all the studied traits for the male x female interaction of backcrosses are shown in Supplementary material 1-3. The results revealed signifi- cant differences among the bi-parental crosses for all the studied traits at both locations. In general, set 2 of backcrosses with P1 obtained the short- est period to physiological maturity (85.00 d and 77.58 d), and the shortest (dwarf) parents (158.65 cm and 152.37 cm) at Kafr-El-Hamam and Al-Arish, respectively. However, the first order of sets was the Set 3 of backcrosses with P2 for head diameter (25.62 cm and 18.29 cm), seed weight/plant (35.46 g and 28.68 g), seed yield ha-1 (2458.88 kg ha-1and TABLE 1. Latitude and longitude of the two experimental field stations and soil composition and chemical properties of the upper 30 cm of the experimental soil. Property Sharkia/ Kafr-El-Hamam North Sinai/ Al-Arish Latitude 30o61’N 31o13’N Longitude 31o50’E 33o80’E Soil composition Sand (%) 17.42 72.35 Silt (%) 35.59 25.22 Clay (%) 46.99 2.43 Soil texture Clay loam Sandy loam Chemical analysis Concentration of N (mg kg-1) 152 14.42 Concentration of P (mg kg-1) 9.52 4.75 Concentration of K (mg kg-1) 456 31.42 Electrical conductivity (ds/m) 0.62 4.76 pH 7.62 7.95 308 Agron. Colomb. 38(3) 2020 TABLE 2. Mean squares of North Carolina design III for all the studied traits in the bi-parental sunflower crosses at Kafr-El-Hamam (K) and Al-Arish (A) in the 2019 season. SOV Df K A K A K A Days to maturity Plant height Head diameter Sets 3 433.13** 204.84** 565.53** 325.12** 124.97** 50.97** Replications 8 5.16 2.38 15.81 11.85 1.11 0.85 Female in sets 4 282.23** 140.51** 765.26** 385.42** 52.09** 20.06** Male in sets 12 153.94** 68.34** 280.75** 141.15** 41.4** 16.57** F x M in sets 12 49.81** 22.37** 120.15** 78.71** 31.4** 10.77** Error 56 3 1.2 7.54 5.85 0.97 0.63 SOV Df 100-seed weight Seed weight/plant Seed yield ha-1 Sets 3 1.09** 2.53** 39.05** 55.03** 253115.07** 233842.96** Replications 8 0.02 0.04 1.93 0.86 1672.04 3267.29 Female in sets 4 3.91** 1.29** 61.58** 86.5** 266613.26** 197240.36** Male in sets 12 1.94** 2.15** 114.2** 84.62** 159450.80** 120517.96** F x M in sets 12 1.59** 0.49** 55** 52.2** 105253.29** 85103.32** Error 56 0.04 0.03 0.94 2.08 2982.03 2862.77 SOV Df Seed oil content Oleic acid Linoleic acid Sets 3 44.42** 12.18** 7.3** 19.7** 252.36** 154.71** Replications 8 0.13 0.33 0.15 0.14 1.2 0.82 Female in sets 4 22.46** 11.81** 4.6** 12.09** 90.66** 70.1** Male in sets 12 28.56** 10.41** 3.38** 8.82** 86.27** 42.02** F x M in sets 12 13.12** 7.51** 2.02** 6.11** 54.76** 36.2** Error 56 0.36 0.39 0.1 0.31 0.97 1.11 SOV: Source of variance; Df: Degree of freedom. *, ** Significant at 0.05 and 0.01 probability level, respectively. TABLE 3. Mean squares of North Carolina design III for all studied traits in the bi-parental sunflower crosses across two locations in the 2019 season. SOV Df DM PH HD 100 SW SWP SYH OC OA LA Replications 16 0.09 0.64 0.44 0.02 0.44 614.6 0.08 0.05 0.1 Location(L) 1 4417.92** 4329.06** 1753.59** 269.21** 2362.51** 10867013.7** 573.01** 105.01** 435.73** Set (S) 3 616.8** 868.88** 167.76** 2.81** 90.59** 482914.4** 51.24** 25.4** 399.51** L x S 3 527.65** 583.42** 190.12** 23.34** 220.4** 1027324.0** 61.9** 15.5** 138.08** Crosses (C) 31 161.13** 368.71** 45.09** 2.3** 132.64** 232122.8** 24.89** 9.31** 101.88** Female (F) 4 410.3** 1104.44** 67.91** 3.36** 140.18** 446944.7** 32.25** 15.68** 158.56** Male (M) 12 165.63** 265.09** 17.47** 2.55** 86.2** 97518.1** 13.47** 3.44** 67.98** F x M 12 113.87** 319.26** 76.38** 2.27** 209.72** 353150.8** 40.07** 15.38** 142.36** L x C 31 9.45** 22.17** 3.77** 0.83** 5.27** 10182.0** 3.12** 0.87** 4.47** L x F 4 28.32** 62.57** 10.37** 2.45** 10.51** 19941.6** 6.05** 2.21** 7.88** L x M 12 8.36* 7.64 0.9 0.27** 3.58 4681.8 1.8** 0.2 3.29* L x M x F 12 6.61** 28.78** 5.38** 1.07** 6.53** 14974.6** 4.26** 1.31** 5.62** Error 112 2.62 8.58 0.88 0.04 1.65 3187.4 0.4 0.22 1.18 SOV: Source of variance; Df: Degree of freedom DM: Days to maturity, PH: Plant height, HD: Head diameter, 100 SW: 100-seed weight, SWP: Seed weight/plant, SYH: Seed yield ha-1, OC: Seed oil content, OA: Oleic acid and LA: Linoleic acid. *, ** Significant at 0.05 and 0.01 probability level, respectively. 309Abdelsatar and Hassan: Gene action and heritability in bi-parental crosses of sunflower 1944.79 kg ha-1), seed oil content (41.76% and 44.08%), oleic acid content (29.42% and 31.62%) and linoleic acid content (52.57% and 54.58%) at Kafr-El-Hamam and Al-Arish, respectively. Moreover, the heaviest weight of 100 seeds was detected in the Set 2 of backcrosses with P1 (7.52 g) at Kafr-El-Hamam and in the Set 3 of backcrosses with P2 (5.25 g) at Al-Arish. Overall, backcrosses with the second parent were slightly inferior (desirable) for earliness in days to maturity (0.91% and 0.74%) than overall backcrosses with P1, which were slightly inferior (desirable) for plant height (0.14% and 0.01%) at Kafr-El-Hamam and Al-Arish, respectively. How- ever, the overall backcrosses with the second parent were slightly superior (desirable) for head diameter (0.88% and 0.31%), seed weight/plant (4.48% and 0.94%), seed yield ha-1 (0.83% and 2.09%), seed oil content (0.19% and 0.98%), oleic acid content (0.44% and 1.30%) and linoleic acid content (3.45% and 3.75%) than overall backcrosses with the first parent at Kafr-El-Hamam and Al-Arish, respectively. Generally, the backcrosses had the highest mean values of all the studied traits at Kafr-El-Hamam, except for quality traits such as seed oil content and oleic and linoleic acids which exhibited the highest proportion at Al-Arish. Based on their mean performance, the backcrosses M4 x P2/Set 3 and M1 x P1/Set 4 performed well for most studied traits and consequently, seed weight/plant at both locations. This is maybe due to the aggregation of favorable genes from different traits in these superior backcrosses. The genetic variability among bi-parental sunf lower crosses is consistent with those reported by Shinde et al. (2016), Abd EL-Satar et al. (2017) and Lakshman et al. (2019). Gene mode of action The estimation of additive (σ2A) and dominance variances (σ2D) along with heritability (h2NS) and average degree of dominance (ā) at both contrasting locations and the com- bined analysis are shown in Table 4. A significant variance of all the studied traits was a good indicator for computing both additive and dominance components. Hence, either hybridization breeding or selection in segregating genera- tions is recommended for bi-parental crosses improvement. The magnitude of the additive variances for all the studied traits at both locations in proportion to the phenotypic variances was larger than dominant ones and the average degree of dominance was greater than the unit. Hence, the selection of these traits in early segregating genera- tions to accumulate additive genes would be successful for producing superior inbred lines to be used in hybrid breeding programs (Tab. 4). Similar results were found by Karasu et al. (2010), Abd EL-Satar (2017) and Abdelsatar et al. (2020). The relative proportion variance of additive to phenotypic variance was larger than the dominance variance for days to maturity, plant height, and 100-seed TABLE 4. Estimation of genetic parameters, additive and dominance variance, and degree of dominance in an F2 population of sunflower backcrosses at the contrasting locations of Kafr-El-Hamam (K) and Al-Arish (A) in the 2019 season. Parameters K A K A K A Days to maturity Plant height Head diameter δ2A 100.62 44.76 182.14 90.20 26.96 10.63 δ2D 31.21 14.12 75.07 48.58 20.29 6.76 h2(b) 97.77 98.01 97.15 95.96 97.99 96.51 h2(n) 74.63 74.51 68.80 62.37 55.91 58.98 (ā) 0.79 0.79 0.91 1.04 1.23 1.13 Parameters 100-seed weight Seed weight/plant Seed yield ha-1 δ2A 1.26 1.41 75.51 55.02 104312.5 78436.79 δ2D 1.03 0.31 36.04 33.41 68180.84 54827.04 h2(b) 98.12 98.20 99.16 97.70 98.30 97.90 h2(n) 54.05 80.61 67.12 60.79 59.45 57.62 (ā) 1.28 0.66 0.98 1.10 1.14 1.18 Parameters Seed oil content Oleic acid Linoleic acid δ2A 18.80 6.68 2.18 5.67 56.86 27.27 δ2D 8.50 4.75 1.28 3.86 35.86 23.39 h2(b) 98.69 96.66 97.16 96.81 98.96 97.85 h2(n) 67.95 56.50 61.32 57.57 60.69 52.67 (ā) 0.95 1.19 1.08 1.17 1.12 1.31 δ2A: Additive variance; δ2D: Dominance variance; h2(b): Broad-sense heritability; h2(n): Narrow-sense heritability and (ā): Average degree of dominance. 310 Agron. Colomb. 38(3) 2020 weight as shown in the combined analysis (Tab. 5). On the other hand, the ratio of dominance to phenotypic variance was more significant than the additive variance, and the average degree of dominance was greater than the unit. These results were obtained for the head diameter, seed weight/plant, seed yield ha-1 and seed oil content as well as for oleic and linoleic acids as shown in the combined analysis (Tab. 5). This effect could be related to genetic interactions (Comstock and Robinson, 1948) and confirm the greatest role of dominance variance in the inheritance of these traits. However, selection in the later generations would be more effective than early selection for these traits. These findings were agreement with Lakshman et al. (2019) and Ahmed et al. (2019). Estimates of broad sense heritability at both locations and their combined analysis were more than 95% for all the studied traits, indicating higher importance of the genetic effects of these traits. Moreover, moderate to high esti- mates of narrow-sense heritability were detected in all the studied traits at both locations, indicating that the effect of additive genetic variance was pronounced compared to dominance genetic variance (Tab. 4). These findings were agreement with Abd EL-Satar (2017) and Abdelsatar et al. (2020). Selection for all traits should be highly effective in this population. Regarding the combined analysis, days to maturity, plant height, and 100-seed weight are the highest heritable traits (Tab. 5). Therefore, selection for these traits should be highly effective in improving this population at early segregating generations. On the other hand, the es- timates of narrow-sense heritability were low to moderate for head diameter, seed weight/plant, seed yield ha-1 and seed oil content as well as for oleic and linoleic acids. The selection of these traits in early segregating generations is not efficient but it would be effective in later generations. These findings were agreement with Abd EL-Satar (2017) and Abdelsatar et al. (2020) who reported a predominant role of additive gene action in the control of days to 50% flowering, days to physiological maturity, plant height, head diameter, number of green leaves/plant, and seed oil con- tent. However, seed weight/plant showed an opposite trend. Association of traits and path analysis as powerful selection criteria Phenotypic and genotypic correlation Phenotypic and genotypic correlations were estimated between seed weight/plant and its attributes for 32 sun- f lower genotypes based on the average of two locations (Tab. 6). Seed weight/plant was positively and significantly or highly significantly associated with days to maturity, plant height, head diameter, and 100-seed weight at phenotypic and genotypic levels. These results indicate that the largest diameter of head with moderate value of narrow-sense heritability and the heaviest weight of 100- seed with a high value of narrow-sense heritability across locations can be considered of the main components to improve seed weight/plant; hence, selection will be more effective for these traits. The same results were obtained by Abd EL-Satar et al. (2017) who found that head dia- meter and 100-seed weight had the highest direct and indirect inf luence on seed weight/plant. TABLE 5. Estimation of genetic parameters, additive and dominance variance, and degree of dominance in an F2 population of sunflower backcrosses across the locations of Kafr-El-Hamam and Al-Arish in the 2019 season. Parameters DM PH HD 100 SW SWP SYH OC OA LA δ2A 52.42 85.82 5.52 0.76 27.54 30945.45 3.89 1.08 21.56 δ2D 20.45 51.47 12.59 0.45 35.32 58825.37 7.08 2.52 24.24 h2(b) 99.40 98.97 99.20 99.46 99.56 99.41 99.40 98.99 99.57 h2(n) 71.51 61.87 30.25 62.54 43.62 34.27 35.26 29.73 46.88 (ā) 0.88 1.10 2.14 1.09 1.60 1.95 1.91 2.16 1.50 DM: Days to maturity, PH: Plant height, HD: Head diameter, 100 SW: 100-seed weight, SWP: Seed weight/plant1, SYH: Seed yield ha-1, OC: Seed oil content, OA: Oleic acid and LA: Linoleic acid. δ2A: Additive variance; δ2D: Dominance variance; h2(b): Broad-sense heritability; h2(n): Narrow-sense heritability and (ā): Average degree of dominance. TABLE 6. Pooled phenotypic (above diagonal) and genotypic (below diagonal) correlations of 32 sunflower bi-parental crosses across two locations. Traits Days to maturity Plant height Head diameter 100-seed weight Seed weight/plant Days to maturity 1.000 0.874** 0.678** 0.466** 0.556** Plant height 0.899** 1.000 0.575** 0.339 0.427* Head diameter 0.695** 0.583** 1.000 0.368* 0.769** 100-seed weight 0.471** 0.342 0.365* 1.000 0.528** Seed weight/plant 0.564** 0.430* 0.786** 0.529** 1.000 *, ** Significant at 0.05 and 0.01 probability level, respectively. 311Abdelsatar and Hassan: Gene action and heritability in bi-parental crosses of sunflower Phenotypic and genotypic path analysis Path analysis is an important tool to partition the correla- tion coefficients into direct and indirect effects of attribute traits on seed weight/plant. Genotypic correlations recorded higher values than pheno- typic correlations regarding the direct and indirect effects of seed weight attributes on seed weight, suggesting the negligible role of the environment on the genotypic expres- sion (Tab. 7 and Fig. 2). Maximum positive direct effects at both, phenotypic and genotypic levels, were observed for head diameter (P=0.695, G=0.732) with a moderate value of narrow-sense heritability, followed by 100-seed weight (P=0.288, G=0.291) with high value of narrow-sense herita- bility. Therefore, a preferred improvement may be achieved by selecting genotypes with a larger head diameter and heavier seed weight. On the other hand, a negative direct effect on seed weight/plant at phenotypic and genotypic levels was exerted by plant height (P=-0.112, G= -0.114). Furthermore, the highest indirect effects at phenotypic and genotypic levels on seed weight/plant were detected in days to maturity (P=0.471, G=0.509) through head diam- eter followed by head diameter via plant height (P=0.400, G=0.426) and 100-seed weight (P=0.256, G=0.267). These results may be considered as identical to the previous results of correlation at both phenotypic and genotypic levels (Tab. 7 and Fig. 2). The above-mentioned results of direct and joint effects indicate that the direct selection of head diameter and 100-seed weight will be more effective in improving seed weight/plant. Days to maturity, head diameter, and 100-seed weight exhibited negative indirect effects at phenotypic and genotypic levels on seed weight/ plant through plant height. These results agreed with those of Abd EL-Satar et al. (2017). The residual effect was recorded with 0.579 and 0.556 at phenotypic and ge- notypic levels, respectively. These results indicate that the independent traits that are included at the phenotypic and genotypic path analysis explained 42.15% and 44.42% of the total variation, respectively, in seed weight/plant. The highest residual effects of phenotypic and genotypic path analyses indicate that the presence of other traits that are TABLE 7. Pooled phenotypic (P) and genotypic (G) path analysis of 32 sunflower backcrosses for seed weight/plant across two locations. Traits Days to maturity Plant height Head diameter 100-seed weight Correlation with seed weight/plant Days to maturity P 0.048 -0.098 0.471 0.135 0.556** G 0.020 -0.102 0.509 0.137 0.564** Plant height P 0.042 -0.112 0.399 0.098 0.427* G 0.018 -0.114 0.426 0.100 0.430* Head diameter P 0.033 -0.065 0.695 0.106 0.769** G 0.014 -0.066 0.732 0.106 0.786** 100-seed weight P 0.023 -0.038 0.256 0.288 0.529** G 0.010 -0.039 0.267 0.291 0.529** Residual P 0.578 G 0.556 *, ** Significant at 0.05 and 0.01 probability level, respectively. Highlighted values indicate a direct effect and normal values indicate an indirect effect. FIGURE 2. Pooled A) phenotypic and B) genotypic path diagram for seed weight/plant across the two locations. Re si du al e ffe ct = 0 .5 78 100-seed weight (P4) 0.288 Head diameter (P3) 0.695 Plan height (P2) -0.112 Days to maturity (P1) 0.048 0. 36 8 0. 87 4 0. 67 8 0. 57 5 0. 33 9 0. 46 6 B Re si du al e ffe ct = 0 .5 56 100-seed weight (P4) 0.291 Head diameter (P3) 0.732 Plan height (P2) -0.114 Days to maturity (P1) 0.020 0. 36 5 0. 89 9 0. 69 5 0. 58 3 0. 34 2 0. 47 1 A Phenotypic path diagram for seed weight/plant Genotypic path diagram for seed weight/plant 312 Agron. Colomb. 38(3) 2020 achieved through the selection of genotypes with a larger head diameter and heavier seed weight. Moreover, the ge- netic variations were observed among backcrosses can be used in the future programs for sunf lower improvement. Literature cited Abd EL-Satar, M.A. 2017. Genetic analysis of half diallel matting with different methods and their comparisons for yield and its associated traits in sunf lower under saline soil stress con- ditions. Helia 40(66), 85-114. Doi: 10.1515/helia-2017-0001 Abd EL-Satar, M.A., A.A.E.H. Ahmed, and T.H.A. Hassan. 2017. Response of seed yield and fatty acid compositions for some sunf lower genotypes to plant spacing and nitrogen fertil- ization. Inf. Process. Agric. 4(3), 241-252. Doi: 10.1016/j. inpa.2017.05.003 Abd EL-Satar, M.A., R.M. Fahmy, and T.H.A. Hassan. 2015. Genetic control of sunf lower seed yield and its components under dif- ferent edaphic and climate conditions. The 9th Plant Breeding International Conference, September 2015. Egypt. J. Plant Breed. 19(5), 103-123. Abdelsatar, M.A., E.M.M. Elnenny, and T.H.A. Hassan. 2020. Inheri- tance of seed yield and yield-related traits in sunf lower. J. Crop Improv. 34(3), 378-396. Doi: 10.1080/15427528.2020.1723767 Acquaah, G. 2012. Principles of plant genetics and breeding. 2nd ed. Wiley-Blackwell, Oxford, UK. Ahmed, M.A., M.A. Abdelsatar, M.A. Attia, and A.A. Abeer. 2019. GGE biplot analysis of line by tester for seed yield and its attri- butes in sunf lower. RUDN J. Agron. Anim. Ind. 14(4), 374-389. Doi: 10.22363/2312-797X-2019-14-4-374-389 AOAC. 1990. Official Methods of Analysis. 15th ed. Association of official analytical chemists, Virginia, USA. Azad, K., G. Shabbir, M.A. Khan, T. Mahmood, Z.H. Shah, F. Al- ghabari, and I. Daur. 2016. Combining ability analysis and gene action studies of different quantitative traits in sunf lower by line x tester. Crop Res. 51(2). Chandra, B.S., S. Kumar, A.R.G. Ranganadha, and M.Y. Dudhe. 2011. Combining ability studies for development of new hybrids over environments in sunf lower (Helianthus annuus L.). J. Agric. Sci. 3(2), 230-237. Doi: 10.5539/jas.v3n2p230 Comstock, R.E. and H.F. Robinson. 1948. The components of genetic variance in populations of biparental progenies and their use in estimating average degree of dominance. Biometrics 4(4), 254-266. Dewey, D.R. and R.H. Lu. 1959. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agron. J. 51(9), 515-518. Doi: 10.2134/agronj1959.0002196200 5100090002x Gvozdenović, S., J. Joksimović, and D. Škorić. 2005. Gene effect and combining abilities for plant height and head diameter in Sun- f lower. Genetica 37(1), 57-64. Doi: 10.2298/GENSR0501057G Jackson, M.L. 1973. Soil chemical analysis. Prentice Hall of India Private Limited. New Delhi. Karasu, A., M. Oz., M. Sincik, A.T. Goksoy, and Z.M. Turan. 2010. Combining ability and heterosis for yield and yield components not included in the present study is associated with the highest effect on seed weight/plant. Genetic divergence Thirty-two sunf lower backcrosses were grouped into eight distinct clusters/groups by canonical analysis, as shown in Supplementary material 4 and Figure 3. Data indicated that the evaluated sunf lower backcrosses were widely divergent, so they were scattered into eight clusters. The first cluster was the largest with 14 backcrosses, followed by the fourth cluster with six and the second with four backcrosses. The third and fifth clusters had two backcrosses and the seventh and eighth clusters had one backcross each. Supplementary material 5 shows a comparison of cluster means on the average of both locations. The shortest period to physiological maturity (78.67 d) along with the lowest values of plant height (151.79 cm), head diameter (14.90 cm), seed yield ha-1 (1899.31 kg ha-1), seed oil content (33.07%), oleic acid (26.42%) and linoleic acid (34.44%) were spotted in the third cluster, whereas the longest period to physiological maturity (101.0 d) and the highest values of plant height (181.84 cm), head diameter (27.73 cm), 100-seed weight (8.40 g), seed weight/plant (42.21 g), seed yield ha-1 (2810.79 kgha-1), seed oil content (42.67%), oleic acid (30.56%) and linoleic acid (54.95%) were found in the seventh cluster. FIGURE 3. Scatter diagram of 32 sunflower backcrosses across two lo- cations based on their canonical vectors superimposed with clustering. Code of figures as in Supplementary material 4. III VIII VI V IV VIIIII9 7 2813 6 24 18 1726 110 22 3129 11 5 2725 3012 382 23 15 14 4 32 19 21 2016 80.0070.0060.0050.00 15.00 10.00 5.00 0.00 -5.00 -10.00 20.00 Z1 Z2 Conclusion Based on the results mentioned above, it can be concluded that selection of all the studied traits at both locations in early segregating generations to accumulate additive genes would produce superior inbred lines to be used in hybrid breeding programs. The genetic improvement may be 313Abdelsatar and Hassan: Gene action and heritability in bi-parental crosses of sunflower in sunf lower. Not. Bot. Hort. Agrobot. Cluj-Napoca 38(3), 260-264. Lakshman, S.S., N.R. Chakrabarty, and P.C. Kole. 2019. Study on the combining ability and gene action in sunf lower through line x tester matting design. Electron. J. Plant Breed. 10(2), 816-826. Doi:10.5958/0975-928X.2019.00109.1 Mahalanobis, P.C. 1936. On the generalized distance in statistics. Proc. Natl. Inst. Sci. India. 2, 49-55. Ortis, L., G. Nestares, E. Frutos, and N. Machado. 2005. Combining ability analysis for agronomic traits in sunf lower (Helianthus annuus L.). Helia 28(43), 125-134. Parameshwarappa, K.G., J. Ram, and B.S. Lingaraju. 2008. Heterosis and combining ability for seed yield, oil content and other agronomic traits involving mutant restorer lines in sunf lower (Helianthus annuus L.). J. Oilseeds Res. 25(1), 8-12. Ram, J.J., U.K. Singh, S.K. Singh, and B. Krishna. 2018. Study of genetic diversity in Sunf lower (Helianthus annuus L.). Int. J. Curr. Microbiol. App. Sci. 7(5), 2266-2272. Doi: 10.20546/ ijcmas.2018.705.263 Rao, C.R. 1952. Advanced statistical methods in biometrical re- search. John Wiley and Sons, New York, USA Shankar, V.G., M. Ganesh, A.R.G. Ranganatha, A. Suman, and V. Shridhar. 2007. Combining ability studies in diverse CMS sources in sunf lower (Helianthus annuus L.). Ind. J. Agric. Res. 41(3), 171-176. Shinde, S.R., R.B. Sapkale, and R.M. Pawar. 2016. Combining ability analysis for yield and its components in sunf lower (Helianthus annuus L.). Int. J. Agric. Sci., 12(1), 51-55. Thirumala, R.V., A.R.G. Ranganatha, M. Ganesh, K. Srinivasulu, and P.V. Rao. 2005. Assessment of genetic divergence in gene pool inbreds and elite lines of sunf lower (Helianthus annuus L.) J. Oilseeds Res. 22(1), 168-171. Vidhyavathi, R., P. Mahalakshmi, N. Manivannan, and V. Muru- lidharan. 2005. Correlation and path analysis in sunf lower (Helianthus annuus L.). Agric. Sci. Digest 25(1), 6-10. Weber, C.R. and B.R. Moorthy. 1952. Heritable and nonheritable relationship and variability of oil content and agronomic traits in the F2 generations of soybean crosses. Agron. J. 44(4), 202-209. Doi: 10.2134/agronj1952.00021962004400040010x Zygadlo, J.A., R.E. Morere, R.E. Abburra, and C.A. Guzman. 1994. Fatty acids composition in seed oils of some Onagraceae. J. Am. Oil Chem. Soc. 71(8), 915-916. SUPPLEMENTARY MATERIAL 1. Mean performance of days to maturity, plant height and head diameter for bi-parental sunflower crosses at Kafr-El- Hamam (K) and Al-Arish (A) in the 2019 season. Bi-parental crosses Days to maturity (d) Plant height (cm) Head diameter (cm) K A K A K A P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 M1/Set 1 87.33 82.00 79.00 76.33 163.40 155.30 156.47 151.33 18.70 15.30 14.10 11.80 M2/Set 1 93.33 87.33 82.67 79.00 165.63 163.80 158.27 156.60 22.92 19.73 16.07 14.10 M3/Set 1 92.00 83.00 81.67 76.33 165.20 155.30 156.97 151.33 22.58 16.60 15.73 12.90 M4/Set 1 97.33 92.33 87.33 82.33 179.91 165.58 163.53 158.27 26.42 22.91 18.67 16.20 M1/Set 2 80.67 83.67 75.67 77.33 153.20 183.24 147.33 167.97 14.50 17.83 11.21 13.17 M2/Set 2 85.33 93.33 78.00 83.33 161.30 168.20 154.23 158.20 17.83 23.40 13.83 16.53 M3/Set 2 83.67 89.33 76.33 80.00 155.30 164.50 151.33 156.60 17.03 20.77 13.10 14.87 M4/Set 2 90.33 97.33 80.33 86.67 164.80 181.17 156.60 163.30 22.00 25.80 15.07 18.30 M1/Set 3 88.67 96.00 79.67 85.33 164.50 174.00 156.60 161.73 24.33 23.97 16.90 17.47 M2/Set 3 96.00 101.67 85.33 90.67 174.13 184.01 161.23 170.77 24.67 27.28 17.63 19.43 M3/Set 3 94.33 92.33 83.67 83.00 168.20 165.87 159.73 157.67 23.60 23.51 16.20 16.10 M4/Set 3 97.00 109.67 86.67 92.33 178.94 186.87 164.90 176.80 25.10 27.73 18.30 20.17 M1/Set 4 100.00 85.33 89.33 78.00 183.24 161.27 167.97 152.63 27.07 18.03 19.23 13.17 M2/Set 4 86.00 91.00 78.00 81.33 159.53 165.00 153.93 157.67 17.40 21.68 13.10 15.43 M3/Set 4 100.00 89.00 89.33 80.00 183.24 164.53 167.97 156.97 27.07 20.27 19.23 14.77 M4/Set 4 109.67 95.00 93.33 84.67 188.43 174.00 181.67 161.13 14.50 23.97 11.80 16.53 LSD 5% 2.83 1.79 4.49 3.96 1.61 1.30 314 Agron. Colomb. 38(3) 2020 SUPPLEMENTARY MATERIAL 2. Mean performance of 100-seed weight, seed weight/plant and seed yield ha-1 for bi-parental sunflower crosses at Kafr-El-Hamam (K) and Al-Arish (A) in the 2019 season. Bi-parental crosses 100-Seed weight Seed weight/plant Seed yield ha-1 K A K A K A P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 M1/Set 1 6.37 6.61 4.24 3.41 28.23 27.48 23.90 18.97 2067.38 1884.31 1603.10 1335.00 M2/Set 1 7.28 6.38 4.84 4.24 34.94 34.02 28.13 23.90 2131.04 2068.97 1775.71 1603.10 M3/Set 1 7.28 5.21 4.80 3.41 35.52 27.97 28.77 18.97 2131.04 1942.72 1732.46 1509.68 M4/Set 1 7.98 7.28 5.27 4.81 38.52 35.95 31.63 28.83 2561.11 2113.65 1992.62 1746.59 M1/Set 2 8.56 6.28 3.41 4.20 25.77 28.37 18.63 20.40 1914.32 2033.02 1335.00 1508.65 M2/Set 2 6.35 7.36 4.20 4.95 25.80 33.12 22.03 29.47 2033.81 2188.02 1521.51 1783.02 M3/Set 2 8.17 6.57 5.37 4.51 27.97 34.52 19.40 26.00 1981.91 2084.36 1501.43 1656.90 M4/Set 2 7.01 7.62 4.58 5.29 35.63 38.49 27.00 31.43 2113.81 2425.11 1709.60 1901.75 M1/Set 3 6.65 7.55 4.51 5.15 27.73 37.48 25.10 30.03 2083.49 2255.48 1656.90 1830.16 M2/Set 3 6.50 6.05 4.42 4.78 28.56 25.62 23.80 20.47 2076.35 2610.40 1639.60 2056.11 M3/Set 3 7.32 7.33 4.84 4.83 37.01 36.51 28.87 28.83 2176.20 2158.81 1799.92 1766.51 M4/Set 3 7.63 8.40 6.26 6.25 38.45 42.21 30.73 35.40 2397.38 2810.79 1874.52 2126.35 M1/Set 4 5.57 7.53 3.97 3.92 39.37 30.43 32.27 21.20 2583.10 2041.83 1992.62 1525.56 M2/Set 4 6.05 6.90 3.95 4.73 24.81 34.95 21.47 27.83 2007.82 2093.44 1535.56 1725.00 M3/Set 4 6.19 7.45 5.30 4.57 39.37 34.52 32.27 25.50 2583.10 2092.57 1992.62 1693.41 M4/Set 4 7.37 8.13 5.22 4.97 27.10 37.28 18.63 29.33 1919.68 2247.94 1341.51 1814.52 LSD 5% 0.34 0.29 1.59 2.36 89.32 87.51 LSD 5%: Least significant difference at α=0.05. SUPPLEMENTARY MATERIAL 3. Mean performance of seed oil content, oleic acid content and linoleic acid content for bi-parental sunflower crosses at Kafr-El-Hamam (K) and Al-Arish (A) in the 2019 season. Bi-parental crosses Seed oil content Oleic acid content Linoleic acid content K A K A K A P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 M1/Set 1 37.84 33.43 42.51 39.17 27.72 26.55 29.35 27.18 44.13 34.42 47.29 42.93 M2/Set 1 40.70 38.05 43.15 42.51 28.45 27.82 30.30 29.35 47.67 44.61 50.86 47.63 M3/Set 1 39.82 36.21 42.78 41.31 28.39 26.55 30.01 27.18 47.11 38.80 48.96 42.93 M4/Set 1 41.82 40.16 44.21 42.97 29.34 28.45 31.14 30.23 43.89 47.44 46.89 49.35 M1/Set 2 32.71 33.35 36.58 41.95 26.29 27.35 25.15 28.11 34.45 37.42 41.11 45.71 M2/Set 2 37.40 40.76 42.22 43.19 27.38 28.22 28.73 29.98 43.41 47.00 46.56 49.78 M3/Set 2 37.12 39.07 41.61 42.51 27.15 28.12 28.07 29.46 42.07 45.93 44.80 48.35 M4/Set 2 40.60 41.64 43.52 44.41 28.18 29.34 29.70 31.35 46.41 53.06 48.49 55.05 M1/Set 3 39.07 41.52 42.51 44.11 28.12 28.95 29.45 31.03 45.70 52.18 48.08 53.70 M2/Set 3 38.49 42.67 42.51 44.41 27.95 29.71 29.41 32.06 45.15 54.95 47.75 56.54 M3/Set 3 41.34 40.16 43.43 43.11 28.49 28.45 30.69 30.51 49.73 48.18 51.78 50.30 M4/Set 3 41.85 42.67 44.41 44.67 29.71 30.56 31.55 32.89 53.06 54.95 54.96 57.76 M1/Set 4 42.47 37.35 44.41 42.26 29.71 27.45 31.93 28.72 42.66 43.63 45.71 46.44 M2/Set 4 37.35 39.94 42.37 42.73 27.59 28.35 28.70 29.82 42.91 46.59 46.40 49.26 M3/Set 4 42.47 39.20 44.41 42.54 29.71 28.01 31.93 29.62 54.26 45.73 55.54 47.86 M4/Set 4 35.34 41.43 38.28 43.80 26.29 28.59 25.92 30.75 38.32 51.82 41.92 53.36 LSD 5% 0.98 1.03 0.52 0.92 1.61 1.73 LSD 5%: Least significant difference at α=0.05. 315Abdelsatar and Hassan: Gene action and heritability in bi-parental crosses of sunflower SUPPLEMENTARY MATERIAL 4. Grouping of 32 sunflower backcrosses in clusters based on the canonical analysis for all the studied traits across two locations. Cluster No. of backcrosses Backcrosses I 14 2 (M2 x P1/Set1), 3 (M3 x P1/Set1), 4 (M4 x P1/Set1), 8 (M4 x P2/Set1),12 (M4 x P1/Set2),14 (M2 x P2/Set2), 15 (M3 x P2/Set2),16 (M4 x P2/Set2), 19 (M3 x P1/Set3), 20 (M4 x P1/Set3), 21 (M1 x P2/Set3), 23 (M3 x P2/Set3),30 (M2 x P2/Set4), 32 (M4 x P2/Set4). II 4 6 (M2 x P2/Set1), 11 (M3 x P1/Set2), 29 (M1 x P2/Set4), 31 (M3 x P2/Set4). III 2 5 (M1 x P2/Set1), 9 (M1 x P1/Set2). IV 6 1 (M1 x P1/Set1), 7 (M3 x P2/Set1), 10 (M2 x P1/Set2), 13 (M1 x P2/Set2), 26 (M2 x P1/Set4), 28 (M4 x P1/Set4). V 2 17 (M1 x P1/Set3), 18 (M2 x P1/Set3). VI 2 25 (M1 x P1/Set4), 27 (M3 x P1/Set4) VII 1 24 (M4 x P2/Set3). VIII 1 22 (M2 x P2/Set3). SUPPLEMENTARY MATERIAL 5. Mean value of nine quantitative traits of the eight clusters for 32 bi-parental crosses across two locations. Cluster DM PH HD 100 SW SWP SYH OC OA LA I 88.57 164.84 23.47 7.38 36.31 2219.81 40.77 28.64 48.58 II 82.33 157.80 18.77 7.38 31.74 2046.31 37.93 27.61 44.01 III 78.67 151.79 14.90 7.59 26.62 1899.31 33.07 26.42 34.44 IV 84.75 164.73 17.14 6.27 27.05 2000.74 36.25 27.15 40.883 V 87.42 164.12 24.50 6.58 28.15 2079.93 38.78 28.04 45.43 VI 94.67 175.60 27.07 5.88 39.37 2583.10 42.47 29.71 48.46 VII 101.00 181.84 27.73 8.40 42.21 2810.79 42.67 30.56 54.95 VIII 96.17 177.39 27.28 6.05 25.62 2610.40 42.67 29.71 54.95 DM: Days to maturity, PH: Plant height, HD: Head diameter, 100 SW: 100-seed weight, SWP: Seed weight/plant, SYH: Seed yield ha-1, OC: Seed oil content, OA: Oleic acid and LA: Linoleic acid.