النشر العلمي

  • Combining ability for seed yield and yield components in sunflower (Helianthus annuus L.)

 

Combining ability for seed yield and yield components

in sunflower (Helianthus annuus L.)

 

Mohamed Y. Mohamed1, Ibrahim N. Elzein1, Mohamed E. Ahmed1 and

Abu Elhassan S. Ibrahim2

 

1 Agricultural Research Corporation, Wad Medani, Sudan.

2 Faculty of Agricultural Sciences, University of Gezira, Wad Medani, Sudan.

ABSTRACT

 

   This study was conducted using four cytoplasmic male sterile lines (cms) and ten restorers of sunflower (Helianthus annuus L.) in a line x tester fashions in summer and winter season of 2007at Sennar Research Station Farm, Agricultural Research Corporation, Sudan. The objectives were to determine the general combining ability and specific combining ability effects among 14 parents and their 40 crosses for seed yield and other agronomic traits. The experiment was carried out under supplementary irrigation using a randomized complete block design (RCBD) with three replicates. The traits studied were days to 50% flowering, plant height (cm), head diameter(cm), number of seeds per head, percentage of empty seeds, 1000-seed weight (g), and seed yield (kg/ha). Analysis of variance of the combining abilities revealed highly significant differences for general combining ability (GCA) and specific combining ability (SCA). The highest contribution to the total variance was expressed by SCA for most important traits. The non-additive gene effects were found to be important for the inheritance of all traits except for plant height which was controlled by additive gene effect in the base material. The cms lines SA3 and SA4 and restorers SR41, SR45, SR10 and SR13 were better general combiners for most of the traits including seed yield across two seasons. The best combining hybrids for seed yield in the combined analysis were SA4 x SR1, SA3 x SR41, SA4 x SR45, SA3 x SR13 and SA3 x SR10. The greatest average contribution to the expression of most of the traits was found in the line x tester interaction, while the contribution of the female (A-line) and the male (R-line) were less significant. The simple correlation analysis revealed that seed yield was positively and significantly correlated with 1000-seed weight, head diameter and number of seeds per head and negatively correlated with percentage of empty seeds. Thus, the highly significant SCA effects in the superior crosses could be utilized in heterosis breeding to develop of high yielding local single-cross hybrids or produce synthetic composite varieties.

 

INTRODUCTION

 

   Sunflower (Helianthus annuus L.) is one of the most important oilseed crops in the Sudan and the world. The development of hybrids with good agronomic performance is an important target in sunflower breeding programs. Therefore, hybrid sunflower became a reality with the discovery of cytoplasmic male sterility and effective male fertility restoration system during 1970. Hybrid vigor has been the main driving force for acceptance of this oilseed crop. In the Sudan, sunflower producers depended almost exclusively on imported seeds and the value of local hybrid and the importance of heterosis breeding were not sufficiently recognized. Nevertheless, one of the long term objectives of sunflower research program at the Agricultural Research Corporation (ARC) is to develop new single-cross sunflower hybrids characterized by uniform plant height, flowering date and seed quality. In addition, the hybrids are more stable, highly responsive to high-input agriculture and highly self-fertile, resulting in higher seedset in areas where pollinators are not abundant. Thus, the development of sunflower hybrids for Sudanese conditions is an important step towards narrowing down the gap between supply and demand in the seed market and boosting sunflower production and productivity in the country. This will also cut down the time and resources being spent on importations from abroad. Sunflower hybrid seed produced locally is likely to be adopted by the majority of sunflower growers, since the seed source is readily available (Mohamed, 2010).

   Sunflower, as a cross-pollinated crop, provided an opportunity for developing new and superior hybrids through the use of breeding for heterosis. In heterosis breeding, the selection of parents/inbreds with good combining ability is very important in producing superior hybrids. The estimation of general combining ability (GCA) and specific combining ability (SCA) helps in identifying the potential parents/inbreds in the production of superior hybrids for seed yield, oil content and oil quality. Hence, estimation of combining ability is a pre-requisite in sunflower breeding where it aimed at the development of hybrids or improvement of lines.

 

 

 

 

 The line x tester analysis(Kempthorne, 1957)is one of the simplest and efficient methods of evaluating large number of inbreds for combining ability and per se performance. Analysis of GCA and SCA is useful in knowing the type of gene action controlling various characters and development of suitable breeding strategies. Therefore, the objectives of this study were to determine the general combining ability and specific combining ability effects among parents and their crosses for seed yield and other agronomic traits, as well as components of genetic variance and to find out the best testers for testing F1 hybrid combinations for important yield traitsas an attempt to develop high yielding local single-cross hybrids or produce synthetic composite varieties.

 

MATERIALS AND METHODS

 

    The initial breeding material of cytoplasmic male sterile lines and their counterparts were introduced from France and Russia, respectively. The full-sib progeny method was used in both the female and maintainer lines (B-lines) for monitoring purity from pollen shedders and branching types. Plants selected as pairs of A and B-lines were characterized by uniformity in height, flowering period and emergence date. The four selected female parents (A-lines) and their maintainers (B-lines) were maintained and increased for further crossing and redesigned as SA1, SA2, SA3 and SA4, respectively. Ten R-lines (male or restorer lines) were developed from finished F1-hybrids introduced by the program. Thus, theten male lines (R-lines SR1, SR2, SR3, SR6, SR7, SR10, SR 13, SR14, SR41 and SR 45) used for this study were characterized by uniformity in flowering, plant height, maturity and high degree of self-fertility and classified as branching types except SR 41, which is non-branching. The experiment was carried out during 2005-2007 at Sennar Research Station Experimental Farm in the clay plains of central Sudan (13ْ 33́ N, 33ْ 34̀ E and 421masl). The soil is a vertisol with 60% clay content, PH of 7.8-8.5, about 0.4-0.5% organic carbon and 0.05% total nitrogen. The region has semi-arid climate with summer rainfall ranging from 300 to 600 mm. The average rainfall, temperature and relative humidity at Sennar from May 2007 to February 2008 are presented in Table 1.

 

 

Table 1. Meteorological data for summer and winter season of 2007-2008 at Sennar Research Farm of ARC, Sudan.

Month

Mean maximum temperature (Cْ )

Mean minimum temperature (Cْ )

Mean relative humidity (%)

Total rainfall (mm)

April 2007

41.7

21.7

23

12.8

May 2007

42.0

26.6

35

Nil

June 2007

37.9

24.7

66

115.8

July 2007

32.6

22.6

86

286.6

August 2007

32.1

22.7

85

266.7

September 2007

33.6

21.9

81

87.0

October 2007

37.6

22.3

68

17.9

November 2007

37.7

19.0

46

Nil

December 2007

35.3

16.3

50

Nil

January 2008

32.5

14.5

47

Nil

February 2008

34.6

17.0

49

Nil

Source:  Sennar Meteorological Station, Sennar-Sudan.

 

   The four cytoplasmic male sterile lines were crossed with the ten restorer lines (testers) in a field experiment in a line x tester mating design during 2005 and 2006 giving 40 F1 hybrids. Hand pollination was used to develop the breeding material. Pollen grains from the male parent (R-line) were collected in toclassic Petri dish and then dusted on the stigmas of the florets of female parents (A-line) using cotton buds and covering with respective butter cotton bag. The pollinations were repeated 4-6 times on alternate days, to ensure the pollination of all florets in the three to five heads per each cross. These 40 F1 hybrids (line x tester) along with 14 parents (including 4 lines and 10 testers) were tested at Sennar for summer and winter of 2007.

   The materials were tested in a randomized complete block design with three replicates on the 15thof July for summer season and on the 13thof November for winter season. Three seeds per hill were sown to ensure uniform stand which was later thinned to one plant per hill. The plot size consisted of four rows per plot, each of 5 m in length with row to row and plant to plant distances of 0.80 m and 0.30 m, respectively. Nitrogen was applied at 80 kg urea per hectare.  Irrigation was applied at intervals of 12-14 days depending on weather conditions. Hand weeding was carried out to keep the crop weed free. The harvest was done manually during the first and

 

second week of December and March for summer and winter seasons, respectively. Data were recorded on ten randomly selected plants from the middle inner two rows in each plot from each replication for the following traits; days to 50% flowering, plant height (cm), head diameter (cm), number of seeds per head, percentage of empty seeds, 1000-seed weight (g), and seed yield (kg/ha).

   The analysis of variance procedure was carried out separately, for each season and then combined, using the IRRISTAT statistical analysis package for windows (2006). Analysis of variance for combining ability was done according to the line x tester analysis method, in which estimates of GCA variances (б2GCA) and SCA variances (б2SCA) were obtained as suggested by Singh and Chaudhary(1985). The significance of GCA and SCA effects were determined at the 0.05 and 0.01 levels using the t-test.

 

RESULTS AND DISCUSSION

 

Combining ability analysis

   The combining ability analysis is an indication of the variances due to GCA and SCA, which represent a relative measure of the additive and non-additive gene action, respectively. The variance components due to GCA and SCA are used to derive conclusions regarding the gene action that is prevalent in determining any trait. Hence, the ANOVA for combining ability (combined over two seasons) from a line x tester design involving four females (lines) and ten males (testers), partitioning the variation due to hybrids (crosses) into three components namely females, males and their interactions is given in Table 2. However, the results revealed that the variance due to females (lines) was highly significant for all tested traits except for percentage of empty seeds. The mean squares due to males (testers) were also highly significant only for days to 50% flowering, plant height, head diameter, one thousand seed weight and seed yield. The variation due to female x male (line x tester) interactions was found significant for all the seven traits studied.

 

 

 

 

 

Table 2. Analysis of variance for combining ability for seven characters of 4 females, 10 males and 40 F1- hybrids evaluated over two seasons of 2007at Sennar Research Station Farm.

Source/ trait

DF

PH

HD

NSH

ES

SW

SY

Crosses

36.86**

656.0**

6.53**

9280.58**

13.1**

126.1**

47472.2**

Females (F)

34.76**

721.3**

6.82**

18017.4**

13.1ns

139.4**

52745.8**

Males (M)

99.67**

4586.3**

26.6**

1542.1ns

2.72ns

541.9**

206182 **

F x M

30.58**

197.53*

4.20**

7228.1**

14.2**

75.39**

28079.9**

Error

1.40

88.71

0.38

3080.74

0.37

6.65

5796.75

Where, DF = days to 50% flowering, PH = plant height (cm), HD = head diameter (cm), NSH = number of seeds per head, ES = percentage of empty seeds, SW = 1000-seed weight (g), and SY = seed yield (kg/ha).

*,** Significant at 0.05 and 0.01 levels of probability, respectively, and  ns = non-significant.

 

Table 3. Estimates of variance due to GCA, variance due to SCA, additive variance (VA), dominance variance (VD), ratio of GCA to SCA  and degree of dominance (VA/VD) of seven sunflower traits combined over two seasons of 2007at Sennar Research Station Farm.

Genetic component

DF

 

PH

 

HD

 

NSH

 

ES

 

SW

 

SY

 

GCA

1.74

116.97

0.60

121.51

-0.02

12.63

4827.81

      F=0,VA

3.49

467.8

2.39

486.03

-0.07

50.53

19311.2

      F=1,VA

6.98

635.7

4.78

972.06

-0.14

101.0

38622.3

SCA

9.73

36.27

1.28

1382.46

4.64

22.92

7427.73

      F=0,VD

9.73

145.0

5.10

5529.8

18.5

91.66

2971.9

      F=1,VD

38.92

580.36

20.40

22119.3

74.28

366.64

11887.6

VD/ VA         F=0

1.67

0.56

1.46

4.77

0.00

1.90

0.55

VD/ VA         F=1

2.36

0.79

2.07

3.37

0.00

1.35

0.39

GCA/SCA

0.18

3.22

0.47

0.09

0.00

0.55

0.65

Where: DF= Days to 50% flowering, PH = Plant height (cm), HD = Head diameter (cm), NSH = No. of seeds per head, ES = Percentage of empty seeds, SW = 1000- seed weight (g), & SY = Seed yield (kg/ha).

 

 

 

 

    On the other hand, the mean square for GCA (б2gca) was highly significant for plant height, indicating the importance of additive gene effect for this trait. The components of variance showed that SCA variance (б2sca) was higher than GCA variance (б2gca) for all traits studied except plant height, indicating the importance of non-additive gene effect for these traits. Also, as the variances due to SCA were highly significant for number of seeds per head and seed yield, these characteristics were influenced by dominant gene actions. Furthermore, the ratio of GCA to SCA variance (б2gca/б2sca) for all traits, except plant height, were less than one, indicating that the inheritance of these traits was due to the non-additive gene action. It revealed that dominance and epistasis played major roles in the inheritance of these traits. It also revealed the possibility of hybrid breeding for these traits. The estimates of non-additive gene actions for all the characters(except plant height) in this study were generally in agreement with the  results reported by Gangappa et al.(1997), Shekar et al.(1998), Ashok et al.(2000), and Goksoy et al.(2004).

General and specific combining ability effects

   The GCA and SCA effects of the parents and hybrids derived for assaying test genotypes in an L x T analysis for seven sunflower traits combined over two seasons is given in Tables 4 and 5, respectively. For days to 50% flowering, favorable GCA effects were shown by parents that manifested earliness by having negative GCA estimates for flowering time. In combined analysis, the favorable GCA effects in the negative direction were obtained by SR41 followed by SA4 and SR10 (good combiners for earliness), while the unfavorable GCA effect was obtained by SR7 (good combiner for late flowering) in a positive direction (Table 4). There is a variation among the crosses for earliness due to variation in SCA. Favorable negative and significant SCA effect was obtained by SA1 x SR1 followed by SA2 xSR10, SA2 x SR14 and SA3 x SR13. The latest flowering cross was SA2 x SR1 (Table 4). These results confirmed the findings of Shekar et al.(1998), Naik et al.(1999), Ashok et al. (2000) and Goksoy et al. (2004) who reported significant and negative SCA effects for early flowering. Also, as mentioned earlier, the greatest average contribution for this trait was 57% due to SCA in crosses, compared to 22% for lines and 21% for testers (Table 6).

   The ratio of GCA/SCA effects was less than unity, which confirmed that this trait was controlled by non-additive gene action (Table 2).

 

Table 4. Estimates of GCA effects of 14 sunflower parents (4 lines and 10 testers) for seven traits evaluated across two seasons of 2007at Sennar Research Station Farm.

Parent

DF

PH

HD

NSH

ES

SW

SY

SA1

0.97**

-9.87**

-1.29**

-8.95

0.45**

-5.52**

-94.75**

SA2

2.07**

15.51**

0.013

-2.49

-0.12

-0.55

-34.22*

SA3

-1.23**

4.41*

0.32**

6.69

-0.22*

4.63**

97.29**

SA4

-1.80**

-10.05**

0.95**

4.76

-0.12

1.44**

31.68

SR1

1.98**

16.10**

-0.35*

-17.04

0.24

-3.62**

-0.01

SR2

-1.35**

-7.02**

-0.05

11.63

1.06**

-2.43**

-37.51*

SR3

-0.85*

-10.10**

-0.98*

-31.57*

0.21

-2.39**

-79.98**

SR6

0.57*

3.65**

-0.31*

-44.47*

-0.47*

-3.19**

-78.11**

SR7

3.15**

-1.85

-0.24

-27.84

2.35**

-3.01**

-55.21*

SR10

-1.43**

7.73**

0.46*

57.73**

-0.40*

3.20**

65.06**

SR13

0.82*

-4.94**

-0.64*

44.28*

-1.20**

1.74**

36.09*

SR14

-0.52*

-3.52

-0.30

-10.75

-0.37*

0.55

-27.96

SR41

-2.52**

-2.94

0.85*

52.40**

-0.99*

6.18**

104.85**

SR45

0.15

2.90

1.57**

-34.38

-0.45*

2.97**

72.77**

SE+  line

0.48

3.85

0.25

22.66

0.25

1.05

31.08

SE+tester

0.31

2.43

0.16

14.33

0.16

0.67

19.66

Where, DF= days to 50% flowering, PH = plant height (cm), HD = head diameter (cm), NSH = number of seeds per head, ES = percentage of empty seeds, SW = 1000- seed weight (g), and SY = seed yield (kg/ha)

*,** Significant at 0.05 and 0.01 levels of probability, respectively, and ns = non-significant

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 5. Estimates of SCA effects of 40 sunflower hybrids (crosses) for seven traits evaluated across summer and winter seasons of 2007 at Sennar Research Station Farm.

Hybrids

DF

PH

HD

NSH

ES

SW

SY

SA1 x SR1

-5.38**

-2.75

-0.54*

27.71

2.74**

0.07

-18.88

SA1 x SR2

0.28

-3.46

-1.44**

-26.64

1.32**

-1.60

5.39

SA1 x SR3

0.12

-2.38

0.39

-18.97

-1.79**

1.57

30.32

SA1 x SR6

2.03**

2.54

0.02

-8.84

0.39

1.93

33.88

SA1 x SR7

-2.88**

-4.30

1.51**

82.81*

-0.47

-1.45

18.62

SA1 x SR10

0.70

-2.21

0.02

58.20*

-0.90*

0.32

59.12

SA1 x SR13

0.78

11.12*

2.12**

-33.65

0.06

5.52**

-5.78

SA1 x SR14

-0.22

-1.63

-0.76*

16.45

-1.28**

-5.28**

-32.13

SA1 x SR41

2.45**

3.79

-0.57*

42.46

-2.82**

3.08*

-19.88

SA1 x SR45

2.12**

-0.71

-0.76*

-23.18

2.75**

-4.16*

-70.66*

SA2 x SR1

9.85**

4.36

0.33

21.55

0.24

-9.87**

-107.3*

SA2 x SR2

-0.15

-4.18

-0.54*

19.64

2.44**

3.69*

64.06*

SA2 x SR3

-0.65

-0.43

0.12

43.14

-0.80*

-0.54

17.86

SA2 x SR6

-0.73

14.49*

0.69*

-37.69

-0.73*

0.62

-50.81

SA2 x SR7

3.68**

-2.35

-1.62**

-51.18*

1.44**

4.32**

42.59

SA2 x SR10

-3.73**

-6.60

-0.31

-1.52

-1.90**

-6.04**

-78.74*

SA2 x SR13

-0.98*

3.07

-1.05**

45.96*

-0.89*

0.14

4.69

SA2 x SR14

-3.65**

5.32

1.88**

31.33

1.02**

7.14**

193.5**

SA2 x SR41

-2.32**

-3.60

0.56*

-21.62

-1.31**

0.80

-57.04

SA2 x SR45

-1.32*

-10.10*

-0.06

-49.61*

0.49

-0.26

-28.85

 

 

 

Table 5. Continued.

SA3 x SR1

-1.18*

-5.54

-0.95*

-75.43*

-2.00**

1.70

-71.06*

SA3 x SR2

0.15

8.25

2.39**

45.43*

-0.03

2.40*

50.25

SA3 x SR3

-0.68

3.34

0.01

-7.97

2.47**

-1.03

-75.72*

SA3 x SR6

0.57

-12.08*

-1.19**

-2.30

-1.59**

-4.71**

-20.39

SA3 x SR7

-1.68*

10.42*

0.74*

16.74

1.05**

2.79*

67.14*

SA3 x SR10

1.23*

2.84

-0.82*

1.76

2.82**

1.40

50.37

SA3 x SR13

-3.35**

-13.83*

0.74*

70.35*

-1.93**

3.70*

68.81*

SA3 x SR14

2.98**

-11.58*

-0.93*

-44.19

-1.34**

-4.83**

-172.8

SA3 x SR41

-0.35

2.50

0.12

-33.34

3.43**

-0.14

133.91**

SA3 x SR45

2.32**

15.67**

-0.10

28.94

-2.90**

-1.27

-30.47

SA4 x SR1

-3.28**

3.93

1.16**

26.17

-0.99*

8.10**

197.2**

SA4 x SR2

-0.28

-0.61

-0.41

-38.44

-3.73**

-4.49**

-119.7*

SA4 x SR3

1.22*

-0.53

-0.52*

-16.21

0.12

0.001

27.53

SA4 x SR6

-1.87*

-4.95

0.48

48.83*

1.93**

2.17*

37.32

SA4 x SR7

0.88

-3.78

-0.63*

-48.36*

-2.02**

-5.66**

-128.3**

SA4 x SR10

1.80*

5.97

1.12**

57.96*

-0.03

4.32**

-30.74

SA4 x SR13

3.55**

-0.36

-1.82*

-82.66*

2.75**

-9.36**

-67.71*

SA4 x SR14

0.88

7.89*

-0.19

-3.59

1.61**

2.97*

11.45

SA4 x SR41

0.22

-2.70

-0.11

12.49

0.70*

-3.74**

-57.00

SA4 x SR45

-3.12**

-4.86

0.91*

43.80

-0.34

5.69**

129.98**

SE+ SCA

0.97

7.69

0.50

45.32

0.50

2.11

62.17

Where, DF= days to 50% flowering, PH = plant height (cm), HD = head diameter (cm), NSH = number of seeds per head, ES = percentage of empty seeds, SW = 1000- seed weight (g), and SY = seed yield (kg/ha)

*,** Significant at 0.05 and 0.01 level of probability, respectively, and ns = non-significant

 

   Development of dwarf or medium plant height is the recent trend in breeding work to avoid lodging of sunflower hybrids. In combining analysis, the favorable negatively significant GCA effects of plant height for parents wereSR3 followed by SA4, SA1, SR2 and SR13 (Table 4). The variation among crosses for plant height and favorable negatively significant SCA effects obtained by SA3 x SR13 (-13.83) followed by SA3x SR6 (-12.08), SA3 x SR14 (-11.58), SA2 x SR45 (-10.10) and SA2 x SR10 (-6.60) as a best cross-combinations for the development of dwarf to moderate height sunflower hybrids. While, the unfavorable SCA effect (in positive direction) was observed by SA3 x SR45 as the tallest cross (Table 5). These results agreed with the results of Kumar et al.(1998), Gokosy et al.(1999),Naik et al.(1999) and Ashok et al.(2000)who reported negative SCA effects for plant height in sunflower.

 

   Head size is influenced greatly by environmental effects especially by plant population, soil moisture and soil fertility. Thus, the portion of total variation in head size attributable to genetic effects was often less than it was for certain other agronomic traits. Hence, for head diameter, positive values of GCA and SCA are desirable. Over two seasons, the favorable large head diameter having positive and significant GCA effects for parents were obtained by SR45, SA4, SR41, SR10 and SA3 (Table 4). The crosses that revealed positive and significant SCA effects were SA3x SR2 (2.39) followed by SA1xSR13 (2.12) and SA2xSR14 (1.88), whereas, the unfavorable negatively SCA effects hybrids were SA4 x SR13, SA2 x SR7 and SA1 x SR2 (Table 5). Also, from Table 2,SCA variances were higher than GCA variances for head diameter. This result provided an information that greater amount of genetic variability was due to SCA effects, indicating non-additive type of gene action being involved in this trait. Also, the greatest average contribution was obtained by SCA (45%) for crosses compared with GCA (24%) for lines and (32%) for testers (Table 6). This is desirable for heterosis breeding and can be exploited in hybrid seed production. These results are in accordance with the findings of Gangappa et al.(1997), and Kumar et al.(1998), who reported positive and significant SCA effects for head diameter in sunflower hybrids.

   Number of seeds per head has a direct relationship with seed yield and consequently positive values of GCA and SCA effects are desirable. In this study, the parents having positive and significant GCA effects for this trait in the two seasons and across them were SR10, SR41 and SR13      (Table 4). The maximum number of seeds per head and favorable positively significant SCA effects were obtained by SA1 x SR7 (82.81) followed by SA3 x SR13 (70.5)and SA1 x SR10 (58.20), and the minimum number of seeds per head of hybrids were SA4 x SR13(-82.66),SA3 x SR1(-75.43)and SA2 x SR7   (-51.18) (Table 5). Thus, in the two seasons and across them and generally, the crosses have SCA variances higher than GCA variances for number of seeds per head, and also the ratio of GCA/SCA effects was less than unity, indicating that this trait is controlled by non-additive gene action. Also, the proportional contribution of SCA for crosses have reached 54% compared with GCA 45% for lines and 1% for tester (Table 6). This is desirable for heterosis breeding needed in hybrid production. These findings are in line with those reported by Gangappa et al. (1997) and Kumar et al.(1998).

 

   In sunflower, disk flowers in the center of the head fail to produce seeds, a phenomena occurring in both open pollinated varieties as well as hybrids. It varies in extent and severity depending on genotypes and environment. In this case, negative GCA and SCA effects are desirable for achieving low percent empty seed in genotypes. Based on estimates of GCA effects for parents in the combining analysis (Table 4), the best combiner parents for lower empty seeds per head was SR13 followed by SR41 and SR6 and the poor combiner for this trait was SR7 followed by SR2 and SA1. With reference to crosses the best cross-combinations that had negative and significant SCA effects were SA4 x SR2 (-3.73), SA3 x SR45 (-2.90), SA1 x SR41 (-2.82), SA4 x SR7 (-2.02) and SA3 x SR1 (-2.00). The crosses that had positive and significant SCA effects were SA3 x SR41 (3.43) and SA3 x SR10 (2.82) (Table 5). Furthermore, the greatest average contribution for this trait in combined analysis was obtained by SCA (75%) compared with GCA    (23% for lines and 2% for testers). Also, the contributions of hybrids               (SCA variances for crosses) were greater than their parents (GCA for lines and testers) and the ratio of GCA/SCA was zero which was less than unity. This indicates that this trait is predominantly controlled by non-additive gene action (Table 2).

    For one thousand seed weight, positive values of GCA and SCA are desirable. In this study the best combining parents having significant positive GCA were SR41, SA3, SR10 and SR45 in the combined analysis. Also, in both seasons and their combined, the unfavorable GCA effects parent was shown by parent SA1 followed by SR1 (Table 4). Among the crosses, the best cross-combinations with positive and significant SCA effects were SA4 x SR1 (8.10) followed by SA2 x SR14 (7.14), SA4 x SR45(5.69) and SA1 xSR13 (5.52), so the negative cross was SA2 x SR1         (-9.87) followed by SA4 x SR13(-9.36) and SA2 x SR10 (-6.04) (Table 5). On the other hand, the SCA variances were higher than GCA variances for 1000-seed weight. Also, the ratio of GCA/SCA was less than unity, which provide information that greater amount of genetic variability, was due to SCA effects. These results indicate non-additive type of gene action being involved for controlling this character. Also, the average contribution in their combined analysis showed higher SCA (41%) than GCA (26% for lines

and 33% for testers) (Table 6). However, these results confirmed those of Ashok et al., (2000)and Goksoy et al. (2004).

 

Proportional contribution

   The proportional contribution of lines, testers, and their interaction to the total variance are presented in Table 6. There were different contribution of lines, testers, and their interaction in expression of the studied traits. The contribution of maternal and paternal interaction (line x tester) was very high for all traits except for plant height. It revealed preponderance of paternal and maternal interaction (line x tester) influence for all these traits, while it revealed preponderance of paternal (testers) influence for plant height. These results confirmed those reported by Goksoy et al. (2004).

 

Table 6. Proportional contribution of lines, testers and their interaction in seven sunflower traits evaluated over two seasons of 2007 at Sennar Research Station Farm.

Character

DF

PH

HD

NSH

ES

SW

SY

Lines

21.76

25.37

24.10

44.80

23.10

25.53

25.64

Testers

20.80

53.78

31.37

1.28

1.60

33.07

33.41

L x T

57.44

20.85

44.54

53.92

75.30

41.42

40.95

Where; DF= Days to 50% flowering, PH = Plant height (cm), HD = Head diameter (cm), NSH = No. of seeds per head, ES = Percentage of empty seeds, SW = 1000- seed weight (g) and SY = Seed yield (kg/ha).

 

   Yield is a polygenic character, influenced by the fluctuating environments. Also, it is a complex trait depending on many components. Therefore, one of the main directions of sunflower breeding in the Sudan and elsewhere is the development of hybrids with high genetic potential for seed yield and with plant architecture adaptable to varying environmental conditions. In the combined analysis, GCA effects for seed yield ranged from – 95 for SA1 to 105 for SR41. The best general combining parents for seed yield were SR41, SA3, SR45, SR10 and SR13(Table 4). With regards to specific combining ability, twenty crosses showed significant SCA effects for higher seed yield. However, the variation among the crosses revealed positive and significant SCA in SA4 x SR1 (197.2), SA3 x SR41 (133.91), SA4 x SR45 (129.98), SA3 x SR13 (68.81)and SA3 x SR10 (50.37). While, the unfavorable negatively SCA effects were in SA3 x SR14(-172.8), SA4 x SR7 (-128.3) and SA4 x SR2           (-119.7) (Table5).

 

 

 

 

   Specific combining ability variances were higher than general combining ability variance for seed yield, an indication of non-additive type of gene action being involved for this trait. Also, the non-additive component of genetic variance was more influential in the inheritance of seed yield, as confirmed by the GCA/SCA ratio being below one in the F1 generation (0.65) (Table 3). Hence, the greatest average contribution to the expression of seed yield (41%) was found in the lines x testers. The contribution of 26% by lines and 33% by testers were insignificant (Table 6).  These results were in agreement with the findings of Gangappa et al.(1997), Kumar et al.(1998), Goksoy et al. (1999), Naik et al.(1999),Ashok et al.(2000)and Goksoy and Turan (2004).

   Therefore, across two seasons, parents including SR10, SR13, SR41, SR45, SA3 and SA4 all have positive GCA effects for seed yield. Out of these parents SR41, SA3 and SR10 had positive effects for 1000-seed weight and head diameter, in addition to positive GCA effects for number of seeds per head except SA3 (male sterile line) which has negative GCA effect for number of seeds per head. Those parents have high potential for seed yield and other desirable traits. Also, those parents are more suitable for recombination for developing parents of hybrids for both rainfed and irrigated systems of Sudan.

Information on the relative amount of GCA and SCA variances is of great value in the development of efficient breeding programs. In the present study, the ratio of GCA and SCA variance was less than one for seed yield, 1000-seed weight, number of seeds per head, head diameter, and days to50% flowering. This implies that gene action for these traits was non-additive in nature. Also, the relative contribution of line x tester interactions was found more important for the above mentioned traits.

   With regards to SCA across two seasons, the best crosses with positive SCA effects for seed yield were SA3 x SR41, SA3 x SR10, SA3 x SR13, SA4 x SR45 and SA4 x SR1. In this study, the estimates of SCA effects revealed that the best crosses were SA1 x SR1 for days to flowering, SA3 x SR13 for plant height, SA3 x SR2 for head diameter, SA1 x SR7 for number of seeds per head, SA4 x SR2 for percentage of empty seeds, SA4 x SR1 for 1000-seed weight and SA4 x SR1 followed by SA3 x SR41 and SA4 x SR45for seed yield.

 

 

   Therefore, high SCA effects resulting from crosses between higher general combiners can be improved through early selection. High SCA effects resulting from low GCA combiners suggest that such crosses may be utilized for further improvement through single plant selection in the later generations.

Trait association

   The rank correlation coefficients of general combining ability (GCA) between agronomic traits of sunflower parents across the two seasons were presented in Table (7). Seed yield was positively and significantly correlated with 1000-seed weight(r =0.88**), and head diameter (r =0.71**) and significantly correlated with days to 50% flowering(r =0.50*).The highest correlation coefficients with seed yield were recorded with seed weight and head diameter, indicating the importance of these components in sunflower seed yield. This result agreed with the results obtained by Nehru and Manjunath(2003) and Wani(2004) who independently reported positive association between seed yield with head diameter and number of seeds per head. Also, seed yield per head (data not shown) and seed yield were highly significant and positively correlated with days to 50% flowering, head diameter, number of seeds per head and 1000-seed weight. This indicates the importance of these characters as yield components. This strong association could be due to the fact that head size is determined by its diameter and the yield per plant is determined by seed number per head and weight per 1000 seeds. Generally, the large –sized heads give more seeds; hence the plant will yield more if the weight of1000 seeds is also high. Thus, number of filled seeds and 1000-seed weight exert the highest direct effects on seed yield.

 

 

 

 

 

 

 

 

 

 

 

Table 7. Coefficient of rank correlation for various traits based on ranking GCA effects for sunflower parental populations and lines across two seasons.

 

DF

PH

HD

NSH

ES

SW

PH

0.32

 

 

 

 

 

HD

0.53 *

0.25

 

 

 

 

NSH

0.52 *

0.11

0.29

 

 

 

ES

0.35

0.11

0.29

0.22

 

 

SW

0.64 **

- 0.10

0.70 **

0.57 *

0.65 **

 

SY

0.50*

- 0.35

0.71 **

0.53 *

-0.60 *

0.88 **

DF = days to 50% flowering, PH = Plant height (cm), HD = Head diameter (cm), NSH

= number of seeds per head, ES = Percentage of empty seeds, SW = 1000- seed weight

(g), and SY = Seed yield (kg/ha).

*,** Significant at, 0.05 and 0.01 levels of probability, respectively

 

   Seed yield was negatively and significantly correlated (r = -0.60*) with percentage of empty seeds; negative but not-significantly correlated with plant height. This indicates that percentage of empty seeds depends mainly on some of sunflower morphological character like plant height. Because the increase in stem diameter leads to increase in plant height and total vegetative surface, hence reducing the seed set and seed yield. On the other hand, days to 50% flowering showed a significant and positive correlation with head diameter with r value of 0.53* (Table 7). This indicates that early flowering provides sufficient time for seed formation process and good seed filling period with respect to head diameter. Also, this result suggested that seed yield is positively correlated with good vegetative growth and the earliness can be reliable selection criteria for seed yield in sunflower for genetic materials under study. Thus, sunflower seed yield per unit area can be enhanced by improving days to flowering, head diameter, number of seeds per head and 1000-seed weight.

 

 

 

 

 

 

 

 

CONCLUSION

 

   Based on the study results, significant differences were found among the genotypes studied (parents and hybrids) in the general and specific combining ability effects. The sunflower female lines SA3 and SA4 and the restorers SR 41, SR 45, SR 10 and SR 13 were better general combiners for most of the traits including seed yield across two seasons. However, the hybrid demonstrated best specific combining ability effects for seed yield was SA4 x SR1 followed by SA3 x SR41, SA4 x SR45, SA3 x SR13and SA3 x SR10.The non-additive component of genetic variance played the main role in the inheritance of all traits except plant height, as shown by the analysis of variance of combining abilities and genetic variance components. This was supported by the GCA/SCA ratio in the F1 generation, which was below the value of one of all traits (except plant height) across two seasons of study. Also, the relative contribution of line x tester was very high, which revealed predominance of parental and maternal interaction influence for all traits except plant height. Hence, testers played an important role towards plant height, which revealed predominance of parental influence for this trait. Furthermore, simple correlation analysis revealed that seed yield was positively and significantly correlated with 1000-seed weight, head diameter and number of seeds per head, and negatively correlated with percentage of empty seeds.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

REFERENCES

 

Ashok, S., S.N. Muhammad and S.L. Narayanan. 2000. Combining ability studies in sunflower (Helianthus annuus L.). Crop Research Hisar. 20(3): 457-462.

Gangappa, E., K.M. Channakri shnaiah, S. Ramesh and M.S. Harini. 1997. Exploitation of heterosis in sunflower (Helianthus annuus L.). Crop Research13(2): 339-348.

Goksoy, A.T., A. Turkec and Z.M. Turan. 1999.Research on determination of superiorhy brid combinations in sunflower (Helianthus annuus L.). Turkish Journal ofAgriculturalandForestry23(1): 25-30.

Goksoy, A.T., and Z.M. Turan. 2004.Combining abilities of certain characters and estimation of hybrid vigour in sunflower (Helianthus nnuus L.). ActaAgronomicaHungarica52(4): 361-368.

IRRSTAT 2006.Statistical analysis package from International Rice Research Institute(IRRI), Philippines.

Kempthorne, O. 1957. Introduction to Genetic Statistics. John Wiley and Sons, Inc. New York, USA.

Kumar, A.A., M. Ganesh and P. Janila. 1998. Combining ability analyses for yield and yield contributing characters in sunflower (Helianthus annuus L.). Annals of Agricultural Research20(4):478-480.

Mohamed, M. Y. 2010. Development and stability of some Sudanese sunflower hybridsunder irrigated conditions. Helia 33 (52): 135-144.

Naik, V.R., S.R. Hiremath and K. Giriraj. 1999. Gene action in sunflower. Kamataka Journal ofAgricultural Sciences 12(1/4): 43-47.

Nehru, S.D. and A. Manjunath. 2003. Correlation and path analysis in sunflower(Helianthus annuus L.). Karnataka Journal of Agricultural Sciences 16(1): 39-43.

Shekar, G.C., H. Jayaramaiah, K. Virupakshappa and B.N. Jagadeesh. 1998. Combining ability of high oleic acid in sunflower. Helia21(28): 7-14.

Singh, R.K. and  B.D. Chaudhary.1985. Biometrical Methods in Quantitaive Genetic Analysis. Kalyani Publishers, New Delhi, India.

Wani, M.A. 2004. Correlation and regression studies in sunflower. Advances in Plant Sciences 17(1): 329-332.

published in Gezira Journal of agricultural science

  • Performance, genetic variability and heritability of sesame (Sesamum indicum L.) seed yield and its components in central Sudan

 

Performance, genetic variability and heritability of sesame

(Sesamum indicum L.) seed yield and its components in central Sudan

                                                                                                        

Badr Eldin K. Eltayeb¹, Khalafalla A. Ali² and Abu Elhassan S. Ibrahim¹

 

¹Faculty of Agricultural Sciences, University of Gezira, Wad Medani, Sudan.

²Agricultural Research Corporation, Gedarif, Sudan.

 

ABSTRACT

 

    Success in any breeding programme depends on the magnitude of genetic variability and heritability present in a plant population. Fifteen genotypes of sesame (Sesamum indicum L.) were evaluated during 2012 and 2013 rainy seasons, at Wad Medani, Rahad (under supplementary irrigation) and Gedarif (under rainfed), to assess performance, genetic variability and heritability of seed yield and its components. A randomized complete block design with four replicates was used in each location. The analysis of variance procedure revealed highly significant differences among the 15 sesame genotypes for seed yield in each of the eight individual environments and across them. Combined under supplementary irrigation (Medani and Rahad), the highest yielding genotypes (649–861 kg/ha) were Um Shagara, Kenana-2, Promo, Abu Sofa, Elgezouli and Abu Radoum. They were of medium height and maturity. Characters like capsule length and 1000-seed weight showed high genotypic coefficient of variation and phenotypic coefficient of variation accompanied with high heritability and genetic advance as percent of mean. These characters could be used as selection criteria for high seed yield. Based on mean performance, genotypes Elgezouli, Um Shagara, Kenana-2 and Promo were the highest seed yielders in the four locations.

 

 

 

 

 

 

 

 

INTRODUCTION

 

   Sesame (Sesamum indicum L.) is an important oilseed crop in warm tropical and sub-tropical regions. The Sudan is the 4th major sesame producing country in the world. India, China, Myanmar and Sudan are the main sesame producing countries, accounting for almost 68% of production worldwide (Laurentin and Karlorsky, 2006). In the Sudan, sesame is an important product of the agricultural sector being the third after pearl millet and sorghum (Khidir, 1997). Sesame is a valuable crop for both local and export markets.

   Increasing yield is always the paramount aim of any breeder. However, assessing factors responsible for increasing yield is always difficult. Yield is the end product of action and interaction of vital activities of the plant throughout its life cycle and is controlled by numerous factors shaped by genetics and environment. Among these factors, the most important is the inherent potential of the plant to produce high yield which depends upon the hereditary make-up of the plant. Therefore, for rational improvement of a crop, understanding of the magnitude of genetic variability and the extent to which the desirable characters are heritable becomes essential.

    The determination of genetic variability and its partitioning into various components, in any crop, is necessary to have an insight into the genetic nature of yield and its components. Thus, the magnitude of heritable variation, particularly its genetic components, is clearly the most important part of the breeding material which has a close bearing on its response to selection.

    Heritability is an important parameter in crop improvement programs as it is used to predict response from selection as well as optimum environment selection. Selection efficiency for a character is dependent on the magnitude of its heritability and genetic variation. According to Singh (2001), if heritability of a character is very high (> 80%), selection for such character could be fairly easy. This is because there would be a close correspondence between the genotype and the phenotype due to the relative small contribution of the environment to the phenotype. Hence, effective utilization of any sesame germplasm in a breeding program requires information of genetic variability and heritability. Knowledge of such information will help the breeders to chose the appropriate breeding methods.

 

 

 

 The objectives of the current study were to assess performance, genetic variability and heritability of sesame seed yield and its components in eight different growing environments in centeral Sudan.

 

MATERIALS AND METHODS

 

   The experiment was conducted during 2012 and 2013 rainy seasons at four locations: Gezira University Experimental Farm (UG), Agricultural Research Corporation Farm (ARC), Wad Medani (latitude 14º 25´ N, longitude 33º 29´ E and 407 masl), Rahad Research Farm (Rah), (latitude 13º 31´ N, longitude 34º 32´ E and 570 masl)  and Gedarif Research Farm (Ged), (latitude 14º 1´ N, longitude 35º 21´ E and 592 masl) to evaluate 15 sesame genotypes. The four locations lied within the central clay plain of the Sudan characterized by heavy alkaline clay soil, with a pH of around 8.5 and poor in nitrogen and organic matter.

   The combination of 4 locations and 2 growing seasons formed 8 different growing environments as: UG 12 and UG 13 (University of Gezira seasons 2012 and 2013, respectively), ARC 12 and ARC 13 (Agricultural Research Corporation seasons 2012 and 2013, respectively), Rah 12 and Rah 13 (Rrahad Research Station, seasons 2012 and 2013, respectively) and Ged 2012 and Ged 2013 (Gedarif Research Station, seasons 2012 and 2013, respectively).  

   The plant material used in this study consisted of four local varieties (Abu Sofa, Jugam, Abu Radoum and Abu Sandoog); six released varieties (Kenana-2, Khidir, Promo, Um Shagra, Gedarif-1 and Elgezouli) and five advanced breeding lines (Gd2003-S-P-S-N23, Gd2002-ob-N2-39, Gd2002-S-P-S-N12, Gd2002-S-P-S-N-14 and Gd2008-S-P-S-N-1) from the breeding program of  Dr. Khalafalla Ahmed Ali at Gedarif Research Station, ARC, Sudan.

    The genotypes were tested in a randomized complete block design with four replicates. The material was sown on the first week and second week of July, 2011 and 2012, respectively, in the four locations. The seeds were sown manually in two rows, 5 m long and 0.8 m apart, in holes spaced 0.1 m apart within the row, with a seed rate of 3 kg/ha. The total rainfall at Medani locations in the first and second seasons were 450 mm and 560 mm, respectively, eight supplementary irrigations were applied in the first season and five in the second season, at Rahad were 545 and 675 mm, respectively,

 

 

four supplementary irrigations were applied in the first season and three in the second season and at Gedarif were 515 and 600 mm, respectively, without

supplementary irrigation. The data were collected on days to 50% flowering, days to maturity, plant height (cm), number of capsules per plant, height to first capsule (cm),  capsule length (cm), number of seeds per capsule, 1000-seed weight (g) and seed yield (kg/ha).

    The analysis of variance procedure was used to test differences among genotypes within each location, season, and combined. Estimates of genetic parameters such as genotypic and phenotypic variances were computed following Singh and Chaudhry (1985), genotypic and phenotypic coefficients of variation and broad sense heritability after Burton and Devane (1953). Genetic advance (10% selection intensity) and genetic advance as percent of mean were computed using the formula given by Johnson et al. (1955) and used by Allard (1960).

 

RESULTS AND DISCUSSION

 

    In the present study, highly significant differences among genotypes (p≤0.01) were observed for most traits studied (Table 1). These findings indicate the presence of a large genetic variation among the tested sesame genotypes. Similarly, Arameshwarappa et al. (2009) and Sumathi and Muralidharan (2010) recorded significant differences among a large number of sesame genotypes for days to 50% flowering, days to maturity, plant height, number of primary branches/plant, number of capsules/plant, capsule length, number of seeds/capsule and seed yield.

    Analysis of variance for individual environments and the mean square value revealed highly significant differences among the 15 sesame genotypes for most of the studied traits, indicating the presence of sufficient genetic variation for its effective management through selection to identify the superior genotypes (Table 1).

 

 

 

 

 

 

 

 

Table 1. Mean squares of agronomic traits of 15 sesame genotypes grown in Gezira University Experimental Farm (UG), Agricultural Research Corporation Farm (ARC), Rahad Research Farm (Rah) and Gedarif Research Farm (Ged) during 2012 and 2013 rainy seasons.

 

 

 

 

 

Traits

 

 

 

 

Environment

DTF

DTM

PH

(cm)

NCP

HTFC

(cm)

CL

 (cm)

NSC

1000-SW (cm)

SY

(kg/ha)

UG 11

74***

134***

513***

730***

348***

0.65***

  95*

0.90***

126984***

ARC 11

53***

  89***

  98

206***

149*

0.47***

102***

0.34***

  53279***

Rah 11

52***

  34***

338***

177

325***

0.88***

109**

0.15***

  83049***

Ged 11

42***

  34***

524***

173*

541***

0.75***

280*

0.54*

111018***

UG 12

47***

  82***

658***

202*

225***

0.83***

164**

0.66***

931787***

ARC 12

23***

  28***

393***

  81*

154***

0.72***

105*

0.53***

  46240***

Rah 12

39*

    9***

204***

  43

  72***

0.60***

137*

0.95***

      584*

Ged 12

37***

  12

267***

335

191***

1.10***

178***

0.69***

  42719**

DTF= days to 50% flowering, DTM= days to maturity, PH= plant height, NCP= number of capsules/plant, HTFC= height to first capsule, CL= capsule length, NSC= number of seeds/capsule, 1000-seed weight and SY= seed yield/ha.

 *, **, ***, Significant at the 0.05, 0.01 and 0.001 probability levels, respectively, otherwise    

  non-significant.

 

    The wide range of variability exhibited by most of the measured characters in the eight different environments indicated potential genetic variability in the material under study. The mean squares and mean values of characters in the different environments were greatly variable. The very highly significant difference among the genotypes is a good base for their ability to suit these variable environments.

   Estimates of genotypic variance) 2g),phenotypic variance ( 2P), phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV) are given in Table 2. The genotypic coefficient of variation ranged from 7.2% for days to maturity to 36% for capsule length. At the same time, the range of phenotypic coefficient of variation was 7.4% for days to maturity to 38% for capsule length (Table 2).

   Phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV) values greater than 20% are regarded as high, whereas values less than 10% are considered to be low and values between 10 and 20% to be medium (Deshmukh et al.,1986). Based on this delineation, number of capsules/plant, height to first capsule, capsule length and       1000-seed weight had high GCV and PCV (Table 2). This finding indicates that selection may be effective based on these characters and their

 

phenotypic expression would be a good indication of genetic potential. The large scope for selection based on these characters and the diversity in genotypes provides huge potential for future breeding programs. The GCV and PCV values for days to flowering, plant height and number of seeds/capsule were medium. Days to maturity had low GCV and PCV values indicating the low scope of selection for improvement (Table 2). Similar findings were reported by Aramesh Warappa et al. (2009) considering number of capsules/plant.

   Heritability values are helpful in predicting the expected progress to be achieved through the process of selection. Genotypic coefficient of variation along with heritability estimate provides a reliable estimate of the amount of genetic advance to be expected through phenotypic selection (Wright, 1921). Heritability estimate and genetic advance as percent of mean for characters under study are given in Table 2. Heritability ranged from 98% for days to 50% flowering to 80% for number of capsules/plant. According to Singh (2001), if heritability of a character is very high, say 80% or more, selection for such character could be fairly easy. Accordingly, all characters, under study had very high heritability estimates.

 

Table 2. Estimates of genotypic ) 2g), and phenotypic ( 2P),  variances, heritability ( %),genotypic (GCV) and phenotypic (PCV) coefficients of variation and genetic advance as percentage of the mean (GAM)  for eight traits of 15 sesame genotypes.

Traits

 

 

 (%)

 

GCV

 (%)

PCV

(%)

GAM

(%)

Days to 50% flowering

  66.8

  68.0

98

16.7

16.8

  4.1

Days to maturity

  52.5

  54.3

96

  7.2

  7.4

  2.0

Plant height (cm)

438.1

469.6

93

18.2

18.8

  1.7

Number of capsules/plant

123.8

153.8

80

30.0

33.0

  4.6

Height to first capsule (cm)

273.9

289.8

95

27.0

28.0

  3.2

Capsule length (cm)

    1.2

    1.3

92

36.0

38.0

62.7

Number of seeds/capsule

  69.0

  83.0

83

14.0

16.0

  3.2

1000-seed weight (g)

  0.83

  0.85

97

27.0

28.0

60.0

           

   

 

 

 

 

   The range for genetic advance as percent of the mean was from 1.7% for plant height to 62.7% for capsule length (Table 2). Capsule length (62.7%) and 1000-seed weight (60) had relatively high genetic advance as percent of the mean. While the other characters showed low genetic advance as percent of the mean. Selection based on those traits with a relativity high GAM will result in the improvement of the performance of the genotypes. According to Johnson et al. (1955), high heritability estimate along with high genetic advance as percent of the mean is usually more helpful in predicting gain under selection than heritability estimates alone. The present study showed that high heritability coupled with high expected genetic advance as percent of the mean were exhibited by capsule length and 1000-seed weight (Table 2). Phenotypic selection for these characters would likely be effective than other characters measured.

   In the present investigation, characters like capsule length and 1000-seed weight showed high GCV and PCV accompanied with high heritability and genetic advance as percent of the mean. These characters could be used as selection criteria for high seed yield.

     Days to 50% flowering is used as an indicator of earliness. Earliness is a trait that has been subjected to selection in the past and current breeding programs. Identification of early, medium and late flowering genotypes is of paramount importance in selecting genotypes for the different growing environments (high, medium and low rainfall as well as supplementary irrigation).

   The overall mean of days to 50% flowering was 46 days and 52 days in the first and second seasons, respectively (Table 3). This result indicated that the genotypes flowered earlier in the first season due to fact that the average rainfall in the second season was higher than that of the first season at the four locations.

   Across environments, the early flowering genotypes were Abu Radoum, Kenana-2, Gd2008-S-P-S-1, Gd2003-S-P-S-N-23, and Um Shagra, respectively, with a range of 45 – 47 days. The medium flowering genotypes were Elgezouli, Gd2002-ob-N-2-39, Gd2002-S-P-S-N-12, Khidir, Promo and Abu Sandoog, respectively, ranging from 48–49 days. The late flowering genotypes were Jugam (57 days), Gedarif-1 (53 days), Gd2002-S-P-S-N-14  (50 days) and Abu Sofa (50 days) (Table 3).

 

 

 

 

Table 3. Means of days to 50% flowering, days to maturity and plant height for 15 sesame genotypes grown in Gezira University Experimental Farm (UG), Agricultural Research Corporation Farm (ARC), Rahad Research Farm (Rah) and Gedarif Research Farm (Ged) during 2012 and 2013 rainy seasons.

                                               

 

DTF

 

 

DTM

 

 

PH

 

Genotype

2012

2013

Mean

2012

2013

Mean

2012

2013

Mean

Gd2003-S-P-S-N-23

43

51

47

96

104

100

128

 109

 119

Gd2002-ob-N-2-39

46

51

48

95

104

  99

127

 105

 116

Gd2002-S-P-S-N-12

44

52

48

90

102

  96

119

   94

 106

Gd2002-S-P-S-N-14

48

52

50

91

103

  97

127

 103

 115

Gd2008-S-P-S-N-1

44

49

47

95

105

100

114

   92

 103

Kenana-2

42

50

46

93

103

  98

123

 104

 113

Khidir

45

51

48

95

104

100

112

 102

 107

Promo

46

53

49

96

106

101

125

 108

 117

Um Shagra

45

50

47

87

101

  94

128

   99

 114

Gedarif-1

49

57

53

99

110

105

137

 121

 129

Abu Sofa

46

53

50

96

104

100

131

 113

 127

Jugam

55

58

57

99

107

103

132

 119

 126

Abu Radoum

40

49

45

93

105

  99

110

   95

 103

Elgezouli

45

51

48

92

105

  98

125

 111

 118

Abu Sandoog

46

52

49

94

105

  99

124

 107

 116

Mean

46

52

49

95

105

100

125

 106

 115

SE (±)

1.5

2.7

2.2

2.8

2.4

  2.6

 9.5

12.2

11.2

C.V%

3.3

5.2

4.5

3.0

2.3

  2.6

 7.4

11.6

  9.7

DTF= days to flowering, DTM= days to maturity and PH= plant height (cm).

 

    Maturity of genotypes differed in the two seasons and over environments, 95,105 and 100 days in the three cases, respectively (Table 3). The early maturing genotypes were Um Shagra (94 days), Gd2002-S-P-S-N-12 (96 days), Gd2002-S-P-S-N-14 (97days), Kenana-2 (98 days) and Elgezouli (98 days).The late maturing genotypes were Gedarif-1 (105 days), Jugam (103 days) and Promo (101 days) (Table 3).

   The three groups of  sesame genotypes, i. e. (promising breeding lines Gd2003-S-P-S-N-23, Gd2002-ob-N-2-39, Gd2002-S-P-S-N-12, Gd-2002-S-P-S-N-14 and Gd2008-S-P-S-N-1), released varieties (Kenana-2, Khidir, Promo, Um Shagra, Elgezouli and Gedarif-1) and local (Beladi) varieties (Abu Sofa, Jugam, Abu Radoum and Abu Sandoog) contain early, as well as , late maturing genotypes. This makes selection for earlines, possible within each group. It seems that earliness is not highly affected by the environment, since the early genotypes were always early under both seasons and irrigation systems.

 

   Plant height is a varietal characteristic, but the influence of the environment on this character is significant. Contrary to the flowering and maturity traits, the mean values for plant heights were higher in the first season (125cm) compared to the second season (106cm) and across environments (115 cm) (Table 3). Ray (2007) reported that plant height is affected by rainfall conditions.

   The tallest genotypes were Gedarif-1, Abu Sofa and Jugam, (129, 127 and 126 cm), respectively. These were also late maturing genotypes, confirming the known fact that late maturing genotypes are tall, while the shortest genotypes were Abu Radoum (103cm) and Gd2008-S-P-S-N-1, (103 cm), (Table 3). These were also early maturing genotypes. The short stemmed, little-branched types are generally early maturing than the taller branched types (Weiss, 1971).

    The number of capsules per plant followed the same trend shown by plant height, i.e. higher in the first season compared to the second season (43 and 32 capsules), respectively. The highest number of capsules per plant were shown by Um Shagra (44 capsules), Gedarif-1 (40 capsules) and Abu Radoum (40 capsules) (Table 4). With the exception of Gedarif-1 these were not the tallest genotypes.

     The lowest number of capsules per plant were exhibited by Gd2002-S-P-S-N12 (27 capsules), Gd2002-S-P-S-N-14 (31 capsules), Gd2008-S-P-S-N-1       (31 capsules), Khidir (33 capsules) and Jugam (35 capsules), an indication that the improved lines showed low values of this character (Table 4), while the released and local varieties gave the highest values. This fact should be taken in consideration when selecting for seed yield since number of capsules per plant was found to have the greatest direct effect on seed yield       (Gupta, 1976).

 

 

 

 

 

 

 

 

 

 

 

 

Table 4. Means of number of capsules/plant, height to first capsule and capsule length for 15 sesame genotypes grown in Gezira University Experimental Farm (UG), Agricultural Research Corporation Farm (ARC), Rahad Research Farm (Rah) and Gedarif Research Farm (Ged) during 2012 and 2013 rainy seasons.

 

Genotype

 

NCP

 

 

HFC

 

 

CL

 

2012

2013

Mean

2012

2013

Mean

2012

2013

Mean

Gd2003-S-P-S-N-23

47

  28

  38

59

53

  56

2.8

2.8

2.8

Gd2002-ob-N-2-39

47

  31

  39

62

51

  57

3.4

3.4

3.4

Gd2002-S-P-S-N-12

31

  24

  27

67

54

  61

3.9

3.9

3.9

Gd2002-S-P-S-N-14

36

  25

  31

71

59

  65

3.5

3.5

3.5

Gd2008-S-P-S-N-1

37

  25

  31

60

47

  53

3.0

2.9

3.0

Kenana-2

43

  35

  39

67

52

  60

2.8

2.7

2.7

Khidir

34

  33

  33

62

54

  58

2.8

2.7

2.7

Promo

45

  34

  39

66

56

  61

2.8

2.7

2.8

Um Shagra

54

  35

  44

70

51

  60

2.9

2.6

2.7

Gedarif-1

44

  36

  40

71

51

  61

2.4

2.4

2.4

Abu Sofa

42

  30

  36

76

59

  68

3.0

2.9

3.0

Jugam

34

  36

  35

77

67

  72

3.2

3.2

3.2

Abu Radoum

51

  28

  40

48

46

  47

2.9

2.6

2.7

Elgezouli

42

  34

  38

69

58

  64

3.3

3.2

3.3

Abu Sandoog

44

  32

  38

70

59

  64

2.6

2.5

2.5

Mean

43

  32

  37

67

55

  61

3.1

3.0

3.0

SE (±)

 10.2

11.7

11.0

7.3

8.3

 8.0

0.3

0.2

0.2

C.V%

 23.8

37.3

29.7

11.0

15.0

13.1

8.3

6.0

7.2

        NCP= number of capsules/plant, HFC= height to first capsule (cm) and CL= capsule   

        length (cm).   

    

    Height to first capsule is an important character for mechanical harvesting and is an indication to height uniformity. The highest values (64 – 72 cm) to first capsule were shown by the local (Beladi) varieties (Jugam, Abu Sandoog and Abu Sofa) while the lowest values (47 – 58 cm) for such character were shown by the improved breeding lines and Abu Radoum and Khidir (Table 4).  This is the same trend shown by number of capsules per plant and plant height. Similar trends are an indication of close positive relationships among traits.

 

 

 

 

   Height to first capsule differed in the two seasons and over environments,      i. e. 67, 55 and 61cm in the three cases, respectively (Table 4). Height to first capsule in the second season was reduced compared to the first season, 55 to 67cm across the four locations. This great reduction was attributed to the adverse effect of high average rainfall during the second season. This result could be attributed to the fact that sesame is more sensitive to excess water (water logging) or cloudy environments.

   It is generally indicated that higher yielding sesame varieties start to form capsules at lower heights. Moreover, for mechanized harvesting, varieties forming capsules at lower height are more suitable and recommended.

   Capsule length is an important yield component that has a great direct effect on seed yield (Singh et al., 1997; Sumathi et al., 2007).The improved  breeding  lines gave the highest capsule length, Gd2008-S-P-S-N-1-(3.0 cm),- Gd2002-ob-N-2-39-(3.4 cm),-Gd2002-S-P-S-N-14-(3.5 cm) and Gd2002-S-P-S-N-12 (3.9 cm), while the released variety Gedarif-1 (2.4 cm) gave the lowest value (Table 4). Capsule length followed a totally different trend compared with height to first capsule and plant height, an indication of their negative associations. The capsule length varied from 2.4 cm to 3.9 cm indicating a wide range of variability in this character among the genotypes.           

   Number of seeds per capsule is a major yield component, which is highly correlated with yield (Dixit et al., 1997). Over environments, the number of seeds per capsule ranged from 51-61, the highest number of seeds per capsule were shown by Elgezouli (61), Abu Sandoog (61) and Gd2002-ob-N-2-39 (60), while the lowest values were shown by Khidir (51), Kenana-2 (52) and Gedarif-1 (52) (Table 4).

   The local variety Abu Radoum gave the highest number of seeds per capsule, (61 seeds), the released variety Elgezouli (61 seeds) and the improved breeding lines, Gd2002-S-P-S-N-12 (58 seeds) and Gd2002-ob-N-2-39 (61 seeds) (Table 4) and the released varieties Gdarif-1 (52 seeds) and Kenana-2 (52 seeds) gave the lowest values. Number of seeds per capsule followed the same trend compared with capsule length, an indication of their positive association. Weiss (1971) reported that number of seeds per capsule varies according to the number of locules/capsule and capsule length.

 

 

 

 

 

 

              The number of seeds per capsule varied from 51 to 61 seeds indicating a wide range of variability in this character among the genotypes (Table 5). Similar results were also observed by Lazim (1973) and Ahmed (1985), who showed considerable variations among sesame cultivars in number of seeds per capsule. Variability in number of seeds per capsule between varieties might be due to genotypic factors as stated by Padmavathi (1997).

 

Table 5. Means of number of seeds/capsule, 1000-seed weight and seed yield/ha  for 15 sesame genotypes grown in Gezira University Experimental Farm (UG), Agricultural Research Corporation Farm (ARC), Rahad Research Farm (Rah) and Gedarif Research Farm (Ged) during 2012 and 2013rainy seasons.

                                               

 

NSC

 

 

1000- SW

 

 

SY

 

Genotype

2012

2013

Mean

2012

2013

Mean

2012

2013

Mean

Gd2003-S-P-S-N-23

   59

   51

   55

  3.0

   3.5

  3.2

   625

   589

   607

Gd2002-ob-N-2-39

   68

   51

   60

  3.0

   3.1

  3.1

   642

   556

   599

Gd2002-S-P-S-N-12

   64

   51

   58

  3.3

   3.7

  3.5

   451

   443

   447

Gd2002-S-P-S-N-14

   60

   46

   53

  3.1

   3.2

  3.1

   645

   468

   557

Gd2008-S-P-S-N-1

   64

   46

   55

  3.5

   4.1

  3.8

   684

   542

   613

Kenana-2

   59

   45

   52

  3.0

   3.6

  3.3

   711

   660

   686

Khidir

   60

   43

   51

  3.1

   3.8

  3.4

   574

   516

   545

Promo

   62

   47

   55

  3.2

   3.4

  3.3

   708

   688

   698

Um Shagra

   63

   49

   56

  3.0

   3.3

  3.2

   889

   615

   752

Gedarif-1

   62

   43

   52

  2.8

   2.9

  2.8

   655

   581

   618

Abu Sofa

   59

   47

   53

  3.1

   3.4

  3.3

   746

   545

   645

Jugam

   64

   43

   54

  2.3

   2.5

  2.4

   371

   509

   440

Abu Radoum

   59

   47

   53

  3.3

   3.7

  3.5

   748

   537

   643

Elgezouli

   66

   57

   61

  3.0

   3.7

  3.4

   800

   694

   747

Abu Sandoog

   72

   50

   61

  2.8

   3.1

  3.0

   576

   523

   550

Mean

   63

   48

   55

  3.1

   3.5

  3.3

   655

   565

   610

SE (±)

  7.8

  7.1

  7.4

  0.3

   0.3

  0.3

126.0

185.0

164.7

C.V%

12.4

14.7

13.4

17.0

   7.4

  7.9

19.2

32.7

27.0

 NCP= number of seeds/capsule, 1000-SW= 1000-seed weight (g) and SY= seed  yield/ha (kg/ha).

 

 

 

   Thousand seed weight is an important trait, which should be considered in selection for high yield. Thousand seed weight in the second season data were better than the first season at the four locations (Table 5). The 1000-seed weight ranged from 2.3 g which was registered by Jugam and 3.5 g which was scored by Gd2008-S-P-S-N-1 in the first season. In the second season, it ranged from (2.5 g) Jugam  to (4.1 g) Gd2008-S-P-S-N-1. Over environments, it ranged between (2.4 g) Jugam  to (3.8 g) Gd2008-S-P-S-N-1 (Table 5).

   The highest 1000-seed weight values were shown by Gd2008-S-P-S-N-1   (3.8 g), Gd2002-S-P-S-N-12 (3.5 g), Abu Radoum (3.5 g) and Khidir (3.4 g). The lowest 1000-seed weight values were shown by Jugam (2.4 g) and Gedarif-1 (2.8 g). It seems that the improved breeding lines gave high values of seed weight than the local varieties. The highest 1000-seed weight genotypes were Gd2008-S-P-S-N-1(3.8g) and Abu Radoum (3.5g) (Table 5).  Variation in 1000- seed weight of sesame among varieties was also reported by Amna et al. (2007).

   Seed yield per unit area is of primary importance as a breeding objective because it affects the economic return to the sesame grower. Sesame genotypes in this study differ in their inherent yield potential.

   Analysis of variance for seed yield showed that there were significant differences among the sesame genotypes within and over environments (Table 1). Seed yield in the first season was higher than the second season for all locations (Table 5). A wide range of variability in seed yield (294 –1227 kg/ha) was detected.

    The highest seed yield (1227 kg/ha) was recorded by Kenana-2 in the second season, whereas the lowest seed yield (294 kg/ha) was recorded by Jugam in the first season.

    In the first season, sesame genotypes Um Shagra and Elgezouli were the highest yielders (889 and 800 kg/ha), respectively, whereas in the second season the genotypes Elgezouli and Promo were the highest yielders (694 and 688 kg/ha) (Table 5). Combined analysis of variance over environments revealed that genotype Um Shagra and Elgezouli were the highest yielders (752and 747 kg/ha), respectively and the lowest genotypes were Jugam and Gd2002-S-PS-N-12 (440 and 447 kg/ha), respectively.

 

 

 

 

 

 

    The over environments data revealed that the improved breeding lines were low yielders compared to the released and local varieties. The released varieties were high yielders compared with the local ones (Table 5). Within released varieties, Um Shagra and Elgezouli were the higher yielders        

 (747 and 752 kg/ha). Seed yield in the second season was reduced compared to the first season 565 to 655 kg/ha at the combined analysis of the four locations (Table 5). This great reduction in seed yield was attributed to the adverse effect of high average rainfall during the second season. This result could be attributed to the fact that sesame is more sensitive to excess water (water logging) or cloudy environments.

    The highest yielders were Um Shagara, Kenana-2, Promo, Elgezouli, Abu Sofa and Abu Radoum for all locations. At the same time, these genotypes were of medium height, flowering and maturity. This result could be attributed to the fact that these genotypes achieve great success in supplementary irrigation and under rainfed areas.

    In general, the high yielding genotypes are characterized by medium plant stature, high number of capsules per plant, long capsules, early to medium maturity and high 1000-seed weight.

 

CONCLUSION

 

     In conclusion highly significant differences among genotypes were observed for most traits studied over environments indicating the presence of a large genetic variation among the tested 15 sesame genotypes. High genotypic coefficient of variation (GCV), phenotypic coefficient of variation accompanied with high heritability and genetic advance as percent of mean were shown by capsule length and 1000-seed weight. These characters could be used as selection criteria for high seed yield.  Varieties Elgezouli, Um Shagara, Kenana-2 and Promo were recommended for the four locations.

 

 

 

 

 

 

 

 

REFERENCES

 

Ahmed, M. E. 1985. Genetic studies in sesame (Sesamum indicum L.). M.Sc. (Agriculture) Thesis. University of Khartoum. Sudan.

Allard, R. W. 1960. Principles of  Plant Breeding. John Wiley and Sons, Inc. New York, USA, p. 236.

Amna, A., A. Abdalla., M. Omer and E. Ahmed. 2007. Response of sesame cultivars to sowing date under irrigation. A paper submitted to the Crop Husbandry Committee Meeting, Agricultural Research Corporation, Wad Medani, Sudan.

Arameshwarappa, S. G., M. G. Palakshappa., P. M. Salimath and K. G. Parameshwarappa. 2009. Studies on genetic variability and character association in germplasm collection of sesame (Sesamum indicum L.). Karnataka Journal of Agricultural Science 22: 252-254.

Burton, G. W. and H. E. Devane. 1953. Estimating heritability in tall fescue (Festuca arundinacea L.) from replicated clonal material. Agronomy Journal 45: 478-481.

Deshmukh, S. N. N., M. S. Basu and P. S. Reddy. 1986. Genetic variability, characters association and path coefficient analysis of quantitative traits in Virginia bunch varieties of groundnut. Indian Journal of Agricultural Science 56: 515-518.

Dixit, J. P., N. S. V. Rao., R. G. Ambabatiya and R. R. Khan.1997. Productivity of sesame cultivars sown as semi-arid under various plant densities and nitrogen levels. Crop Research 1311: 27-31.

Gupta, T. R. 1976. Path coefficient analysis of seed in sesame. Oilseeds Journal 6: 27-29.

Johnson, H. W., F. H. Robinson and E. R. Comstock. 1955. Estimates of genetic and environmental variability in soybeans. Agronomy Journal 47: 311-318.

Khidir, M. O. 1997. Oilseed Crops in the Sudan (in Arabic). Khartoum University Press, Sudan.

Lazim, M. A. 1973. Population and varietal effects on growth and yield of sesame under irrigation. M.Sc. Thesis, University of Khartoum, Sudan.

Laurentin, H and P. Karlorsky. 2006. Genetic relationship and diversity in sesame (Sesamum indicum L.) germplasm collection using amplified fragments length polymorphisms (AFLP). Sesame, Safflower News 17: 37-39.

 

Padmavathi, N. 1997. Genetic variability for seed yield and its components in sesame. Sesame and Safflower Newsletter 12: 64-65.

Ray, D. 2007. Edible Oilseed, Grains and Grain Legumes. Phenology of Sesame. Issues in New Crops and Uses. J. Januck and A. Whipkey (eds). ASHS Press. Alexandria VA. pp. 155.

Singh, R. K. and V. D. Chaudhry. 1985. Biometrical Methods in Quantitative Genetic Analysis. Kalyani Publishers, New Delhi, India. pp. 55-54.

Singh, P. K., K. R. Dixit and K. R. Yadva. 1997. Estimates of genetic parameters, character association and path analysis in sesame. Crop Research 13: 115-119.

Singh, B. D. 2001. Plant Breeding Principles and Methods. Kalyani Publishers, New Delhi. India, pp: 896.

Sumathi, V., V. Muralidharan and N. Manivannan. 2007. Trait association and path analysis for yield and yield attributing traits in sesame (Sesamum indicum L.). Madras Agricultural Journal 94: 174-178.

Sumathi, P and V. Muralidharan. 2010. Analysis of genetic variability, association and path analysis in the hybrids of sesame (Sesamum indicum L.). Tropical Agricultural Research 13: 63-67.

Weiss, E. A. 1971. Castor, Sesame and Sunflower. Leonard Hill. London.

Wright, S. 1921. Correlations and causations. Journal of Agricultural Research 20: 557-558.

 

 

 

 

 

published in Gezira Journal of agricultural science

  • Genotype x environment interaction and stability analysis of grain sorghum (Sorghum bicolor (L.) Moench) yield under rainfed and irrigation conditions in central Sudan

 

Genotype x environment interaction and stability analysis of grain sorghum (Sorghum bicolor (L.) Moench) yield  under rainfed and irrigation conditions  in central Sudan

 

Mohammed¹  H. Mohammed, Abu Elhassan S. Ibrahim² and Ibrahim N.Elzein.¹

¹ Agricultural Research Corporation ,Wad Medani ,Sudan.

² Faculty of Agricultural Sciences, University of Gezira ,Wad Medani,

  Sudan.

ABSTRACT

     An experiment was conducted over three consecutive seasons (2009, 2010, and 2011) at three locations , Rahad Research farm. Gedarif  Research Station farm (North Gedarif and South Gedarif region) of the Agricultural Research Corporation (ARC), Sudan. Both North and South Gedarif were rainfed, while Rahad station was irrigated. A randomized complete block design with four replicates was used. Sorghum production is highly influenced by the environment where it is grown, thus, the genotype by environment interaction is highly significant when breeding for specific adaptation. The objective was   to assess the genotype x environment interaction and stability of grain yield. The mean squares due to environment, genotypes and genotype x environment interaction were highly significant for grain yield. Significant differences among  genotypes for the studied characters were found in almost all seasons, indicating that these sorghum genotypes were highly variable for the characters studied and , therefore, expected to respond to selection. The interaction effects of genotype x location were highly significant for most traits indicating that genotypes responded differently to different environments and some are environmentally specific. The present study showed that the first two axes PCA1,PCA2 in Additive Main Effect and Multiplicative Interaction (AMMI ) accounted for the GE sum of squares by 56.7% and 19.3%, respectively, while the regression analysis accounted for GE sum of squares by 21.9% .Hence, AMMI analysis was superior to regression techniques and more effective in partitioning the interaction sum of squares. From both statistical  stability models used in this study, i.e. Eberhart and Russell (1966) as well as the Aditive Main Effect and Multiplicative Interaction (AMMI) analysis, they pointed out  that genotypes Mugod (1510 kg/ha), Tabat (1299 kg/ha), Wad-Ahmed (1471 kg/ha), Gadambalia bloom (1428 kg/ha), Safra (1410 kg/ha) and Tetron (1323) were high yielding and stable under the favorable environments of South Gedarif and Rahad irrigated Scheme. Genotypes Wad Baku(1225 kg/ha), Farhoda (1252 kg/ha),Gesheish (1194 kg/ha) and Wad Fahal (1230 kg/ha) were low yielders but quite stable under low rainfall environments like North Gedarif environment.

 

INTRODUCTION

 

     Sorghum (Sorghum bicolor (L) Moench) is an important food and feed crop. As an energy supplier for the world’s population, it ranks sixth, and it is fifth in importance among cereals. Semi-arid tropical Asia and semi-arid tropical Sub-Saharan Africa grow about 60% of the world area (ICRISAT and FAO, 1996), while Sudan grows about 24% of Africa area and produces 17% of its production. The national average yield in the Sudan (250 kg/fed) was 18% of that obtained at the research stations (Ishag and Ageeb 1987). This was attributed among many other factors, to the use of low yielding cultivars as well as to poor cultural practices.

    During the last 15 years, plant breeders in the Agricultural Research Corporation (ARC) have successfully developed high yielding open pollinated varieties such as Feterita, Wad Ahmed, Ingaz (Osman and Mahmoud,1992) and Tabat (Osman et al.1996). In addition, many other varieties suitable for both irrigated and rainfed sectors were also developed such as Butana and Bashayer (Elzein  et al.,2008), and AG-8 (Abdalla et al., 2009).

     Estimation of stability performance has  become an important tool to identify consistently high-yielding genotypes (Kang,1998). Many stability statistical methods have been used to determine whether or not cultivars evaluated in multi-environment trials were stable (Lin et al.,1986; Flores et al.,1998; Hussein et al.,2000; Robert,2002).The use of a method that integrated yield  performance and stability  for superior genotypes becomes important  because the  most stable  genotypes  were not often the highest yielding (Kang and Magari ,1996). 

     Conventional methods of partitioning total variation into components due to variety, environment and variety-environment interaction conveyed little information on individual patterns of response (Kempton ,1984). Other methods used include regression analysis to partition  genotype x environment interaction (Gauch,1988), and multivariate analysis (Westcoff,1987). Development of sorghum with high yielding and desirable grain quality for different environments is one of the exciting research that leads to successful evaluation of stable genotypes which could be used for  general cultivation or as breeding material. Therefore, the objective of this study was to assess genotype x environment interaction and stability of sorghum  grain yield using regression method of Eberhart and Russels, (1966). The deviation from regression is used to assess unpredictable part of variability

 

of any genotype with respect to environment that could not be predicted by the regression. It is a measure of reliability of the linear regression and the stable genotype  was defined as one with bi = 1, S2d = 0 and higher than the overall mean grain yield , and more recent application methods such as Additive Main and Multiplicative Interaction analysis (AMMI). Multivariate analysis such as AMMI analysis groups genotype or environments in a qualitative manner according to their similarity of performance rather than quantitative manner of the stability parameters. AMMI analysis involves the clustering analysis to classify genotypes under the most adapted sites for them depending on the AMMI principle components scores (Gauch and Zobel,1988; Nachit et al. 1992). Non parametric approach (multivariate) has been proposed to overcome problems associated with parametric approach (Lin et al,1986).

 

MATERIALS AND METHODS

Location

        The experiments were conducted over three consecutive seasons  (2009, 2010  and 2011) at three locations,viz. Rahad  Research Farm , North and South Gedarif regions of the Gedarif Research Station farm of the Agricultural Research Corporation (ARC), Sudan . The three locations lied within the central clay plain of the Sudan, characterized by heavy alkaline clay soil, with a pH of around 8.5 and low in nitrogen and organic matter.

 Plant material

     Eighteen accessions of sorghum collected from Gedarif and from the gene bank (Wad Medani) were used in this study. These accessions were five released varieties (Wad-Ahmed, Tabat, Butana, Bashayer and Arffagadamak-8), and 13 local land races preferred by farmers (Korakollo, Mugod, Saffra, Wad-Bako, Tetron, Faki-Mustahi, Farhoda, Gadambalia bloom, Ajeb-seido, Arafah, Gesheish,Wad-fahal and Milo) .

 Cultural practices

     The standard cultural practices adopted for sorghum at the  ARC were followed. Land was prepared by disc ploughing, disc- harrowing, leveling and ridging in irrigated site and by disc- harrowing  in rain-fed sites.  Treatments were laid out in a randomized complete block design with four replicates in the different locations and seasons. Sowing was done in the

 

 

 

first week of July under irrigation and the first to the third week of July under rainfed conditions depending on the onset of rainfall. Under irrigation, the entries were sown in five rows, 5 m long on ridges; 0.8 m apart at 0.3 m intra - row spacing and thinned to two seedlings per hill. Under rainfed conditions, they were also sown in five rows 5 m long on flat; 0.8 m apart at 0.2 m intra row spacing and thinned to two seedlings per hill. Urea at the  rates of 80 kg and 40 kg /fed was applied under irrigation and rainfed sites, respectively, as recommended by the ARC. The crop was irrigated every two weeks or whenever necessary and irrigation was withheld three weeks before harvest. In irrigated and rainfed  experiments, assessments were made in the central three rows of the plot discarding one row or more at each side. Data were collected on days to 50% flowering, plant height, number of heads/m², head length (cm), head width (cm),1000 seed weight(g) and grain yield (kg/ha). 

Statistical analysis 

    The analysis of variance procedure was used to test differences among genotypes within each season, location and combined. Eberhart and Russell (1966) stability model was performed. In addition, the Additive Main Effect and Multiplicative Interaction (AMMI)  was carried out to show the stability and pattern of adaptation of sorghum genotypes in nine environments, using IRRISTAT(2005) statistical analysis package for grain yield data.

 

RESULTS AND DISCUSSION

 

    The combined analysis of variance showed highly significant differences among  seasons for all the traits studied with the exception of head length (Table 1).It also showed that differences among locations were highly significant for all traits under study. Differences among genotypes were highly significant for all traits with the exception of number of plants/m²  and head length. The interaction effect of genotype x location was highly significant for most traits except  number of plants /m² and number of heads /m² and this may be due to genetic factors.

    The significance of genotype x environment indicated that genotypes responded differently to environments and some are environmentally specific. Also, this finding indicated the importance of these components in affecting the phenotypic performance of the evaluated genotypes in the different environments. Similar results were reported by Abdalla et al.

 

 (2009), who found that the mean squares of genotypes, environments and genotypes x environments interactions are highly significant (P=0.01). Also Elasha et al.(2011) found significant differences between environments, genotypes and environment x genotype.

     Genotypes significantly interacted with seasons for almost all traits except number of plants/m² and head width and this may be due to genetic factors. However the significant interactions of  genotypes with seasons shown by all of the characters studied reflect their instability over seasons. Similar results were reported by Shivanna et al.(1992) and Santos et al. (1995). The second degree interaction of season x location x genotype was significant for all traits except for number of plants/m². Kambal and Mahmoud (1978) reported that variety x year interaction was small and not significant, while the variety x location and variety x location x year interaction were highly significant in sorghum.

     The current findings indicated that there is a wide range of genetic variability among tested genotypes, which could be attributed to both genetic and environmental factors and their interactions. Similar results were reported by Hashim (2008) and  Bello et al (2007) who studied genetic variability in sorghum and reported significant differences among cultivars for days to 50% flowering, plant height, 1000- seed weight and grain yield .

     Shinde and Jagadshwar (1986) in F1 and F2 generations of 8x8 diallel cross evaluated for grain yield and yield components in three environments showed significant genotype x environment interaction for all studied  traits. From the present study, and on  the basis of the importance of genotype x environment  interactions  as shown it could be concluded  that  sorghum genotypes show differential responses when grown under different environments, suggesting that these genotypes should be tested  in different environments.

 

 

 

 

 

 

 

 

 

 

 

S X LXG

S X G

L X G

Genotype (G)

Location (L)

Season (S)

Trait

141**

245*

126**

1274**

7091**

2068**

  DF

 

1775**

2333**

2057**

1323**

246696**

19500**

  PH

 

7.93 ns

9.11ns

9.60 ns

7.84 ns

909**

55**

  P/m

 

10.30*

13.3**

7.67 ns

28.7**

4096**

83.9**

  H/m

 

22.83**

56.23**

18.84**

29834 ns

15.5**

573 ns

  HL

 

1.5**

1.13 ns

1.8**

2.5**

75.2**

74.8**

  HW

 

46**

55**

113**

423**

7251**

660**

  Sw/

 

143870**

165846**

165846**

166993**

7349446**

24873664**

  GY

 

 Table 1. Means for seasons , locations ,genotypes and their interactions for18 sorghum genotypes combined over three seasons and three locations, grown at North Gedarif, South Gedarif, and Rahad Research farm (RRF) during season 2009,2010,and 2011.

*,** Significant at 0.05and 0.01 of probability levels, respectively; ns=not  significant.         

DF= days to 50% flowering, PH= plant height (cm), P/m= number of plants /m²,        

H/m²=number of heads /m²,HL= head length (cm), HW=head width (cm),Sw (g)           

= 1000  seed weight (g)  , GY=grain yield (kg/ha).

 

Grain yield stability

    The data on the three stability parameters, mean performance, regression coefficient(bi) and deviation from regression (S²d) for grain yield are presented according to Eberhart and Russell (1966) stability model      (Table 2). The mean grain yields of sorghum genotypes ranged from 846 kg/ha as minimum to the 1510 kg/ha as maximum , with an average of 1302 kg/ha. Seven genotypes recorded higher yield than the mean of all genotypes (Table 2).These  genotypes were Tetron (1323 kg/ha), Butana (1333 kg/ha), Safra (1401 kg/ha), Gadambalia bloom (1428 kg/ha), Wad-Ahmed (1471kg/ha), Bashaiyer (1503 kg/ha), and Mugod (1510 kg/ha).

 

 

 

 

 

 

     Genotypes with bi > 1 and  mean grain yield greater than the  general mean, were Mugod, Safra, Tetron, W-Ahmed and Gadambalia bloom indicating that they were more responsive to environmental changes and, therefore, suitable for favorable environments of irrigation  conditions (Rahad) and high rainfall conditions (South Gedarif).

    These findings  agreed with those reported by  Elasha et al. (2011) who studied stability and adaptability of seven hybrids and three open pollinated varieties under twelve environments. They found that the genotypes DIA-07666, DMN 15P 1003, PAC-501 and E-1 showed slopes (bi) of 2.67, 2.49, 2.34 and 1.18 with deviation from regression of 0.08, 0.45, 0.68 and 0.12 under irrigation, respectively, and a mean grain yield above the general mean of the traits meaning those are more adaptable under irrigation conditions.

    Genotypes with (bi) close to 1.0 but low yielding (below the general mean),and so quite stable with relatively small S2d were Korakollo,Wad Baku, Farhoda, Gesheish, and Wad Fahal (1272, 1225, 1252,1194 and 1230kg/ha, respectively).This means that these genotypes have better response under unfavorable environments and are, therefore, stable and adaptable. Similar findings were reported by Abdalla et al.(2009) who studied stability and adaptability in some sorghum lines grown under nine environments . They found that genotypes, AG15 and AG8 had b values of 0.900 and 0.928, repectively,The Genotypes, AG15 and AG-8 also had mean grain yield of 894.35 and 862.32 kg/fed respectively, and Wad Ahmed had 862 kg/fed which were above the overall mean of 777.73 kg/fed of the trials, while CAG had a mean of 543kg/fed, which was lower than the overall mean. This means that both genotypes have better response in unfavorable environments and are, therefore, adaptable stable and predictable (high R2 value) than the two checks. Similar results were also reported by Elzein et al. (2008) who studied stability and adaptability in some sorghum lines and they found regression coefficients greater than one and  had higher(S2d) observed for Gew  22-15 and Gew 3-2 with mean grain yield below the general mean yield, indicating that these two lines were not stable under adverse conditions but may respond better to favorable environments.

     The most stable genotypes as indicated by this stability parameter were Mugod, Tabat, Gadambalia bloom, Safra, Wad Ahmed and Tetron when the mean yield, regression coefficient and the deviation from regression were considered together.

 

 Table 2 .Stability parameters for grain yield (kg/ha) of 18 sorghum genotypes tested  at North Gedarif, South Gedarif, and Rahad during 2009,2010,and 2011 growing seasons.

Genotypes

Yield (kg/ha)

bi

S²d

Korakollo

1272

1.07

4.7

Mugod

1510

1.58

11.4

Safra

1401

1.16

5.5

Wad Baku

1225

0.96

2.3

Tetron

1323

1.14

9.0

Faki Mustahi

846

0.75

2.3

Farhoda

1252

1.01

1.0

Gadambalia bloom

1428

1.18

5.7

Ajeb seido

1261

0.83

2.8

Arafa

1298

0.73

3.3

AG-8

1214

0.59

1.4

Butana

1333

0.97

1.2

Bashayier

1503

0.94

8.4

Tabat

1299

1.16

2.5

Wad Ahmed

1471

1.26

3.2

Gesheish

1194

0.92

5.5

Wad Fahal

1230

0.95

6.6

Milo

1286

0.72

3.5

Mean

1302

 

 

bi =slopes of regression , S²d =Deviations from regression.

 

     In the present study, multivariate analysis such as AMMI analysis groups genotype or environments in a qualitative manner according to their similarity of performance rather than quantitative manner of the stability parameters. AMMI analysis involves the clustering analysis to classify genotypes under the most adapted sites for them depending on the AMMI principle components scores (Gauch and Zobel,1988;Nachit et al. 1992). The combined analysis of variance according to the AMMI model is presented in Table 3.

    The partitioning of GE interaction through AMMI model analysis revealed that the four multiplicative terms (PCA1, PCA2, PCA3, and PCA4) were significant and were captured 56.7%, 19.3%, 10.1%, and 7.2% of variation due to GE interaction sum of squares, respectively. Together they accounted for 93.3% of GE interaction sum of squares. However, most of the variation was explained by the first principle components (PCA1).

 

According to  Crossa et al. (1990), AMMI with two, three or four PCA1 axes is the best predictive model. Similarly, in the present study, the AMMI analysis further revealed that the first two interaction principle component axes (PCA1 and  PCA2) explained 76% of the GxE sum of squares. This was in agreement with Sneller et al., (1997),who suggested that GxE pattern is collected in the first principal components of analysis.

 

Table 3. AMMI analysis of variance of the significant effects of genotypes (G), and  environment (E) and genotype- environment interaction (GE) on grain yield  (kg/ha) and the partitioning of the GE into AMMI scores.

Source of variation

DF

SS

MS

 Efficiency (%)

Environment (E)

Genotypes (G)

GE I

PCA1

PCA2

PCA3

PCA4

Residual

8

17

136

24

22

20

18

52

0.25367E+0.8

706700

0.604554E +0.7

0.342852E +0.7

0.116852 E +0.7

615978

392181

440344

0.317095E+0.7

41570.6

44452.5

142855***

53114.5***

30798.9**

21787.8**

 

 

 

100

56.7

19.3

10.1

07.2

  **,*** Significant at the 0.01 and 0.001 probability levels, respectively.

  DF, degree of freedom; SS sum of square, MS  mean square and Efficiency % of GE sum  

  of squares.

 

     Variation among the studied genotypes for grain yield and their reactions to the environments were determined (Table 4). The highest average yield was obtained in E-7 followed by the E-9 (representing Rahad environment), whereas E-1 (representing North Gedarif environment) had  obtained the lowest grain yield. E-7 exhibited the  largest absolute PCA1 score (i.e. had the highest interaction effect), whereas the smallest score was shown by the E-4 ( representing  South Gedarif environment) (i.e. had the least interaction effects). Based on AMMI biplot, G and E having PCA values close to zero have small interaction effects, whereas those having large positive or negative PCA absolute values largely contribute to GE interaction. Hence, E-7 was the most interactive, while E-4 was the least interactive among the nine environments.

 

 

 

 

 

Table 4. PCA1 and PCA2 scores for the nine growing environments

of sorghum genotypes.  

Environment

E-Mean

IPCAe (1)

IPCAe (2)

E1

113.8

3.51928

6.33557

E2

144.3

3.39088

6.43972

E3

310.3

5.28205

-4.16294

E4

631.2

-3.50394

4.74171

E5

671.1

5.81292

17.75241

E6

475.3

7.83628

-1.66849

E7

1215.9

-39.053

-5.03838

E8

191.4

4.84558

0.39725

E9

1184.2

11.86991

-24.7968

  E1,E2,E3 (North Gedarif ),E4,E5,E6 (South Gedarif),E7,E8,E9 (Rahad).

 

    To analyze genotype-environment interaction and adaptation graphically, AMMI biplot was used with the PCA score plotted against the mean yields (main effects).

    A graphical display of the GE interaction of PCA1 and their effects (yields) is useful for revealing favorable pattern in genotypes response across environments (Crossa et al.1990). The AMMI bi-plot of mean on yield explained a large proportion of the treatment sum of squares. The PCA scores, negative or positive, more specific or  adaptive genotype to certain environments. The more  PCA score approximate to zero, the more stable or adapted genotype over all environments. Accordingly, the genotypes Mugod, Safra ,Tetron, Gadambalia bloom, Butana ,and Bashaiyer revealed good stability across environments and high grain yields. This indicated that these genotypes Butana,and Bashaiyer were stable over all environments, while the genotypes W-Ahmed,Tabat, Mugod, and Safra were adapted for specific environments. W-Ahmed and Tabat for favorable environment, while Mugod and Safra were adapted for specific  environment   (South Gedarif environments). Genotype Mugod exhibited high yield in the environment 6 which represent South Gedarif environment,followed by the genotypes Gadambalia bloom,Safra, and genotype Tetron, respectively. (Fig 1).

 

 

 

 

 

 

       To further explain the GE and stability,a bi-plot between the PCA1 and PCA2 scores were given in Fig2. AMMI bi-plot of the first two principle component axes is a  powerful way of detecting important score of GE effects (Zobel et al.1988).This analysis represents stability of the genotypes across environments in terms of principle component analysis. It is used to identify broadly adapted genotypes that offer stable performance across sites, as well as genotypes that perform well under specific conditions. In this study, the first two principal component axes (PCA1 and PCA2) in bi-plot analysis explained a large proportion of the variation 76% of the total GE sum of squares (Table 3).On this AMMI bi-plot ,genotypes and environment ِhaving PCA values close to zero (near the origin) have small interaction effects, whereas those having large positive or negative PCA values (distant from zero) largely contribute to GE interaction  (Yau,1995). Hence, the genotypes Butana, Farhoda, Faki-Mustahi, ashaiyer, Gadambalia bloom, Safra, and Wad baku were the most interactive ,while the genotypes W-Ahmed, Tabat, Wad Fahal,and Gesheish were the least interactive. On the other hand, environments E-9 and E-6 appeared we distant from the origin (large PCA score), hence they had large interaction effects, whereas E-2 had small interaction effects (Fig.2). Genotypes Tabat,W-Ahmed, and Wad Fahal were more stable and responsive for good environments (Rahad environment),while the genotypes Mugod, Tabat ,Wad-Ahmed, Safra, and Tetron  were responsive and suitable for South Gedarif environment. Hence, in this investigation, visual observations of AMMI bi-plot analysis enable the identification of  genotypes and testing environments that exhibited major sources of GE interaction as well as those that were stable. Similar results were reported by Sneller et al (1997). From the result shown in Table 4 and Fig 2, it was found that the genotypes Mugod, Wad-Ahmed, Tabat, Gadambalia bloom, Safra and Tetron were high yielding and stable under favorable environments, and they could  be grown under high rainfall and Rahad conditions. The others (Wad Baku, Farhoda, Gesheish and Wad Fahal) were quite stable under  unfavorable conditions, and it could be  grown under low rainfall conditions of North Gedarif. In this study, comparing the effectiveness of joint regression and AMMI analysis for analyzing GE interaction, it was found that PCA1 in AMMI accounted for the GE sum of squares by 56.7%, while regression analysis accounted forGE sum of squares  by 21.9%. Hence, AMMI analysis was superior to regression techniques in accounting for GE sum of squares and more effective in partitioning the interaction sum of squares.

 

 

    From these two models of stability used in this study, it was found that the genotypes Mugod, Wad-Ahmed, Tabat, Gadambalia bloom, Safra and Tetron were high yielding and stable under favorable environment, and could  be grown under high rainfall and Rahad conditions, others (Wad Baku, Farhoda, Gesheish and Wad Fahal) were quite stable under unfavorable conditions, and could be  grown under low rainfall conditions of North Gedarif.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CONCLUSION

 

    Based on the results of  this study, it could be concluded that the genotypes Mugod,Tabat, Wad-Ahmed, Gadambalia bloom, Safra and Tetron were high yielding and stable under favorable environment, and  could be grown under high rainfall and Rahad irrigation conditions. Genotypes Wad Baku, Farhoda, Gesheish and Wad Fahal were quite stable under unfavorable conditions, and could be grown under low rainfall conditions of North Gedarif. Further testing of the unsuitable genotypes is necessary for further breeding manipulations. Both parametric and non- parametric approaches of stability analysis (Eberhart and Russell as well as AMMI) agreed in identifying stable genotypes over different environments.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

REFERENCES

 

Abdalla, H. M, Y. A. Gamar, A. H. Abu-Assar, T. Y. Elagib. M. H. Elgada, and O. M. Elhassan. 2009. A proposal for the release of two early maturing,high yielding and drought tolerant sorghum genotypes, ARC , Wad Medani , Sudan.                

Bello, B., A.M. Kadmas, S.Y.Simon and D.S. Mashi.2007. Studies on genetic variability in cultivated sorghum. American-Eurasian Journal of  Agriculture and Environmental Science 2(3):297-302.

Crossa, J., H.G. Gauch, and  R.W. Zobel.1990. Additive main effects and  multiplicative interaction analysis of two international maize  cultivar trials. Crop Science 30:493-500.

Eberhart, S.A. and W.A. Russel. 1966. Stability parameters for comparing crop varieties. Crop Science 6: 36-40.

Elasha, A., I. N. Elzein, A. H. A. Assar, M. K. Hassan, A. E. Hassan, O. M. Alhassan, A. A. Elmustafa and H. A. Hassan 2011. A proposal for sorghum (Sorghum  bicolor (L.) Moench) hybrids release for irrigated and rain- fed sectors of the Sudan.National Variety Release Committee, Khartoum , Sudan.

 Elzein, I. N., E. I. Hassan, A. M. Ali, A. B. Elahmadi, E. A. Elasha and T. Eltaib. 2008. A proposal for the release of short maturing sorghum genotypes for drought prone areas of the Sudan. National Variety Release Committee, Khartoum, Sudan.

Flores, F., M.T. Moreno and J.I. Cubero.1998. A comparison of univariate and multivariate methods to analyze G x E interaction. Field Crops Research 56: 271–286.

Gauch, H.G. 1988. Model selection and validation for yield trials with interaction. Biometrics 44:705-715.

Gauch, H. G. and R.W.Zobel.1988. Predictive and postidictive success of statistical analysis of yield trials. Theoretical and Applied Genetics 76: 1-10.

Hashim,A.A.2008. Tapping the Sudanese Sorghum Germplasm for Grain Yield Stability and Tolerance to the Spotted Stem borer Damage. Ph.D. Thesis, University of Gezira, Wad Medani, Sudan.

Hussein, M.A, A. Bjornstad  and  A.H. Astaveit. 2000. SASG x ESTAB : a SAS program for computing genotype x environment stability statistics. Agronomy Journal 92:454-459.

 

ICRISAT (International Crops Research Institute for the Semi – Arid Tropics) and FAO (Food and Agriculture Organization of the United Nations ).1996. Part I, Sorghum, pp. 5-27. In: The World  Sorghum and Millet Economics: Facts, Trends and Outlook. Rome.

IRRISTAT,2005.IRRISTAT for windows © 1998-2005. International Rice Research Institute, DAPO, Box 7777, Metro Manila, Philippines.

Ishag, H.M. and, O.A. Ageeb.1987. Agricultural research /extension linkage in Sudan. Workshop on Planning and Programming of Agricultural Research. ARC/ISNAR, Wad Medani, Sudan, 30 Nov. – 2 Dec., 1987. 

 Kambal, A.E. and M.A.Mahmoud.1978.Genotype X environment interactions in sorghum variety test in the Sudan Central Rain lands. Experimental Agriculture 14:41-48.

Kang. M.S. and R. Magari. 1996. Stable: basic program for calculating stability and yield-stability statistics. Agronomy Journal 87: 276-277.

Kempton,R.A.1984.The use of biplots in interoperating variety by environment interactions. Journal of Agricultural Science 103: 123-135.

Lin, C. S; M.R. Binnas and L.P.lefkovitch.1986.Stability analysis: Where do we stand? Crop Science 26:894-900.

Nachit, M.M.G, Nachit, H.K, Guach, and R.W. Zobel.1992. Use of AMMI and liner regression models to analyze  genotype x environment interaction in durum wheat. Theoretical and Applied genetics 83:597-601.       

 Osman, E.I.,I.N Elzein, E.A.Babikir, and I. A. Suliman. 1996. Evaluation of the improved sorghum varieties and hybrids for yield potential, stability and quality, under Sudan irrigated and rainfed conditions. Proceedings of Sudan National Variety Release Committee, ARC, Wad Medani, Sudan.

 Osman. E .I and M.A. Mahmoud. 1992.  Improved sorghum genotypes suitable for irrigated and rain-fed land of  Sudan; Proceedings of Sudan  National Variety Release Committee. Wad Medani, Sudan.

Robert, N. 2002. Comparison of stability statistics for yield and quality traits in bread wheat. Euphytica 128:333-341.

 

 

 

 

 

Santos, J.P.O., G.A. Manciel, M.R.N. Araujo and J.N. Genbosa.1995. Genotype x environment interaction in grain sorghum hybrids. Brazil, ISMN 36: 69-70.

Shinde, V.K. and K. Jagadeshwar. 1986. Genetic analysis of yield in rabi sorghum . Sorghum Newsletter 29:1-13.

Shivanna, H., S. S. Patil and R. Parmeshwarappa.1992. Assessment of  general combining ability of sorghum cultivars over different environments. Journal of Maharashtra Agricultural Universities 17 : 216-218.

Sneller,C. H., L. Kilgore-Norquest and D. Dombek. 1997. Repeatability of yield stability statistics in soybean. Crop Science 37:383-390.

Westcoff, B. 1987. A method of analysis of  the yield stability of  crops. Journal of   Agricultural Science 40: 1-13.

Yau,S.K.1995. Regression and AMMI analysis of genotype x environment interaction: An empirical comparison. Agronomy Journal 87:121-126.

Zobel, R.W.,M.J.Wright, and H.G.Gauch.1988.Statistical analysis of a yield trial. Agronomy Journal 80: 388-393.

 

 

 

 

 

 

 

 

 

 

 

 

 

published in Gezira Journal of agricultural science

  • Performance, genetic variations and interrelationships in different traits of sorghum (Sorghum bicolor L. Moench) genotypes

 

Performance, genetic variations and interrelationships in  different traits of  sorghum  (Sorghum bicolor L. Moench) genotypes

Mohammed H. Mohammed¹, Abu Elhassan S. Ibrahim² and Ibrahim N. Elzein¹.

¹ Agricultural Research Corporation, Wad Medani, Sudan.

² Faculty of Agricultural Sciences, University of Gezira, Wad Medani,  

   Sudan.

ABSTRACT

 

   Eighteen sorghum genotypes were evaluated over three consecutive seasons (2009,2010,and 2011) at three locations ( Rahad Research farm of the Agricultural Research Corporation (ARC), Sudan, Gedarif Research Station farm North Gedarif and South Gedarif region). Both experiments conducted in  North and South Gedarif were rainfed, while that conducted at Rahad station was irrigated. A randomized complete block design with four replicates was used .The objective was to estimate the general performance, genetic variability and interrelationships of grain yield and its components. Data were collected on days to 50% flowering, plant height, number of heads/m², head length (cm), head width (cm),1000 seed weight (g) and grain yield (kg/ha). The study found  that there were highly significant differences among genotypes in  all the characters studied  except head width in season 2011.The early maturing genotypes were Milo (59-64 days), Gesheish      (60-67 days) , AG-8 (59-65 days) and Butana (62-68 days), an indication that these genotypes would fit quite well in short rainy seasons of  North Gedarif environment. The late maturing genotypes were Tabat (68-83days),Wad Ahmed (69-83 days), Faki Mustahi (68-88 days) and Tetron (73-88 days) which were suitable to be grown under Rahad and South Gedarif environments. The highest grain yields (kg/ha) were exhibited by the genotypes Butana (735 kg/ha), Wad Ahmed (2572 kg/ha), and Mugod   (2545 kg/ha). Grain yield was positively   and highly significantly correlated with head width (0.65**)  and number of heads/m² (0.46**) .Accordingly, the simple linear correlation and path coefficient analysis indicated that head width and number of heads/m² could be used as potential selection criteria in breeding programs for developing high yielding sorghum genotypes.

 

 

INTRODUCTION

 

    Sorghum (Sorghum bicolor (L.) Moench) is one of the major cereal crops of the semi-arid tropics. It is the fifth most important cereal crop of the world. Major producers of sorghum in the world are USA, India, Nigeria, China, Mexico, Sudan and Argentina. Twenty one percent of the world sorghum area is in India. In the Sudan, sorghum is the most widely produced and consumed cereal crop. It is utilized in various forms as stable food for humans, feed for animals and contributes  about 70% of total grain produced in the country. It ranks first in total area cultivated as well as total tonnage produced. However, the areas as well as production vary year after year due to many biotic, abiotic, technical and policy factors. The area is reported to be 4-8 million hectares with an average of 5.5 million/ha, about 90% of it is under rain; while total grain production varies between 3-4.5 million tonnes with an average of 0.6 tonnes/ha. Of the abiotic factors limiting sorghum productivity, rainfall stands out as the most important factor. The climatic change seriously affected the traditional sorghum growing areas of northern Gedarif, Gezira, Sennar, White Nile States as well as northern parts of Kordofan and Darfur States. This area is estimated to be > 50% of the total sorghum production area. (Elzein et al, 2008).

    In these dry areas (250 mm– 400 mm), farmers used to grow their own local sorghums, which are low yielders and suffer drought stress at almost all stages of crop growth. The outcome is either low yield or straw and chaff. In fact, sorghum is loosing ground in these important areas. The improved, medium maturing, high yielding varieties and hybrids such as Feterita , Wad Ahmed , Hageen Dura-1 and Tabat require 550 mm– 650 mm which is not available and accordingly these varieties/hybrids, were not recommended  for these low rainfall regions. (Osman and Mahmoud, 1992). The short maturing varieties released earlier by ARC, such as Umbenien 7, 11, 22, feterita Maatuog, etc, were out of cultivation due to  their poor grain quality, small seed, pigmented  seed coat and  hence low market value.

    Progress in plant breeding depends on the extent of genetic variability present in a population that permits effective selection procedures, based on locally adapted land races (Swarp and Chaugale,1962). Therefore, the first step in any plant breeding program is the study of genetic variability, which cannot easily be measured. The ultimate objective of most sorghum breeding programs is to improve yield which is genetically a complex character, that requires a reasonable level of genetic diversity (Sprague, 1966).

 

   Correlation studies are important in breeding programs, as they give information on direction and magnitude of association between different traits. This could be utilized to select for one character indirectly by selection for another one (Sharaan and Ghallab,1997). One of the objectives of the sorghum breeding program in Agricultural Research Corporation (ARC) of the Sudan is to increase productivity and sustainability of sorghum production in irrigated  and  low rainfall regions of the country and thereby making better use of natural resources.

    The present study consists of eighteen sorghum genotypes to be evaluated under different environments (rainfed and irrigation conditions) to contribute to sorghum improvement in the Sudan. The objectives of this study were to evaluate  the performance for yield potential, the extent of genetic variability, interrelationships in nine different growing environments of sorghum in the Sudan.

 

MATERIALS AND METHODS

Location

       The experiments were conducted over three consecutive seasons   (2009, 2010, and 2011) at three locations,viz. Rahad  Research Farm, North and South Gedarif regions of the Gedarif Research Station farm.The three locations lied within the central clay plain of the Sudan characterized by heavy, alkaline clay soil, with a pH of around 8.5 and poor in nitrogen and organic matter.

 Plant material

     Eighteen accessions of sorghum collected from Gedarif and from the gene bank (Wad Medani) were used in this study. Five of these accessions (Wad-Ahmed, Tabat, Butana, Bashayer and Arffagadamak-8). were released varieties and 13 local land races preferred by farmers Korakollo, Mugod, Saffra, Wad-Bako, Tetron, Faki-Mustahi, Farhoda, Gadambalia bloom, Ajeb-seido, Arafah, Gesheish,Wad-fahal and Milo.

 

 

 

 

 

 

 

 

Cultural practices

    The standard cultural practices adopted for sorghum production at ARC were followed. Land was prepared by disc ploughing, disc- harrowing, leveling and ridging in irrigated site and by wide level disc  in rain-fed sites.  Treatments were laid out in a randomized complete block design with four replicates in the different locations and seasons.

    Sowing was done in the first week of July under irrigation and the first to the third week of July under rainfed conditions depending on the onset of rainfall. Under irrigation, the entries were sown in  five rows, 5 m long on ridges; 0.8 m apart at 0.3 m intra - row spacing and thinned to two seedlings per hill. Under rain fed conditions, they were also sown in  five rows 5 m long, on flat; 0.8 m apart at 0.2 m intra row spacing and thinned to two seedlings per hill. Urea at the rates of 80 kg and 40 kg /fed was applied under irrigation and rain-fed sites, respectively, as recommended by the ARC. The crop was irrigated every two weeks or whenever necessary and irrigation was withheld three weeks before harvest.

    In irrigated and rainfed  experiments, assessments were made in the central three rows of the plot discarding one row or more on each side. The data were collected on days to 50% flowering, plant height, number of heads/m², head length (cm),head width (cm),1000 seed weight (g) and grain yield (kg/ha).                                              

 Statistical analysis

     Analysis of variance was performed for each season; location and combined  to test for significant differences among genotypes. Means for seasons were used to compute simple linear correlation coefficients between all possible combinations. The path coefficients procedure was used in order to partition correlation coefficients between grain yield and its components which is divided into direct and indirect effects.

 

 

 

 

 

 

 

 

 

 

 

RESULTS AND DISCUSSION

Mean performance

Days to 50%flowering

    This trait is used as an earliness index. Across locations, it showed significant differences among genotypes under the three locations       (Table 1).The highest general mean was observed at season 2010 while the lowest general mean was obtained in season 2011. The range for days to 50% flowering was 59 (Milo) to 83 (Wad Ahmed)  in 2009, from 60 (Gesheish) to 88 days (Faki Mustahi and Tetron) in 2010, and from 59   (AG-8) to 74 days (Tetron) in 2011.                          

    Identifying early and medium maturing genotypes is important for choosing genotypes to suit the different growing environments (irrigation and rainfed). Hence, from these findings, the early maturing genotypes were Milo, Gesheish , AG-8 and Butana, an  indication that these lines would fit quite well in short rainy seasons, i.e suitable for North Gedarif environment, while the late maturing ones were Tabat, Wad Ahmed, Faki Mustahi and Tetron which were suitable to grow under Rahad and South Gedarif environments (Table 1). These findings were in agreement with those of  Abdalla et al. (2009), who reported that lines AG-8 and AG-15 were 18 days and 14 days earlier than Wad-Ahmed. Elzein et al. (2008) found a wide range of variability in days to 50% flowering

Plant height (cm)

    Development of short, medium genotypes is  important for any plant breeding program, because these genotypes will be  suitable for mechanical harvesting and for resistance to lodging. Across locations, plant height showed significant differences among genotypes under the  three locations  (Table 1). The highest general means were observed in season 2010, while the lowest general means were observed in season 2011. The range for plant height was 112 cm (Tabat) to 189 cm (Wad Baku) in 2009, from 137cm (Butana) to 216 cm (Tetron) in 2010 and  from 97 cm (Bashaiyer and AG-8) to 139 cm (Tetron)  in 2011.Thus, in this study, the short genotypes were Tabat, Butana , Bashaiyer and AG-8, while the tall genotypes were Wad Baku and Tetron (Table 1). From these results, tall genotypes such as Wad Baku and Tetron are not appropriate for drought areas such as North Gedarif , because they were susceptible to drought, while Butana and Bashayier were suitable to grow under North Gedarif conditions. These findings were in agreement with Elasha et al (2011) , who found significant

 

differences (P< 0.01) between the entries in their plant height. Also Bushara (1999) , recorded highly significant differences in plant height among

hybrids of grain sorghum, in the Sudan.

 

Table 1. Means of days to 50% flowering and plant height (cm) for 18 sorghum genotypes grown at North Gedarif (NG),South Gedarif (SG),and Rahad (RH),seasons 2009,2010, 2011.

Genotypes

Days to 50% flowering

Plant height (cm)

2009

2010

2011

Mean

2009

2010

2011

Mean

Korakollo

70

67

66

67.7

171

162

110

147.6

Mugod

74

77

69

73.3

153

179

119

150.3

Safra

64

65

68

65.6

177

187

119

161.0

Wad Baku

60

69

73

67.3

189

209

119

172.3

Tetron

73

88

74

78.3

181

216

139

178.6

Faki Mustahi

68

88

68

74.6

177

185

128

163.3

Farhoda

80

83

73

78.7

175

187

123

161.6

Gadambaliabloom

71

66

69

68.7

168

154

110

144.0

Ajeb seido

69

71

73

71.0

143

155

110

136.0

Arafa

82

83

73

79.3

171

176

114

153.6

AG-8

60

65

59

61.3

125

144

97

122.0

Butana

62

68

68

66.o

119

137

104

120.0

Bashayier

72

74

67

71.0

118

141

97

118.6

Tabat

83

84

68

78.3

112

142

98

117.3

Wad ahmed

83

82

69

78.0

116

147

108

123.6

Gesheish

65

60

67

64.0

143

159

105

135.6

Wad Fahal

73

80

70

74.0

182

151

108

147.0

Milo

59

63

64

62.0

134

149

112

131.6

Mean

70

74

69

71.0

153

166

112

143.7

CV%

6.8

2.8

6.5

5.3

6.8

21.7

10.6

13.1

SE±

1.39

0.60

1.36

1.12

2.9

10.36

4.48

5.91

 

**

**

**

 

**

**

**

 

** Significant at 0.01 of probability level.

                                       

Number of heads /m²

    This character is an indicator for high grain yield. Good crop establishment will result in increasing  number of heads/m² which lead to an increase in grain yield. Across locations, this trait showed  highly significant differences among the genotypes in the three locations (Table 2). The highest general mean was  observed in season 2010, while the lowest

 

 

general mean was observed in season 2011. The range for this trait  was from 6  (Wad Fahal, Mugod, Tabat) to 11 (Wad Ahmed) in season 2009,  from 8 (Wad Fahal)  to 13  (Wad Baku) in season 2010, from 5 (Arafa and Faki Mustahi) to 9 (Korakollo) in season 2011 (Table 2 ). In this study, the highest numbers of heads/m² were obtained by the genotypes Wad Ahmed, Wad Baku, and Korakollo, while the lowest number of heads/m² were obtained by the genotypes Wad Fahal, Tabat and Arafa. This is because Wad Fahal  and Arafa are late in maturity and some plants failed to produce heads due to moisture stress.

Head length (cm)

   Highly significant differences among genotypes were observed for this trait under the  three locations (Table 2). Across locations, the highest general mean(19 cm) was observed in season 2010, while the lowest general mean (14 cm) was observed in season 2011. The range for this trait was from 13 cm (Mugod, Farhoda and Wad Baku) to 24 cm (Faki Mustahi) in season 2009, from 12 cm (Mugod) to 27 cm (Tetron) in season 2010, from 12 cm (Safra and Farhoda) to 18 cm (Tabat) in season 2011. In this study, the longest heads were obtained by the genotypes Faki Mustahi ,Tetron, and Tabat ,while the shortest heads were obtained by the genotypes Mugod and Safra (Table 2 ).This result agreed with that reported by  Elasha et al (2011),who found that during both seasons at the irrigated and the rainfed sites, there were significant differences (P< 0.01) between the entries in their panicle length.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 2. Means of number of heads/m² and head length (cm) for 18 Sorghum genotypes grown at North Gedarif (NG),South Gedarif (SG),and

Rahad (RH), season 2009, 2010, 2011.

                                      

No. of heads/m²

Head length (cm)

Genotypes

2009

2010

2011

Mean

2009

2010

2011

Mean

Korakollo

8

11

9

9

17

18

15

16

Mugod

6

11

7

8

13

12

14

13

Safra

8

11

7

8

17

14

12

14

Wad Baku

8

13

8

9

13

14

14

13

Tetron

7

10

6

8

17

27

17

20

Faki Mustahi

10

9

5

8

24

26

17

22

Farhoda

7

10

6

8

13

13

12

12

Gadambaliabloom

9

11

6

8

16

17

13

15

Ajeb seido

9

11

8

9

15

19

13

15

Arafa

8

9

5

7

20

19

15

18

AG-8

10

10

7

9

15

18

14

15

Butana

9

10

7

9

23

24

15

20

Bashayier

7

11

7

8

22

20

14

18

Tabat

6

9

7

7

20

22

18

20

Wad ahmed

11

11

8

10

17

18

15

16

Gesheish

7

11

8

8

19

17

14

16

Wad Fahal

6

8

7

7

21

20

14

18

Milo

9

11

7

9

16

17

13

15

Mean

8

10

7

8

18

19

14

17

CV%

29.6

23.9

8.5

20.6

12.4

22.0

14.6

16.6

SE±

0.67

0.72

0.71

6.7

0.62

1.18

0.80

0.86

 

**

**

**

 

**

**

**

 

** Significant at 0.01 of probability level.

 

Head width (cm)

    Across locations, this trait showed highly significant differences among genotypes except for season 2011. The highest general mean (5cm) was observed in season 2010, while the lowest general mean (3cm) was observed in season 2011(Table 3). The range for head width varied  from 3 cm (Faki Mustahi) to 4 cm (Wad Fahal), from 4 cm (Faki Mustahi)  to 6cm (Butana and Gesheish), from 3cm (Bashaiyer) to 4cm (Mugod) in seasons 2009, 2010 , and 2011, respectively (Table 3). From this study,  genotype Mugod had the  largest head width coupled with the  longest head length. This means that tall, late maturing genotypes are not suitable for drought areas such as  North Gedarif but suitable for South Gedarif and Rahad environments.

 

1000 seed weight (g)

    Across  locations, highly significant differences among genotypes  were observed for this trait.The highest general mean (31g) was observed in season 2010,while the lowest general mean (25 g) was obtained in seasons 2009 and 2011(Table 3). The range  for 1000 -seed weight (g)  varied from 18 g (Butana)  to 36 g (Mugod), from 25g (Tetron and Wad Ahmed) to 42g (Wad Fahal) , from 21g (Wad Ahmed) to 31g (Wad Fahal) in season 2009, 2010, and 2011, respectively. Hence, in this study, the highest 1000 seed weight was obtained by the genotypes Mugod and Wad-fahal,while the lowest 1000 seed weight (g) was exhibited by the genotypes Butana , Tetron and Wad Ahmed (Table 3). These findings agreed with those reported by  Geremew (1993), who recorded a wide range of variability in 1000 seed weight. From these findings genotypes Mugod and Wad Fahal had large seed size compared to the genotypes Butana,Wad Ahmed and Tetron with  medium seed size. Tall, late maturing genotypes with large  to medium seed size such as Tetron are not suitable to grow under North Gedarif conditions because they need higher rainfall.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3.  Means of head width(cm) and 1000 seed weight (g) for 18 Sorghum genotypes grown at North Gedarif (NG),South Gedarif (SG),and Rahad (RH), seasons 2009,2010 , 2011.

Genotypes

Head width (cm)

            Seed weight ( g)

2009

2010

2011

Mean

2009

2010

2011

Mean

Korakollo

3

5

4

    4

25

30

25

26.6

Mugod

4

5

4

4

36

39

26

33.6

Safra

3

6

3

4

25

31

27

27.6

Wad Baku

4

5

3

4

27

31

27

28.3

Tetron

3

5

4

4

27

25

22

24.6

Faki Mustahi

3

4

4

3

25

39

27

30.3

Farhoda

3

5

4

4

25

35

24

28.0

Gadambaliabloom

4

5

3

4

27

33

25

28.3

Ajeb seido

3

5

3

3

20

27

22

23.0

Arafa

4

5

3

4

28

28

24

26.6

AG-8

3

5

4

4

26

30

25

27.0

Butana

3

6

3

4

18

23

22

2.0

Bashayier

3

5

3

3

23

27

25

25.0

Tabat

3

5

3

3

20

27

25

24.0

Wad ahmed

3

5

4

4

22

25

21

22.6

Gesheish

3

5

3

3

26

30

26

27.3

Wad Fahal

4

6

4

4

29

42

31

34.0

Milo

4

5

3

4

24

31

25

26.6

Mean

3.34

5

3

3.78

25

31

25

27.0

CV %

21.9

2.6

4.3

9.6

13.5

16.3

4.2

18.0

SE±

0.2

.3

0.2

0.2

0.9

1.4

0.8

1.1

 

**

**

 

 

**

**

**

 

*,** significant at  0.01 probability level .

 

Grain yield ( kg/ha)

    Across locations, this trait showed highly significant differences among the genotypes. The highest general mean (1973 kg/ha) was observed in season 2010, while the lowest general mean (450 kg/ha) was obtained in season 2009 (Table 4).The range for this trait was from 225 kg (Farhoda) to 735 kg (Butana) in season 2009,  from 1408 kg (Faki Mustahi) to 2572 kg (Wad Ahmed) in season 2010, from 862 kg (Faki Mustahi) to 2545 kg (Mugod)  in season 2011, respectively. Hence, in this study the highest grain

 

 

 

 

yields (kg/ha) were exhibited by the genotypes Butana, Wad Ahmed, and Mugod. Similar results were reported by Elasha et al .(2011);Elzein et al, 2008, and Abdalla et al (2009). While the lowest grain yields were obtained by the genotypes Farhoda and Faki Mustahi (Table 4).This study indicated that Butana was an early maturing and short genotype which is suitable for North Gedarif conditions ,while medium or tall genotypes that are late maturing,  coupled with high grain yield such as Mugod and Wad Ahmed  were suitable for growing under irrigation and high rainfall.

   Across seasons, grain yield showed highly significant differences    (Table 4). The highest general mean (1563 kg/ha) for this trait was obtained at Rahad location, while the lowest general mean (803 kg/ha ) was observed at North Gedarif. The range for this trait was from 409 kg (Wad Fahal) to 1129 kg (Arafa), from 916 kg (Faki mustahi) to 2572 kg (Mugod) , and from 1060 kg (Wad Baku) to 2120 kg (Wad Ahmed) for North Gedarif , South Gedarif and Rahad , respectively. Hence, in this study the highest grain yields were exhibited by the genotypes Mugod at South Gedarif and Wad Ahmed in Rahad location ,while the lowest grain yield were obtained by the genotypes Wad Fahal , Faki Mustahi and Wad Baku at North Gedarif, South Gedarif and Rahad, respectively. This is because all of them are tall and late maturing genotypes and they produce small seeds under drought spell conditions. From this finding, genotype Tetron was late maturing, tall ,with medium seed size and high grain yield, genotype Gesheish was early maturing and has low yield, while genotype Mugod is medium maturing, having medium height with big seed size and high yielding , while Faki Mustahi was a  late maturing genotype with a small seed size and has low yield.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Yield  (  kg/ha)

Seasons

Locations

Genotype

2009

2010

2011

NG

SG

RH

Korakollo

412

1674

1729

721

2004

1090

Mugod

447

1998

2545

679

2572

1738

Safra

363

2087

1753

941

2163

1100

Wad Baku

361

1804

1509

838

1777

1060

Tetron

346

1894

1731

799

1789

1382

Faki Mustahi

355

1408

862

532

916

1178

Farhoda

225

2134

1397

899

1375

1483

Gdambaliabloom

576

2044

1666

789

1799

1698

Ajeb seido

620

2188

977

923

1191

1670

Arafa

434

2059

1400

1129

1080

1684

AG-8

616

1834

1190

885

1156

1510

Butana

735

1963

1302

687

1615

1698

Bashayier

677

2121

1709

1001

1559

1948

Tabat

354

2215

1329

704

1310

1884

Wad Ahmed

391

2572

1450

698

1598

2120

Gesheish

487

1714

1380

726

1659

1196

Wad Fahal

266

1674

1750

409

1207

2073

Milo

450

2121

1287

1101

1231

1526

Mean

450

1973

1498

803

1556

1563

CV%

45.0

20.8

33.6

28.3

27.6

30.6

SE±

24.6

49.7

61.0

27.4

52.01

1.1

 

**

**

**

**

**

**

Table 4.  Means of  grain yield (kg/ha) combined over locations and over seasons for 18 Sorghum genotypes grown at North Gedarif (NG), South Gedarif (SG), and Rahad (RH), seasons 2009, 2010, 2011.

    **significant at 0.01 probability level.

 

Simple linear correlation coefficients

    Grain yield was positively , and highly significantly correlated with head width (0.65**), number of heads/m² (0.46**) , and 1000-seed weight (0.32*). It was positively and non-significantly correlated with days to 50% flowering (Table 5). Similar results were reported by Bittinger et al. (1981) and  Elagib (1999) who found  that grain yield was positively correlated with days to 50% flowering and 1000-grain weight. Also, positive association between grain yield and days to 50% flowering was reported by many authors; Liang et al.  (1969) in sorghum and  Umakanth et al. (2001) found

 

 

that correlation coefficients were moderate to high for days to anthesis, plant height,1000-seed weight, number of plant/m² and number of heads/m². Liang et al. (1969) found that 1000-grain weight was significantly correlated with grain yield. Shukla (1966) Chigwe (1984) found that 1000-seed weight was significantly correlated with grain yield under dry conditions in all maturity groups. Hadjichrislodoulu (1990) and Krishnasamy (1986) found that days to 50%

flowering showed a significant positive correlation with plant height in some hybrids. Also, a positive correlation was reported by Rana et al. (1984) who found an association between fodder yield and days to 50% flowering.Grain yield was significantly and negatively correlated with number of plants at establishment/m² but negatively and non-significantly correlated with head length and plant height. Plant height was positively and highly significantly correlated with one thousand seed weight (0.48**), head width (43**), number of heads /m² (0.40**), and number of plants/m² (0.35**).These results indicated that selection for these traits may be effective in improvement of grain yield, in addition, these findings indicated that tall plants possess heavier heads than short ones.                                                  

     One thousand seed weight was positively and highly significantly correlated with head width (0.59**) and plant height (0.48**) (Table 5), but positively and not significantly correlated with number of heads/m² and days to 50% flowering.1000-seed weight was negatively correlated with number of plants/m² and head length, Significant and positive correlations of 1000-seed weight with plant height were reported by Ezeaku and Mohamed (2006). Hence,  in this study,  head width, number of heads/m² and 1000-grain weight had strong correlation with grain yield (0.65**), (0.46**), and (0.32*), respectively, while it was negatively correlated with plant height, number of plants/m², and head length. The positive and significant association of grain yield with head width and number of heads/m² was mainly due to their positive  direct effect with negligible indirect effects through other characters. This suggested the direct use of these two characters as selection criteria.

 

 

 

 

 

 

Table 5. Simple linear correlation coefficients among various pairs of 8 characters of sorghum genotypes combined over three seasons (2009,2010,2011) and three locations (North Gedarif, South Gedarif and Rahad).

 

50%F

PH

#P/E

#H/m2

H L

H W

1000SW

GY

50%F

-

 

 

 

 

 

 

 

PH

0.30

-

 

 

 

 

 

 

#P/E

0.03

.348**

-

 

 

 

 

 

#H/m2

-0.01

0.40**

0.18

-

 

 

 

 

H L

0.34*

0.28*

0.16

0.18

 

 

 

 

H W

0.13

0.43**

-0.27

0.62**

0.23

-

 

 

1000SW

0.22

0.48**

-0.13

0.26

-0.03

0.59**

-

 

GY

0.13

-0.04

-0.59*

0.46**

-0.02

0.65**

0.32*     

-

*,** Significant at 0.05 and 0.01 probability levels, respectively.

50%F: Days to 50% flowering, PH: Plant height, #P/E=number of plants/establishment, #H/m²: Numbers of heads/m²,HL: Head length, HW: Head width,1000.SW,GY: Grain yield (kg/ha).

 

Path coefficient analysis

    The relatively large, positive and significant simple linear correlation coefficient between grain yield and number of heads/m² was (0.46**) .The positive direct effect of number of heads/m² on grain yield was the  highest (0.47) (Table 6). The highest positive direct effect on grain yield was exhibited by head width (0.33). Its indirect effect through number of heads/m² is large while too small through the other characteristics. This suggested the use of this character as a selection criterion for the improvement of grain yield. Its indirect effects on grain yield were negligible through the other traits.

 

 

 

 

 

 

 

 

 

 

 

 

    The relatively small, negative simple linear correlation between grain yield and head length (-0.02) is explained via the negative direct effect of head length on grain yield (-0.12) ,so it is difficult to recommend this character as a  selection criterion for yield (Table 6).

    Head width was highly significantly and positively correlated with grain yield (o.65**) (Table 6),such strong association is explained via the high positively direct effect of head width on grain yield(0.33). Low negative indirect effect were observed via plant height, head length, and 1000-seed weight. Also,  it had low positive indirect effect on grain yield via  number of plants/m², number of heads/m² and days to 50% flowering (Table 6).

     In this study, correlation and path analysis may measure two different aspects. Hence, the study of correlation alone does not give accurate indications of yield association. For example, in this study correlation between days to 50% flowering and grain yield was very small (0.13). This means that this character had no in influence on  grain yield, but the path analysis expressed days to 50% flowering as an important trait influencing yield.   

    From the present study, the direct effect of the tested traits on grain yield  indicated that among yield components head width and number of heads/m² had the highest correlation coefficient with grain yield. These traits also showed a positive direct effect on grain yield and therefore, these characters  may be considered as selection criteria for  developing high yielding sorghum genotypes.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 6. Path coefficient analysis of direct(in bold) and indirect effect

 of 8 characters on sorghum grain yield of 18 genotypes grown seasons 2009,2010,2011 at North Gedarif, South Gedarif and Rahad.

 

X1

X2

X3

X4

X5

X6

X7

Rij

X1

0.23

-0.07

-0.02

0.01

-0.04

0.04

-0.00

0.14

X2

0.07

-0.24

-0.17

0.19

-0.03

0.14

-0.00

-0.04

X3

0.01

-0.08

-0.50

0.09

-0.02

-0.09

0.00

-0.59**

X4

-0.00

-0.09

-0.09

0.47

-0.02

0.20

-0.00

0.46**

X5

0.08

-0.07

-0.08

0.08

-0.12

0.07

0.00

-0.02

X6

0.01

-0.10

0.13

0.29

-0.03

0.33

-0.00

0.65**

X7

0.05

-0.11

0.07

0.12

0.00

0.19

-0.01

0.32*

 

 

 

 

 

 

 

 

 

 

 

 

 

 

*,** Significant at P=0.5 and 0.01 level of probability, respectively. rij= Simple linear correlation coefficient.

X1: Days to 50% flowering,X2:Plant height,X3:number of plants at establishment/m2,X4:number of heads/m2,X5: Head length,X6: Head width,X7:1000 SW(g).

 

CONCLUSION

 

     Based on the results of this study, it could be concluded that a wide range of genetic variability was observed among sorghum genotypes for most of the  characters studied. The early maturing genotypes were Milo, Gesheish, AG-8 and Butana, an indication that these genotypes would fit quite well in short rainy seasons, which were suitable for North Gedarif environment, while the late maturing ones were Tabat,Wad Ahmed, Faki Mustahi and Tetron which were suitable to grow under Rahad and South Gedarif environments. The highest grain yields ( kg/ha) were exhibited by the genotypes Butana (735 kg/ha),Wad Ahmed (2572 kg/ha),and Mugod (2545kg/ha). Simple linear correlation and path coefficient analysis indicated that head width and numbers of heads/m² could be used as potential selection criteria in breeding programs for developing high yielding sorghum genotypes.

 

 

 

 

REFERENCES

 

Abdalla, H. M, Y. A. Gamar, A. H. Abu-Assar, T. Y. Elagib. M. H. Elgada, and O.M. Elhassan. 2009. A proposal for the release of two early maturing ,high yielding and drought tolerant sorghum genotypes, ARC, Wad Medani, Sudan.

 Bittinger ,T.S., R.P. Contrell, J. Axtell Dan and W.E. Nuqist .1981. Analysis of quantitative traits in NP 9 random-mating sorghum population. Crop Science 21:664-669.

Bushara, M.A.1999. Line x tester analysis for heterosis and combining ability in some genotypes of grain sorghum. M.Sc. Thesis, University of  Khartoum, Sudan.

Chigwe,C.F.B.1984. Quantitative  and  morphological  characteristics of NP9 BR random – mating  population of sorghum after nine cycles of  selection .(Abstract) Dissertation Abstract International, B. (science and  Engineering) 45(2) 419 B , Arizona University,  Tucson, USA.

Elagib, T. Y. 1999. Combining ability and heterosis for forage and grain yield in line x tester crosses of sorghum. M. Sc. Thesis, University of Gezira, Wad Medani, Sudan.

 Elasha, A., I. N. Elzein, A. H. A. Assar, M. K. Hassan, A. E. Hassan, O. M. Alhassan, A. A. Elmustafa and H. A. Hassan. 2011. A proposal for sorghum (Sorghum  bicolor (L.) Moench) hybrids release for irrigated and rain- fed sectors of the Sudan. ARC, Wad Medani, Sudan.

Elzein, I. N., E. I. Hassan, A. M. Ali, A. B. Elahmadi, E. A. Elasha and T. Eltaib. 2008. A proposal for the release of short maturing sorghum genotypes for drought prone areas of the Sudan. National Variety Release Committee, Khartoum, Sudan.

Ezeaku, I.E. and S.G. Mohamed. 2006. Character association and path analysis in grain sorghum. African Journal of Biotechnology 5(14):1337-1340.

Geremew.1993. Characterization and evaluation of sorghum germplasm collected from Gambella. Sorghum Newesletter 29:97.

Hadjichristodoulou, A.1990. Evaluational correlations among grain yield and other important traits of wheat in dry lands. Euphytica 44:143-150.

Osman, E.I and M.A. Mahmoud. 1992.  Improved sorghum genotypes suitable for irrigated and rain-fed land of  Sudan; Proceedings of Sudan  National Variety Release Committee. Wad Medani, Sudan.

 

Krishnasamy,V.1986.  Association of growth parameters with days to half-bloom in the parental lines of few sorghum hybrids. Madras Agricultural Journal 73(11) :653-654.

Liang, G.H., C.B. Overley, and A.J. Casady.1969. Interrelations among agronomic characters in grain sorghum. Crop Science 9:299-302.

Rana, B.S., B.C. Barah, H.P. Binswanger, and N.G.P. Rao. 1984. Breeding optimum plant types in sorghum. Indian Journal of Genetics and Plant Breeding 11(6):385-398.

 Sharaan, A.N. and K.H. Ghallab.1997. Character association at different locations in sesame. Sesame and Safflower Newsletter 12:66-75.

Shukla,P.T.1966. Studies of hybrid vigor in some of the Jowar Nagpur. Agricultural College Magazine , PP.121.

Sprague, G.  F. 1966. Evaluation of genetic variation in two open pollinated varieties of maize and their reciprocal and F1 Hybrids. Crop Science 4: 332-334.

Swarp, V. and D. S. Chaugale. 1962. Studies on genetic variability in hybrid sorghum seed production. Agronomy Journal 46: 20-23.

Umakanth. A.V. Madhusudhana, SwarnlataKaul, and B.S. Rana. 2001.Genetic  diversity studies in Sorghum National Research Centre for sorghum (NRCS), Rajendranagar, Hyderabad 500030, Andhra Pradesh University12:318-379.

 

 

 

 

 

 

 

 

 

published in Gezira Journal of agricultural science

  • Combining ability for grain yield and yield components in local inbred lines and introduced open pollinated varieties of maize (Zea mays L).

ABSTRACT

 

   The development of hybrids is the main objective of maize breeding. However, success depends largely on the identification of the best parents to ensure maximum combining ability. This study was conducted to estimate genetic variability and combining ability for grain yield and yield components of seven local inbred lines and four introduced open pollinated varieties of maize (Zea mays L.) across two irrigated locations, Medani and Matuq, Gezira, Sudan in 2008. The experiment was arranged in a randomized complete block design with three replicates. The traits measured were days to 50% tassel, plant height, ear length, ear diameter, hundred kernels weight and grain yield. Significant differences were observed among the parents and crosses for most of studied traits in both seasons. The crosses showed high genetic variability and tall plants than their parents which suggested some degree of hybrid vigor. The tallest hybrids across locations were T3 x L5 and T4 x L3. This indicates that the crosses were late maturing than their parents. The highest yielding hybrids had long ears and better shape, e.g., T2 x L1 and T1 x L7.The top five ranking crosses for grain yield across locations were T2 x L7 (3.45 t/ha), T1 x L2 (3.44 t/ha), T2 x LI (3.32 t/ha), T4 x L4 (3.30 t/ha) and T1 x L1 (3.13 t/ha).   The inheritance of most traits was controlled by non-additive gene action except ear height and grain yield. The best combiners for grain in Medani were T4, L4 and L5, while in Mutaq were L2, L4 and L6. The ratio of GCA to SCA variance for the most traits was less than one, suggesting that the inheritance was due to non additive gene effect with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and was largely controlled by additive gene action in the base material. From these results it is recommended that parents T4, L1 and L6 to be used in recurrent selection, while, crosses T3 x L5, T1 x L5 and T4 x L6 to be tested in multi-locations trials for commercial utilization.

 

INTRODUCTION

       Maize generally is one of the most diverse crop both genetically and phenotypically. Due to its wide adaptability and productivity, maize spread rapidly around the world after the Europeans brought the crop from the Americas in the 15th and 16th centuries (McCann, 2005). The Portuguese introduced the crop to Africa at the beginning of the 16th century and since then the crop has replaced sorghum and millet as the main staple food in most of the continent where the climatic conditions are favorable (McCann, 2005). Today, there  is an increasing interest in maize production in Sudan due to its suitability to cultivation in the agricultural irrigated schemes, especially in the Gezira.It can occupy an important position in the economy of the country due to the possibility of blending it with wheat for making bread (Nour et al., 1997; Meseka, 2000).

    The grain yield of existing maize varieties and local landraces in Sudan is low. Also, maize   hybrids have been reported to show high potential for grain yield than the open pollinated varieties and landraces (Alhussein, 2007). Advantages of hybrids over open pollinated cultivars are higher yield, uniformity, high quality and resistance to diseases and pests. In spite of having yield potential, the production of maize in Sudan is very low. One of the reasons for this is the cultivation of exotic hybrids, which are not well adapted to our agro-climatic conditions. One of the strategies of the Agricultural Research Corporation (ARC) of the Sudan for maize breeding program is to develop new hybrids as an attempt to incorporate both advantages for higher yield and adaptability to environmental conditions. Thus, getting the benefit from the use of hybrids is the main purpose in maize breeding program of ARC.    Therefore, the objective of this study is to estimate the magnitude of combining ability in 28 topcross hybrids of maize for grain yield and its components across two irrigated locations and to identify high yielding topcross hybrids for future testing and commercial utilization.

 

 

MATERIALS AND METHODS

    The plant material used consisted of 7 local inbred lines used as lines (L), and 4 introduced open pollinated varieties used as testers (T) crossed in line x tester arrangement (Table 1). Hand pollination was used to develop the breeding material. Pollen grain was collected into a paper bag from the tassel of male parent (tester) and then dusted on the silk of the female parent (line). The ear was covered with a bag and information regarding the cross was written on the bag. A total of 28 cross combinations was obtained through hand pollination. In July 2008, the 11 parental material and 28 cross hybrids were grown and evaluated at two irrigated locations, Medani, Gezira Research Station (GRS) and Matuq, Matuq Research Station (MRS), Gezira State, Sudan. The trials were arranged a randomized complete block design with three replicates. The plot size was maintained as 2 rows x 3 m long with inter and intra row spacings of 80 and 25 cm, respectively.  Seeds were sown at the rate of 3- 4 seeds per hill.  Plants were thinned to one plant per hill after three weeks from sowing. Nitrogen was applied at 86 kg/ha in a split dose after thinning and before flowering. The crop was irrigated at intervals of 10-14 days, and plots were kept free of weeds by hand weeding.  Data were analyzed using the Statistical Analysis System (SAS) computer package. The analysis was done for each season for characters days to 50% tasseling, plant height, ear length, ear diameter, kernels weight and grain yield and then combined. Mean performance was separated using Duncan's Multiple Range Test (DMRT). Data from each location was analyzed separately and across locations to determine the general and specific combining ability of each line was measured according to Griffing,s Method 2 (1956).

 

Table 1. Pedigree of the lines and testers used in the study.

Parents

Pedigree

Source 

L1

RING-B-S1-2    

Inbred line developed by ARC

L2

PR-89 B-5655-S1-1

Inbred line introduced from CIMMYT, Mexico

L3

RING-B- S1-3   

Inbred line developed by ARC

L4

RING- B-S1-1

Inbred line developed by ARC

L5

RING-A-S1-1

Inbred line developed by ARC

L6

RING-A-S1-2

Inbred line developed by ARC

L7

PR-89 B-5655-S1-3

Inbred line introduced from CIMMYT, Mexico

T1

SOBSIY-HG AB                        

OPV introduced from CIMMYT, Kenya

T2

ACROSS- 500 HGY-B             

OPV introduced from CIMMYT, Kenya

T3

CORRALE10 -02 SIYQ           

OPV introduced from CIMMYT,  Kenya

T4

BAILO- 02SIYQ                        

OPV introduced from CIMMYT,  Kenya

RESULTS AND DISCUSSION

 

   The performance of the material tested for most traits is high across the two locations. However, significant differences among the parents and their hybrids for most traits were shown indicating the diversity of the material.

Mean separation and ranking

    Mean days to 50% tasseling indicates that the pollen shedding ability of maize genotypes is an indicator of the earliness of genotypes. Mean days to tasseling across locations for parents scored 52 days as the general mean. Mean of parents ranged between 49 and 55 days for L6 and T3, respectively (Table 2). The mean of crosses ranged between 46 days for (T4 x L5) to 52 days for (T2 x L1) (Table 3). Identification of early tasseling genotypes is very important in developing hybrids and choosing hybrids to suit different agro-ecological zones as well as grower’s requirements. Earliness was a desirable trait especially under rainfed conditions. It is important for better use of water resources and avoidance of late season infestation with stem borers. Hence, the earliest crosses were T1 x L7 (47 days), T4 x L7 (47 days), T4 x L4 (48 days) and T4 x L6 (48 days) (Table 3).

 

Table 2. Mean performance of eleven parents for the measured traits in maize at the two locations, season 2008.

Traits /

Parents

       DT   

      PH   

        EL    

       ED    

        KW  

      GY  

Mean   Rank

Mean  Rank

 Mean Rank 

Mean  Rank  

 

Mean Rank

 

Mean   Rank

L1

49.1      10

131.4     10

14.2         4

3.7          3

20.7         6

   2.8         2

L2

50.0        9

148.5       4

15.0         1

3.6          7

19.9       11

   2.6         5

L3

51.7        6

145.2       6

13.2         9

3.6          6

20.7         8

   2.4         8

L4

50.0        8

152.0       3

14.3         3

4.1          1

20.3       10

   2.1        11

L5

51.7        5

145.6       5

13.7         5

3.6          4

22.6         2

   2.7         3

L6

49.1      11

139.1       9

13.4         8

3.4        11

22.1         3

   2.2         9

L7

50.1        7

131.1     11

12.7       11

3.4        10

20.7         7

   2.4         7

T1

52.7        4

139.3       8

13.6         7

3.9          2

21.3         5

   2.2       10

T2

54.2        2

155.9       2

14.8         2

3.6          5

21.7         4

   2.4         6

T3

55.2        1

157.7       1

13.7         6

3.5          8

22.8         1

   2.6         4

T4

52.8        3

143.2       7

12.9       10

3.5          9

20.5         9

   2.9         1

Mean

52.3

144.4

13.5

3.5

21.4

   2.4

CV%

  6.7

  10.0

13.0

9.8

14.5

 27.8

S.E±

  0.98

    2.33

0.38

0.08

  0.81

   0.15

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha).

 

    Tallness is not a good character in grain maize production, since tall maize plants tend to be susceptible to stem and root lodging.  Highly significant differences for tallness were detected among the studied parents with the general mean being of 144.4 cm. The trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. In the studied parents mean plant height ranged between 131.1 cm for L7 to 158 cm for T3 which was the tallest and latest parent across locations (Table 2). The crosses mean varied from 135.1 cm for (T3 x L7) to 155.9 cm for (T2 x L1).The tallest hybrids across locations were T4 x L6 and T4 x L3 (154 cm) (Table 3).

 

Table 3.  Performance of 28 crosses for the measured traits in maize at the two locations combined,  season 2008.

Traits/

Crosses

         DT                    PH                      EL                    ED                    KW                   GY

 

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

 

T1 x L1

48.5

   22

  14.6

13

14.2

    6

3.8       

 4

22.0    

  9

3.1       

  5

 

T1 x L2

48.5   

20

148.3   

14

14.2     

 7

3.5      

22

23.2    

  1

3.4         

  2

 

T1 x L3

50.0   

13

149.8   

 7

13.7    

18

3.7     

  9

21.7    

14

2.9       

12

 

T1 x L4

50.1   

12

145.0   

18

13.3    

22

3.7      

14

22.1    

  7

2.9       

11

 

T1 x L5

49.0    

19

145.6   

16

12.9    

25

3.5      

23

22.2    

  6

3.0       

10

 

T1 x L6

50.1   

11

152.3   

4

14.3    

  5

3.7      

11

21.8    

18

2.7       

21

 

T1 x L7

46.8   

27

138.9   

25

15.2      

  2

3.4      

26

20.3    

24

2.9       

16

 

T2 x L1

52.3   

  1

155.9   

 1

14.1    

  8

4.0      

  1

20.8    

22

3.3       

  3

 

T2 x L2

49.5   

17

149.2   

10

13.2    

21

3.7    

15

19.9    

27

2.4       

26

 

T2 x L3

51.2   

  4

145.2   

17

12.2    

27

3.7    

16

22.8    

  3

3.1       

  7

 

T2 x L4

50.2   

  9

141.0   

22

13.2    

24

3.7    

17

22.1    

  8

2.4       

15

 

T2 x L5

49.5     

18

140.8   

24

14.0    

10

3.7    

13

21.3    

17

2.0       

28

 

T2 x L6

50.0     

14

143.4   

19

14.6    

  4

3.3     

27

20.1    

25

3.1       

  8

 

T2 x L7

48.2     

21

149.1   

11

13.9    

14

3.4    

25

19.7    

28

3.5       

  1

 

T3 x L1

50.3     

  7

150.3   

 6

13.9    

12

3.6    

20

21.6    

16

2.8       

18

 

T3 x L2

49.7     

16

149.8   

 8

13.7    

16

3.7    

  7

21.7    

13

2.9       

13

 

T3 x L3

48.0     

23

139.2   

24

13.3    

20

3.8    

  2

22.4    

  5

2.7       

22

 

T3 x L4

50.2     

10

142.9   

21

11.9    

28

3.7    

12

20.6    

23

3.0       

  9

 

T3 x L5

51.2     

  3

151.4   

  5

16.1    

  1

3.6    

21

22.5    

  4

2.9       

17

 

T3 x L6

50.8     

  5

138.8   

26

13.9    

13

3.3    

28

20.9    

21

2.6       

24

 

T3 x L7

52.2     

  2

135.1   

28

14.1    

  9

3.5    

24

21.0    

20

2.2       

27

 

T4 x L1

50.3     

  8

146.1   

15

12.8    

26

3.7    

18

21.7    

12

2.5       

25

 

T4 x L2

50.0     

15

149.8   

 9

13.7    

17

3.7    

  8

21.7    

15

2.9       

14

 

T4 x L3

50.3     

  6

154.2   

 3

14.0    

11

3.8    

  5

23.2    

  2

3.1       

  6

 

T4 x L4

47.5     

25

148.9   

12

13.6    

19

3.7    

  6

21.8    

10

3.3       

  4

 

T4 x L5

45.7     

28

135.1   

27

13.8    

15

3.7    

10

21.8    

11

2.8       

19

 

T4 x L6

48.0     

24

154.2   

 2

13.2    

23

3.8    

  3

20.1    

26

2.7          

20

 

T4 x L7

47.2     

26

143.1   

20

15.2    

  3

3.6    

19

21.2    

19

2.6       

23

 

Mean

49

 

145.9

 

13.8

 

3.7

 

21.3

 

2.8

 

 

CV%

  6.7

 

10

 

13

 

9.8

 

14.5

 

 27.8     

 

 

S.E±

  0.64

 

    3.8

 

  0.46

 

0.08

 

  0.56

 

0.14

 

 
















 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm),  KW= kernels weight and GY= grain yield (t/ha).

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 

    The results indicate that crosses were later than their parents. Also, the taller crosses were late maturing than short ones. Generally, the crosses were taller than their parents which suggested some degree of hybrid vigor.

   Ear length trait is an important selection index for grain yield in maize. The ear length means of parents, as expected, were found to be shorter than those of the crosses at the two sites, with the general mean of 13.5 cm. The parents mean ranged between 12.7 cm for L7 to 15 cm for L2 (Table 2). The crosses mean varied from 11.9 cm for (T3 x L4) to 16.1 cm for (T3 x L5). However, long ear length were recorded for crosses T1 x L7 (15.2 cm), and T2 x L6 (14.6cm) (Table 3).Vedia and Claure (1995) found that ear length was the most important yield component and when used as a selection index genetic gain in recurrent selection reached 9.94% for yield and 5.75% for the ear traits. Therefore, any increase in ear length would be expected to increase number of kernels/row and hence increase grain yield.

    Ear diameter is a good indicator of the number of kernel rows/ear. The mean of ear diameter across sites for parents ranged between 3.4 cm for L6 and L7 to 4.1 cm for L4 (Table 2). Among the crosses, the large ear diameter ranged from 3.3 cm for T3 x L6 to 4.0 cm for T2 x L1. The crosses which had a big ear diameter were T3 x L3 and T4 x L6 (3.8cm) (Table 3). This result was in agreement with the findings of Tracy (1990) who found that the maize hybrids with high yield had more ears/plant, longer ears and a better ear shape and row configuration.

The mean of one hundred kernels weight for parents was 21.4 g, and it ranged between 19.9 g for L2 to 22.8 g for T3 (Table 2). Among the crosses, the mean was 21.3 g. The best crosses which obtained the highest kernel weight were T1 x L2 and T4 x L3 (23.2) followed by T2 x L3 (22.8 g) (Table 3).

Yield is a polygenic character is influenced by the fluctuating enviro-nment. Moreover, it is a complex trait depending on many components (Sharaan and Ghallab, 1997). In this study, there was a considerable amount of variability among the genotypes for this trait. The studied parents in the two locations showed a general mean of 2.4 t/ha. The parents means ranged between 2.12 t/ha for L4 to 2.93 t/ha for T4 (Table 2), while, the crosses means ranged between 2.0 t/ha for (T2 x L5) to 3.55 t/ha for (T2 x L7) (Table 3).  Most of the crosses (19 hybrids) had significantly higher mean grain yield than the overall mean. It is of interest to mention that the top ranking and the best yielder hybrids were T1 x L2 (3.4 t/ha), T2 x L1 (3.3 t/ha), T4 x L4 (3.3 t/ha), T1 x L1 (3.30 t/ha) and T4 x L3 (3.1 t/ha). These results agreed with those of Khalafalla and Abdalla (1997), who pointed to the fact that hybrids (crosses) produce higher grain yields than the open pollinated varieties due to the good performance of hybrids under Sudan conditions.

 Combining ability

    The breeding method to be adopted for improvement of a crop depends primarily on the nature of gene action involved in the expression of quantitative traits of economic importance. Combining ability leads to identification of parents with general combining ability effects and in locating cross combining showing high specific combining ability effects. In this study the ratio of GCA to SCA mean variance for most traits was less than one, suggesting that the inheritance of these traits was due to non additive gene action, with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and largely

controlled by additive gene action in the base material (Table 4).

Table 4. Mean squares of six agronomic traits for maize parents and 28 lines x tester crosses tested at two locations, Medani and Mutaq 2008.

Source of variation

DF

DT

PH

EL

ED

KW

GY

Location

  1

 3322.70**

13721**   

4.48**

50.90**

287.8**

 26.9**

Line

  6

     04.22

119.19

4.03

  0.02

    5.69

   0.27

Tester

  3

     18.81

   46.63

0.85

  0.07

    3.06

   0.28

Line x tester

18

     05.85**

   93.70*

2.38*

  0.05*

    1.62*

   0.44*

Line x tester x

location

18

     11.91**

217.90**

2.80

  0.11

     6.68

   0.63

Pooled error

76

     05.24

108.60

1.76

  0.04

     3.08

   0.19

GCA

 

       0.2

    -5.0

0.2

  0.00

    -0.7

   0.08

SCA

 

       0.6

    13.7

0.5

  0.02

     0.7

   0.03

GCA/SCA

 

       0.4

    -0.4

0.4

 -0.15

   -1.0

3.07

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 









 

   This result indicates that dominance and epistatic interaction effects seemed to be predomint for this trait and therefore heterosis breeding may be gratifying. The good combiner parents, those having negative GCA effects in Medani, for 50% days to tasseling were L5 followed by T4, T1 and L7, indicating earliness for flowering time, while, the latest, having positive GCA effect was T3 (Table 5).The earliest crosses having negative SCA effects were T3 x L6, T2 x L7 and T2 x L4, while, the latest crosses were T2 x L5, T4 x L5 and T4 x L4 (Table.6).

    The earliest parent in Mutaq was L7 (Table 5) and the earliest crosses were T2 x L4, T4 x L4 and T3 x L4 (Table 6). Common parents across locations that contributed to earliness were T4 and L5. The latest were L6 followed by T3 and T2 (Table 5). Parent L4 had good contribution for earliness to their hybrids progeny across locations.

Thus, the inbred lines which exhibited good general combining ability for at least one character can be used for development of early maturity and high grain yield. The contribution of the total variance for general and specific combining ability for this trait differs from location to another, but SCA was high in both locations (50.4% and 71.7%) compared with GCA which indicates that this trait is  controlled by additive gene action (Figs 1 and 2).

Trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. Only three top cross hybrid parents in Medani have   negative GCA effects for plant height, i.e., L7, L3 and T3; they were best combiners for short plant type. Tallness which is an undesirable trait is shown by parents L1, L2 and T1 (Table 5). Crosses having negative SCA effects and consequently short plant type were T4 x L2, T1 x L4 and T2 x L4,  while, tall hybrids with positive SCA effects were T3 x L1, T2 x L5 and T4 x L5 (Table 6).




 

   The best combiners for the short plant type with negative GCA in Mutaq were L7, L6 and L2 while, the taller parents with high positive GCA effects were L5 and L1 (Table 5). Among the crosses the shortest hybrids were T2 x L4, followed by T3 x L5 and the tallest hybrids were T2 x L5 and T3 x L4 (Table 6). This showed that, there is a relationship between late flowering and tall plant type. This is quite obvious among the hybrids such as T3 x L1 and T3 x L4.  Contribution for this trait is higher in crosses (80% and 53%) compared to parents (20% and 40%) at the two locations (Figs 1 and 2). The earliness and shortness are desirable traits especially under rainfed conditions for better water use efficiency and the escape of drought and avoidance of late season infestation with stem borer.

    Ear length is a good index for higher grain yield, therefore any increase in ear length would be expected to increase number of kernels/row and hence directly improve grain yield. In Medani site, the long ear length parents having a positively significant GCA effects such as L5, L7 and T1, while parents showing the short ear length were L4 and L2 (Table 5). The best crosses for this trait having a positive SCA effects and hence the longest ear length were T2 x L5 and T4 x L7. On the other hand the best combiners in Mutaq were L7 and T3 (Table 5), while the best crosses were T1 x L1 and T4 x L4 (Table 6). In the two locations, the best contribution was (73% and 65.9%) obtained by SCA compared with (27% and44.1) for GCA (Figs1 and 2). These results emphasized that ear length has a direct effect for improving grain yield. This is in agreement with the finding of Vedia and Claure (1995) who found that ear aspect was the most important yield component.

     Based on GCA estimates, the best combiners for ear diameter and length in Medani are L1 and L5, while best crosses were T1 x L2, T3 x L5 and T3 x L7. The good combiners in Mutaq site are L2, L3 and L4, while the best crosses are T3 x L4, T4 x L1 and T1 x L5 (Tables 5 and 6). A higher contribution among this trait is obtained by SCA (55.9% and 65.9%) in both locations compared with GCA (Figs 1and 2).

    Favorable GCA values were given by T1 and L3 as the good combiners for kernel weight in Medani and the best crosses were T4 x L7 and T2 x L7.  Among the studied parent material in Mutaq, only three parents have positive GCA effects (L3, L4 and T3).

 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

 Figure 1. Parent contribution of the maize GCA and SCA to the total variance

                of yield and its componets at Medani, season 2008.

 

 

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

Figure 2. Parent contribution of the maize GCA and SCA to the total variance of yield and its componets at Mutaq, season 2008.

 

The best crosses were shown by T1 x L4 and T3 x L4 (Tables 5 and 6). The higher average contribution was given by the SCA (50.8% and 61) compared with the GCA at two locations (Figs 1and 2). This indicted that the inheritance of this trait was controlled by non additive gene effects.

    At Medani site, all the results depicted in Table 5 showed that the parents differ considerably with respect to estimate of GCA effects for grain yield. The parents having positive GCA effects were T1 followed by L4 and L6. Parents having negative GCA effects were L2 and L6. The best crosses having positive SCA effects were T3 x L3 followed by T4 x L5 and T1 x L2. Negative SCA effects were shown by T3 x L4, T2 x L2 and T1 x L4 (Table 6). The higher combiner in Mutaq, were L2, L1 and L4. The best crosses were T3 x L5, T1 x L5 and T4 x L5, while negative SCA effects were shown by T1 x L3, T2 x L5 and T4 x L3 (Tables 5 and6). The great contribution was given by SCA (62.4% and 62%) compared with GCA at the two locations (Figs1 and 2).

     General combing ability variance for grain yield is greater than the mean square for specific combining ability indicating the importance of additive gene action in controlling grain yield. This finding is in agreement with that of Barakat and Osman (2008) who found GCA effects are larger than SCA effects for grain yield indicating that the additive genetic variance is a major source of variations responsible for inheritance of grain yield.

 

CONCLUSION

   The ratio of general combining ability variance for grain yield was greater than specific combining ability indicating the importance of additive gene action in controlling this trait hence the good combiner parent for grain yield across locations was L4 so it could be used in recurrent selection. Also enormous variability was detected in the studied population which makes cyclic selection more effective. The best cross was T4 x L5 indicating that dominance and epesitic interaction seemed to be predomint, hence, higher heterosis gratified and recommended cross T4 x L5 for future testing in multi-locations trials for commercial utilization in order to be released as a hybrid.

REFERENCES

Alhussein, M.B. 2007. Growth Performance and Grain Yield Stability of   some Open Pollinated arieties of Maize (Zea mays L.). M.Sc. Thesis, University of Gezira, Wad Medani Sudan.

Barakat, A.A and M.M. Osman. 2008. Gene action and combining ability estimates for some romising maize inbred lines by top cross system. Journal of Agricultural Sciences. Mansoura niversity Journal 33:280-709

Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel  rossing system. Australian Journal of Biological Sciences 9: 463-493.

 Khalafalla, M.M. and H.A. Abdalla. 1997. Performance of some maize genotypes (Zea mays L.) nd  their  F1  hybrid  for  yield  and  its components. University of Khartoum Journal of Agricultural Sciences 5(2): 56-68.

Meseka, S.K. 2000.DiallelAnalysis for Combining Ability of Grain Yield and Yield Components n Maize (Zea  mays L.). M.Sc. Thesis, Faculty of Agricultural Sciences, University of ezira, Wad Medani, Sudan.

McCann, J. 2005. Maize and Grace: Africa’s Encounter with a New Crop, 1500-2000. Harvard niversity Press, New York

Nour, A.M., I. N. Elzain and M.A. Dafalla. 1997. Crop Development and   Improvement. Annual Report f the Maize Research Program. Agricultural Research Corporation, Wad Medani,Sudan.

Sharaan, A.N. and K.H. Ghallab. 1997. Character association at different location in sesame. Sesame and Safflower Newsletter 12: 66-75.

Tracy, W.F. 1990.  Potential of field corn germplasm for improvement of sweet corn. Crop Science 30:1041-1045.

Vedia, M.L. and E.T. Claure. 1995. Selection index for yield in the maize population. Crop Science 7: 505-510.

 

 

ABSTRACT

 

   The development of hybrids is the main objective of maize breeding. However, success depends largely on the identification of the best parents to ensure maximum combining ability. This study was conducted to estimate genetic variability and combining ability for grain yield and yield components of seven local inbred lines and four introduced open pollinated varieties of maize (Zea mays L.) across two irrigated locations, Medani and Matuq, Gezira, Sudan in 2008. The experiment was arranged in a randomized complete block design with three replicates. The traits measured were days to 50% tassel, plant height, ear length, ear diameter, hundred kernels weight and grain yield. Significant differences were observed among the parents and crosses for most of studied traits in both seasons. The crosses showed high genetic variability and tall plants than their parents which suggested some degree of hybrid vigor. The tallest hybrids across locations were T3 x L5 and T4 x L3. This indicates that the crosses were late maturing than their parents. The highest yielding hybrids had long ears and better shape, e.g., T2 x L1 and T1 x L7.The top five ranking crosses for grain yield across locations were T2 x L7 (3.45 t/ha), T1 x L2 (3.44 t/ha), T2 x LI (3.32 t/ha), T4 x L4 (3.30 t/ha) and T1 x L1 (3.13 t/ha).   The inheritance of most traits was controlled by non-additive gene action except ear height and grain yield. The best combiners for grain in Medani were T4, L4 and L5, while in Mutaq were L2, L4 and L6. The ratio of GCA to SCA variance for the most traits was less than one, suggesting that the inheritance was due to non additive gene effect with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and was largely controlled by additive gene action in the base material. From these results it is recommended that parents T4, L1 and L6 to be used in recurrent selection, while, crosses T3 x L5, T1 x L5 and T4 x L6 to be tested in multi-locations trials for commercial utilization.

 

INTRODUCTION

       Maize generally is one of the most diverse crop both genetically and phenotypically. Due to its wide adaptability and productivity, maize spread rapidly around the world after the Europeans brought the crop from the Americas in the 15th and 16th centuries (McCann, 2005). The Portuguese introduced the crop to Africa at the beginning of the 16th century and since then the crop has replaced sorghum and millet as the main staple food in most of the continent where the climatic conditions are favorable (McCann, 2005). Today, there  is an increasing interest in maize production in Sudan due to its suitability to cultivation in the agricultural irrigated schemes, especially in the Gezira.It can occupy an important position in the economy of the country due to the possibility of blending it with wheat for making bread (Nour et al., 1997; Meseka, 2000).

    The grain yield of existing maize varieties and local landraces in Sudan is low. Also, maize   hybrids have been reported to show high potential for grain yield than the open pollinated varieties and landraces (Alhussein, 2007). Advantages of hybrids over open pollinated cultivars are higher yield, uniformity, high quality and resistance to diseases and pests. In spite of having yield potential, the production of maize in Sudan is very low. One of the reasons for this is the cultivation of exotic hybrids, which are not well adapted to our agro-climatic conditions. One of the strategies of the Agricultural Research Corporation (ARC) of the Sudan for maize breeding program is to develop new hybrids as an attempt to incorporate both advantages for higher yield and adaptability to environmental conditions. Thus, getting the benefit from the use of hybrids is the main purpose in maize breeding program of ARC.    Therefore, the objective of this study is to estimate the magnitude of combining ability in 28 topcross hybrids of maize for grain yield and its components across two irrigated locations and to identify high yielding topcross hybrids for future testing and commercial utilization.

 

 

MATERIALS AND METHODS

    The plant material used consisted of 7 local inbred lines used as lines (L), and 4 introduced open pollinated varieties used as testers (T) crossed in line x tester arrangement (Table 1). Hand pollination was used to develop the breeding material. Pollen grain was collected into a paper bag from the tassel of male parent (tester) and then dusted on the silk of the female parent (line). The ear was covered with a bag and information regarding the cross was written on the bag. A total of 28 cross combinations was obtained through hand pollination. In July 2008, the 11 parental material and 28 cross hybrids were grown and evaluated at two irrigated locations, Medani, Gezira Research Station (GRS) and Matuq, Matuq Research Station (MRS), Gezira State, Sudan. The trials were arranged a randomized complete block design with three replicates. The plot size was maintained as 2 rows x 3 m long with inter and intra row spacings of 80 and 25 cm, respectively.  Seeds were sown at the rate of 3- 4 seeds per hill.  Plants were thinned to one plant per hill after three weeks from sowing. Nitrogen was applied at 86 kg/ha in a split dose after thinning and before flowering. The crop was irrigated at intervals of 10-14 days, and plots were kept free of weeds by hand weeding.  Data were analyzed using the Statistical Analysis System (SAS) computer package. The analysis was done for each season for characters days to 50% tasseling, plant height, ear length, ear diameter, kernels weight and grain yield and then combined. Mean performance was separated using Duncan's Multiple Range Test (DMRT). Data from each location was analyzed separately and across locations to determine the general and specific combining ability of each line was measured according to Griffing,s Method 2 (1956).

 

Table 1. Pedigree of the lines and testers used in the study.

Parents

Pedigree

Source 

L1

RING-B-S1-2    

Inbred line developed by ARC

L2

PR-89 B-5655-S1-1

Inbred line introduced from CIMMYT, Mexico

L3

RING-B- S1-3   

Inbred line developed by ARC

L4

RING- B-S1-1

Inbred line developed by ARC

L5

RING-A-S1-1

Inbred line developed by ARC

L6

RING-A-S1-2

Inbred line developed by ARC

L7

PR-89 B-5655-S1-3

Inbred line introduced from CIMMYT, Mexico

T1

SOBSIY-HG AB                        

OPV introduced from CIMMYT, Kenya

T2

ACROSS- 500 HGY-B             

OPV introduced from CIMMYT, Kenya

T3

CORRALE10 -02 SIYQ           

OPV introduced from CIMMYT,  Kenya

T4

BAILO- 02SIYQ                        

OPV introduced from CIMMYT,  Kenya

RESULTS AND DISCUSSION

 

   The performance of the material tested for most traits is high across the two locations. However, significant differences among the parents and their hybrids for most traits were shown indicating the diversity of the material.

Mean separation and ranking

    Mean days to 50% tasseling indicates that the pollen shedding ability of maize genotypes is an indicator of the earliness of genotypes. Mean days to tasseling across locations for parents scored 52 days as the general mean. Mean of parents ranged between 49 and 55 days for L6 and T3, respectively (Table 2). The mean of crosses ranged between 46 days for (T4 x L5) to 52 days for (T2 x L1) (Table 3). Identification of early tasseling genotypes is very important in developing hybrids and choosing hybrids to suit different agro-ecological zones as well as grower’s requirements. Earliness was a desirable trait especially under rainfed conditions. It is important for better use of water resources and avoidance of late season infestation with stem borers. Hence, the earliest crosses were T1 x L7 (47 days), T4 x L7 (47 days), T4 x L4 (48 days) and T4 x L6 (48 days) (Table 3).

 

Table 2. Mean performance of eleven parents for the measured traits in maize at the two locations, season 2008.

Traits /

Parents

       DT   

      PH   

        EL    

       ED    

        KW  

      GY  

Mean   Rank

Mean  Rank

 Mean Rank 

Mean  Rank  

 

Mean Rank

 

Mean   Rank

L1

49.1      10

131.4     10

14.2         4

3.7          3

20.7         6

   2.8         2

L2

50.0        9

148.5       4

15.0         1

3.6          7

19.9       11

   2.6         5

L3

51.7        6

145.2       6

13.2         9

3.6          6

20.7         8

   2.4         8

L4

50.0        8

152.0       3

14.3         3

4.1          1

20.3       10

   2.1        11

L5

51.7        5

145.6       5

13.7         5

3.6          4

22.6         2

   2.7         3

L6

49.1      11

139.1       9

13.4         8

3.4        11

22.1         3

   2.2         9

L7

50.1        7

131.1     11

12.7       11

3.4        10

20.7         7

   2.4         7

T1

52.7        4

139.3       8

13.6         7

3.9          2

21.3         5

   2.2       10

T2

54.2        2

155.9       2

14.8         2

3.6          5

21.7         4

   2.4         6

T3

55.2        1

157.7       1

13.7         6

3.5          8

22.8         1

   2.6         4

T4

52.8        3

143.2       7

12.9       10

3.5          9

20.5         9

   2.9         1

Mean

52.3

144.4

13.5

3.5

21.4

   2.4

CV%

  6.7

  10.0

13.0

9.8

14.5

 27.8

S.E±

  0.98

    2.33

0.38

0.08

  0.81

   0.15

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha).

 

    Tallness is not a good character in grain maize production, since tall maize plants tend to be susceptible to stem and root lodging.  Highly significant differences for tallness were detected among the studied parents with the general mean being of 144.4 cm. The trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. In the studied parents mean plant height ranged between 131.1 cm for L7 to 158 cm for T3 which was the tallest and latest parent across locations (Table 2). The crosses mean varied from 135.1 cm for (T3 x L7) to 155.9 cm for (T2 x L1).The tallest hybrids across locations were T4 x L6 and T4 x L3 (154 cm) (Table 3).

 

Table 3.  Performance of 28 crosses for the measured traits in maize at the two locations combined,  season 2008.

Traits/

Crosses

         DT                    PH                      EL                    ED                    KW                   GY

 

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

 

T1 x L1

48.5

   22

  14.6

13

14.2

    6

3.8       

 4

22.0    

  9

3.1       

  5

 

T1 x L2

48.5   

20

148.3   

14

14.2     

 7

3.5      

22

23.2    

  1

3.4         

  2

 

T1 x L3

50.0   

13

149.8   

 7

13.7    

18

3.7     

  9

21.7    

14

2.9       

12

 

T1 x L4

50.1   

12

145.0   

18

13.3    

22

3.7      

14

22.1    

  7

2.9       

11

 

T1 x L5

49.0    

19

145.6   

16

12.9    

25

3.5      

23

22.2    

  6

3.0       

10

 

T1 x L6

50.1   

11

152.3   

4

14.3    

  5

3.7      

11

21.8    

18

2.7       

21

 

T1 x L7

46.8   

27

138.9   

25

15.2      

  2

3.4      

26

20.3    

24

2.9       

16

 

T2 x L1

52.3   

  1

155.9   

 1

14.1    

  8

4.0      

  1

20.8    

22

3.3       

  3

 

T2 x L2

49.5   

17

149.2   

10

13.2    

21

3.7    

15

19.9    

27

2.4       

26

 

T2 x L3

51.2   

  4

145.2   

17

12.2    

27

3.7    

16

22.8    

  3

3.1       

  7

 

T2 x L4

50.2   

  9

141.0   

22

13.2    

24

3.7    

17

22.1    

  8

2.4       

15

 

T2 x L5

49.5     

18

140.8   

24

14.0    

10

3.7    

13

21.3    

17

2.0       

28

 

T2 x L6

50.0     

14

143.4   

19

14.6    

  4

3.3     

27

20.1    

25

3.1       

  8

 

T2 x L7

48.2     

21

149.1   

11

13.9    

14

3.4    

25

19.7    

28

3.5       

  1

 

T3 x L1

50.3     

  7

150.3   

 6

13.9    

12

3.6    

20

21.6    

16

2.8       

18

 

T3 x L2

49.7     

16

149.8   

 8

13.7    

16

3.7    

  7

21.7    

13

2.9       

13

 

T3 x L3

48.0     

23

139.2   

24

13.3    

20

3.8    

  2

22.4    

  5

2.7       

22

 

T3 x L4

50.2     

10

142.9   

21

11.9    

28

3.7    

12

20.6    

23

3.0       

  9

 

T3 x L5

51.2     

  3

151.4   

  5

16.1    

  1

3.6    

21

22.5    

  4

2.9       

17

 

T3 x L6

50.8     

  5

138.8   

26

13.9    

13

3.3    

28

20.9    

21

2.6       

24

 

T3 x L7

52.2     

  2

135.1   

28

14.1    

  9

3.5    

24

21.0    

20

2.2       

27

 

T4 x L1

50.3     

  8

146.1   

15

12.8    

26

3.7    

18

21.7    

12

2.5       

25

 

T4 x L2

50.0     

15

149.8   

 9

13.7    

17

3.7    

  8

21.7    

15

2.9       

14

 

T4 x L3

50.3     

  6

154.2   

 3

14.0    

11

3.8    

  5

23.2    

  2

3.1       

  6

 

T4 x L4

47.5     

25

148.9   

12

13.6    

19

3.7    

  6

21.8    

10

3.3       

  4

 

T4 x L5

45.7     

28

135.1   

27

13.8    

15

3.7    

10

21.8    

11

2.8       

19

 

T4 x L6

48.0     

24

154.2   

 2

13.2    

23

3.8    

  3

20.1    

26

2.7          

20

 

T4 x L7

47.2     

26

143.1   

20

15.2    

  3

3.6    

19

21.2    

19

2.6       

23

 

Mean

49

 

145.9

 

13.8

 

3.7

 

21.3

 

2.8

 

 

CV%

  6.7

 

10

 

13

 

9.8

 

14.5

 

 27.8     

 

 

S.E±

  0.64

 

    3.8

 

  0.46

 

0.08

 

  0.56

 

0.14

 

 
















 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm),  KW= kernels weight and GY= grain yield (t/ha).

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 

    The results indicate that crosses were later than their parents. Also, the taller crosses were late maturing than short ones. Generally, the crosses were taller than their parents which suggested some degree of hybrid vigor.

   Ear length trait is an important selection index for grain yield in maize. The ear length means of parents, as expected, were found to be shorter than those of the crosses at the two sites, with the general mean of 13.5 cm. The parents mean ranged between 12.7 cm for L7 to 15 cm for L2 (Table 2). The crosses mean varied from 11.9 cm for (T3 x L4) to 16.1 cm for (T3 x L5). However, long ear length were recorded for crosses T1 x L7 (15.2 cm), and T2 x L6 (14.6cm) (Table 3).Vedia and Claure (1995) found that ear length was the most important yield component and when used as a selection index genetic gain in recurrent selection reached 9.94% for yield and 5.75% for the ear traits. Therefore, any increase in ear length would be expected to increase number of kernels/row and hence increase grain yield.

    Ear diameter is a good indicator of the number of kernel rows/ear. The mean of ear diameter across sites for parents ranged between 3.4 cm for L6 and L7 to 4.1 cm for L4 (Table 2). Among the crosses, the large ear diameter ranged from 3.3 cm for T3 x L6 to 4.0 cm for T2 x L1. The crosses which had a big ear diameter were T3 x L3 and T4 x L6 (3.8cm) (Table 3). This result was in agreement with the findings of Tracy (1990) who found that the maize hybrids with high yield had more ears/plant, longer ears and a better ear shape and row configuration.

The mean of one hundred kernels weight for parents was 21.4 g, and it ranged between 19.9 g for L2 to 22.8 g for T3 (Table 2). Among the crosses, the mean was 21.3 g. The best crosses which obtained the highest kernel weight were T1 x L2 and T4 x L3 (23.2) followed by T2 x L3 (22.8 g) (Table 3).

Yield is a polygenic character is influenced by the fluctuating enviro-nment. Moreover, it is a complex trait depending on many components (Sharaan and Ghallab, 1997). In this study, there was a considerable amount of variability among the genotypes for this trait. The studied parents in the two locations showed a general mean of 2.4 t/ha. The parents means ranged between 2.12 t/ha for L4 to 2.93 t/ha for T4 (Table 2), while, the crosses means ranged between 2.0 t/ha for (T2 x L5) to 3.55 t/ha for (T2 x L7) (Table 3).  Most of the crosses (19 hybrids) had significantly higher mean grain yield than the overall mean. It is of interest to mention that the top ranking and the best yielder hybrids were T1 x L2 (3.4 t/ha), T2 x L1 (3.3 t/ha), T4 x L4 (3.3 t/ha), T1 x L1 (3.30 t/ha) and T4 x L3 (3.1 t/ha). These results agreed with those of Khalafalla and Abdalla (1997), who pointed to the fact that hybrids (crosses) produce higher grain yields than the open pollinated varieties due to the good performance of hybrids under Sudan conditions.

 Combining ability

    The breeding method to be adopted for improvement of a crop depends primarily on the nature of gene action involved in the expression of quantitative traits of economic importance. Combining ability leads to identification of parents with general combining ability effects and in locating cross combining showing high specific combining ability effects. In this study the ratio of GCA to SCA mean variance for most traits was less than one, suggesting that the inheritance of these traits was due to non additive gene action, with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and largely

controlled by additive gene action in the base material (Table 4).

Table 4. Mean squares of six agronomic traits for maize parents and 28 lines x tester crosses tested at two locations, Medani and Mutaq 2008.

Source of variation

DF

DT

PH

EL

ED

KW

GY

Location

  1

 3322.70**

13721**   

4.48**

50.90**

287.8**

 26.9**

Line

  6

     04.22

119.19

4.03

  0.02

    5.69

   0.27

Tester

  3

     18.81

   46.63

0.85

  0.07

    3.06

   0.28

Line x tester

18

     05.85**

   93.70*

2.38*

  0.05*

    1.62*

   0.44*

Line x tester x

location

18

     11.91**

217.90**

2.80

  0.11

     6.68

   0.63

Pooled error

76

     05.24

108.60

1.76

  0.04

     3.08

   0.19

GCA

 

       0.2

    -5.0

0.2

  0.00

    -0.7

   0.08

SCA

 

       0.6

    13.7

0.5

  0.02

     0.7

   0.03

GCA/SCA

 

       0.4

    -0.4

0.4

 -0.15

   -1.0

3.07

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 









 

   This result indicates that dominance and epistatic interaction effects seemed to be predomint for this trait and therefore heterosis breeding may be gratifying. The good combiner parents, those having negative GCA effects in Medani, for 50% days to tasseling were L5 followed by T4, T1 and L7, indicating earliness for flowering time, while, the latest, having positive GCA effect was T3 (Table 5).The earliest crosses having negative SCA effects were T3 x L6, T2 x L7 and T2 x L4, while, the latest crosses were T2 x L5, T4 x L5 and T4 x L4 (Table.6).

    The earliest parent in Mutaq was L7 (Table 5) and the earliest crosses were T2 x L4, T4 x L4 and T3 x L4 (Table 6). Common parents across locations that contributed to earliness were T4 and L5. The latest were L6 followed by T3 and T2 (Table 5). Parent L4 had good contribution for earliness to their hybrids progeny across locations.

Thus, the inbred lines which exhibited good general combining ability for at least one character can be used for development of early maturity and high grain yield. The contribution of the total variance for general and specific combining ability for this trait differs from location to another, but SCA was high in both locations (50.4% and 71.7%) compared with GCA which indicates that this trait is  controlled by additive gene action (Figs 1 and 2).

Trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. Only three top cross hybrid parents in Medani have   negative GCA effects for plant height, i.e., L7, L3 and T3; they were best combiners for short plant type. Tallness which is an undesirable trait is shown by parents L1, L2 and T1 (Table 5). Crosses having negative SCA effects and consequently short plant type were T4 x L2, T1 x L4 and T2 x L4,  while, tall hybrids with positive SCA effects were T3 x L1, T2 x L5 and T4 x L5 (Table 6).




 

   The best combiners for the short plant type with negative GCA in Mutaq were L7, L6 and L2 while, the taller parents with high positive GCA effects were L5 and L1 (Table 5). Among the crosses the shortest hybrids were T2 x L4, followed by T3 x L5 and the tallest hybrids were T2 x L5 and T3 x L4 (Table 6). This showed that, there is a relationship between late flowering and tall plant type. This is quite obvious among the hybrids such as T3 x L1 and T3 x L4.  Contribution for this trait is higher in crosses (80% and 53%) compared to parents (20% and 40%) at the two locations (Figs 1 and 2). The earliness and shortness are desirable traits especially under rainfed conditions for better water use efficiency and the escape of drought and avoidance of late season infestation with stem borer.

    Ear length is a good index for higher grain yield, therefore any increase in ear length would be expected to increase number of kernels/row and hence directly improve grain yield. In Medani site, the long ear length parents having a positively significant GCA effects such as L5, L7 and T1, while parents showing the short ear length were L4 and L2 (Table 5). The best crosses for this trait having a positive SCA effects and hence the longest ear length were T2 x L5 and T4 x L7. On the other hand the best combiners in Mutaq were L7 and T3 (Table 5), while the best crosses were T1 x L1 and T4 x L4 (Table 6). In the two locations, the best contribution was (73% and 65.9%) obtained by SCA compared with (27% and44.1) for GCA (Figs1 and 2). These results emphasized that ear length has a direct effect for improving grain yield. This is in agreement with the finding of Vedia and Claure (1995) who found that ear aspect was the most important yield component.

     Based on GCA estimates, the best combiners for ear diameter and length in Medani are L1 and L5, while best crosses were T1 x L2, T3 x L5 and T3 x L7. The good combiners in Mutaq site are L2, L3 and L4, while the best crosses are T3 x L4, T4 x L1 and T1 x L5 (Tables 5 and 6). A higher contribution among this trait is obtained by SCA (55.9% and 65.9%) in both locations compared with GCA (Figs 1and 2).

    Favorable GCA values were given by T1 and L3 as the good combiners for kernel weight in Medani and the best crosses were T4 x L7 and T2 x L7.  Among the studied parent material in Mutaq, only three parents have positive GCA effects (L3, L4 and T3).

 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

 Figure 1. Parent contribution of the maize GCA and SCA to the total variance

                of yield and its componets at Medani, season 2008.

 

 

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

Figure 2. Parent contribution of the maize GCA and SCA to the total variance of yield and its componets at Mutaq, season 2008.

 

The best crosses were shown by T1 x L4 and T3 x L4 (Tables 5 and 6). The higher average contribution was given by the SCA (50.8% and 61) compared with the GCA at two locations (Figs 1and 2). This indicted that the inheritance of this trait was controlled by non additive gene effects.

    At Medani site, all the results depicted in Table 5 showed that the parents differ considerably with respect to estimate of GCA effects for grain yield. The parents having positive GCA effects were T1 followed by L4 and L6. Parents having negative GCA effects were L2 and L6. The best crosses having positive SCA effects were T3 x L3 followed by T4 x L5 and T1 x L2. Negative SCA effects were shown by T3 x L4, T2 x L2 and T1 x L4 (Table 6). The higher combiner in Mutaq, were L2, L1 and L4. The best crosses were T3 x L5, T1 x L5 and T4 x L5, while negative SCA effects were shown by T1 x L3, T2 x L5 and T4 x L3 (Tables 5 and6). The great contribution was given by SCA (62.4% and 62%) compared with GCA at the two locations (Figs1 and 2).

     General combing ability variance for grain yield is greater than the mean square for specific combining ability indicating the importance of additive gene action in controlling grain yield. This finding is in agreement with that of Barakat and Osman (2008) who found GCA effects are larger than SCA effects for grain yield indicating that the additive genetic variance is a major source of variations responsible for inheritance of grain yield.

 

CONCLUSION

   The ratio of general combining ability variance for grain yield was greater than specific combining ability indicating the importance of additive gene action in controlling this trait hence the good combiner parent for grain yield across locations was L4 so it could be used in recurrent selection. Also enormous variability was detected in the studied population which makes cyclic selection more effective. The best cross was T4 x L5 indicating that dominance and epesitic interaction seemed to be predomint, hence, higher heterosis gratified and recommended cross T4 x L5 for future testing in multi-locations trials for commercial utilization in order to be released as a hybrid.

REFERENCES

Alhussein, M.B. 2007. Growth Performance and Grain Yield Stability of   some Open Pollinated arieties of Maize (Zea mays L.). M.Sc. Thesis, University of Gezira, Wad Medani Sudan.

Barakat, A.A and M.M. Osman. 2008. Gene action and combining ability estimates for some romising maize inbred lines by top cross system. Journal of Agricultural Sciences. Mansoura niversity Journal 33:280-709

Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel  rossing system. Australian Journal of Biological Sciences 9: 463-493.

 Khalafalla, M.M. and H.A. Abdalla. 1997. Performance of some maize genotypes (Zea mays L.) nd  their  F1  hybrid  for  yield  and  its components. University of Khartoum Journal of Agricultural Sciences 5(2): 56-68.

Meseka, S.K. 2000.DiallelAnalysis for Combining Ability of Grain Yield and Yield Components n Maize (Zea  mays L.). M.Sc. Thesis, Faculty of Agricultural Sciences, University of ezira, Wad Medani, Sudan.

McCann, J. 2005. Maize and Grace: Africa’s Encounter with a New Crop, 1500-2000. Harvard niversity Press, New York

Nour, A.M., I. N. Elzain and M.A. Dafalla. 1997. Crop Development and   Improvement. Annual Report f the Maize Research Program. Agricultural Research Corporation, Wad Medani,Sudan.

Sharaan, A.N. and K.H. Ghallab. 1997. Character association at different location in sesame. Sesame and Safflower Newsletter 12: 66-75.

Tracy, W.F. 1990.  Potential of field corn germplasm for improvement of sweet corn. Crop Science 30:1041-1045.

Vedia, M.L. and E.T. Claure. 1995. Selection index for yield in the maize population. Crop Science 7: 505-510.

 

 

 

 

 

ABSTRACT

 

   The development of hybrids is the main objective of maize breeding. However, success depends largely on the identification of the best parents to ensure maximum combining ability. This study was conducted to estimate genetic variability and combining ability for grain yield and yield components of seven local inbred lines and four introduced open pollinated varieties of maize (Zea mays L.) across two irrigated locations, Medani and Matuq, Gezira, Sudan in 2008. The experiment was arranged in a randomized complete block design with three replicates. The traits measured were days to 50% tassel, plant height, ear length, ear diameter, hundred kernels weight and grain yield. Significant differences were observed among the parents and crosses for most of studied traits in both seasons. The crosses showed high genetic variability and tall plants than their parents which suggested some degree of hybrid vigor. The tallest hybrids across locations were T3 x L5 and T4 x L3. This indicates that the crosses were late maturing than their parents. The highest yielding hybrids had long ears and better shape, e.g., T2 x L1 and T1 x L7.The top five ranking crosses for grain yield across locations were T2 x L7 (3.45 t/ha), T1 x L2 (3.44 t/ha), T2 x LI (3.32 t/ha), T4 x L4 (3.30 t/ha) and T1 x L1 (3.13 t/ha).   The inheritance of most traits was controlled by non-additive gene action except ear height and grain yield. The best combiners for grain in Medani were T4, L4 and L5, while in Mutaq were L2, L4 and L6. The ratio of GCA to SCA variance for the most traits was less than one, suggesting that the inheritance was due to non additive gene effect with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and was largely controlled by additive gene action in the base material. From these results it is recommended that parents T4, L1 and L6 to be used in recurrent selection, while, crosses T3 x L5, T1 x L5 and T4 x L6 to be tested in multi-locations trials for commercial utilization.

 

INTRODUCTION

       Maize generally is one of the most diverse crop both genetically and phenotypically. Due to its wide adaptability and productivity, maize spread rapidly around the world after the Europeans brought the crop from the Americas in the 15th and 16th centuries (McCann, 2005). The Portuguese introduced the crop to Africa at the beginning of the 16th century and since then the crop has replaced sorghum and millet as the main staple food in most of the continent where the climatic conditions are favorable (McCann, 2005). Today, there  is an increasing interest in maize production in Sudan due to its suitability to cultivation in the agricultural irrigated schemes, especially in the Gezira.It can occupy an important position in the economy of the country due to the possibility of blending it with wheat for making bread (Nour et al., 1997; Meseka, 2000).

    The grain yield of existing maize varieties and local landraces in Sudan is low. Also, maize   hybrids have been reported to show high potential for grain yield than the open pollinated varieties and landraces (Alhussein, 2007). Advantages of hybrids over open pollinated cultivars are higher yield, uniformity, high quality and resistance to diseases and pests. In spite of having yield potential, the production of maize in Sudan is very low. One of the reasons for this is the cultivation of exotic hybrids, which are not well adapted to our agro-climatic conditions. One of the strategies of the Agricultural Research Corporation (ARC) of the Sudan for maize breeding program is to develop new hybrids as an attempt to incorporate both advantages for higher yield and adaptability to environmental conditions. Thus, getting the benefit from the use of hybrids is the main purpose in maize breeding program of ARC.    Therefore, the objective of this study is to estimate the magnitude of combining ability in 28 topcross hybrids of maize for grain yield and its components across two irrigated locations and to identify high yielding topcross hybrids for future testing and commercial utilization.

 

 

MATERIALS AND METHODS

    The plant material used consisted of 7 local inbred lines used as lines (L), and 4 introduced open pollinated varieties used as testers (T) crossed in line x tester arrangement (Table 1). Hand pollination was used to develop the breeding material. Pollen grain was collected into a paper bag from the tassel of male parent (tester) and then dusted on the silk of the female parent (line). The ear was covered with a bag and information regarding the cross was written on the bag. A total of 28 cross combinations was obtained through hand pollination. In July 2008, the 11 parental material and 28 cross hybrids were grown and evaluated at two irrigated locations, Medani, Gezira Research Station (GRS) and Matuq, Matuq Research Station (MRS), Gezira State, Sudan. The trials were arranged a randomized complete block design with three replicates. The plot size was maintained as 2 rows x 3 m long with inter and intra row spacings of 80 and 25 cm, respectively.  Seeds were sown at the rate of 3- 4 seeds per hill.  Plants were thinned to one plant per hill after three weeks from sowing. Nitrogen was applied at 86 kg/ha in a split dose after thinning and before flowering. The crop was irrigated at intervals of 10-14 days, and plots were kept free of weeds by hand weeding.  Data were analyzed using the Statistical Analysis System (SAS) computer package. The analysis was done for each season for characters days to 50% tasseling, plant height, ear length, ear diameter, kernels weight and grain yield and then combined. Mean performance was separated using Duncan's Multiple Range Test (DMRT). Data from each location was analyzed separately and across locations to determine the general and specific combining ability of each line was measured according to Griffing,s Method 2 (1956).

 

Table 1. Pedigree of the lines and testers used in the study.

Parents

Pedigree

Source 

L1

RING-B-S1-2    

Inbred line developed by ARC

L2

PR-89 B-5655-S1-1

Inbred line introduced from CIMMYT, Mexico

L3

RING-B- S1-3   

Inbred line developed by ARC

L4

RING- B-S1-1

Inbred line developed by ARC

L5

RING-A-S1-1

Inbred line developed by ARC

L6

RING-A-S1-2

Inbred line developed by ARC

L7

PR-89 B-5655-S1-3

Inbred line introduced from CIMMYT, Mexico

T1

SOBSIY-HG AB                        

OPV introduced from CIMMYT, Kenya

T2

ACROSS- 500 HGY-B             

OPV introduced from CIMMYT, Kenya

T3

CORRALE10 -02 SIYQ           

OPV introduced from CIMMYT,  Kenya

T4

BAILO- 02SIYQ                        

OPV introduced from CIMMYT,  Kenya

RESULTS AND DISCUSSION

 

   The performance of the material tested for most traits is high across the two locations. However, significant differences among the parents and their hybrids for most traits were shown indicating the diversity of the material.

Mean separation and ranking

    Mean days to 50% tasseling indicates that the pollen shedding ability of maize genotypes is an indicator of the earliness of genotypes. Mean days to tasseling across locations for parents scored 52 days as the general mean. Mean of parents ranged between 49 and 55 days for L6 and T3, respectively (Table 2). The mean of crosses ranged between 46 days for (T4 x L5) to 52 days for (T2 x L1) (Table 3). Identification of early tasseling genotypes is very important in developing hybrids and choosing hybrids to suit different agro-ecological zones as well as grower’s requirements. Earliness was a desirable trait especially under rainfed conditions. It is important for better use of water resources and avoidance of late season infestation with stem borers. Hence, the earliest crosses were T1 x L7 (47 days), T4 x L7 (47 days), T4 x L4 (48 days) and T4 x L6 (48 days) (Table 3).

 

Table 2. Mean performance of eleven parents for the measured traits in maize at the two locations, season 2008.

Traits /

Parents

       DT   

      PH   

        EL    

       ED    

        KW  

      GY  

Mean   Rank

Mean  Rank

 Mean Rank 

Mean  Rank  

 

Mean Rank

 

Mean   Rank

L1

49.1      10

131.4     10

14.2         4

3.7          3

20.7         6

   2.8         2

L2

50.0        9

148.5       4

15.0         1

3.6          7

19.9       11

   2.6         5

L3

51.7        6

145.2       6

13.2         9

3.6          6

20.7         8

   2.4         8

L4

50.0        8

152.0       3

14.3         3

4.1          1

20.3       10

   2.1        11

L5

51.7        5

145.6       5

13.7         5

3.6          4

22.6         2

   2.7         3

L6

49.1      11

139.1       9

13.4         8

3.4        11

22.1         3

   2.2         9

L7

50.1        7

131.1     11

12.7       11

3.4        10

20.7         7

   2.4         7

T1

52.7        4

139.3       8

13.6         7

3.9          2

21.3         5

   2.2       10

T2

54.2        2

155.9       2

14.8         2

3.6          5

21.7         4

   2.4         6

T3

55.2        1

157.7       1

13.7         6

3.5          8

22.8         1

   2.6         4

T4

52.8        3

143.2       7

12.9       10

3.5          9

20.5         9

   2.9         1

Mean

52.3

144.4

13.5

3.5

21.4

   2.4

CV%

  6.7

  10.0

13.0

9.8

14.5

 27.8

S.E±

  0.98

    2.33

0.38

0.08

  0.81

   0.15

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha).

 

    Tallness is not a good character in grain maize production, since tall maize plants tend to be susceptible to stem and root lodging.  Highly significant differences for tallness were detected among the studied parents with the general mean being of 144.4 cm. The trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. In the studied parents mean plant height ranged between 131.1 cm for L7 to 158 cm for T3 which was the tallest and latest parent across locations (Table 2). The crosses mean varied from 135.1 cm for (T3 x L7) to 155.9 cm for (T2 x L1).The tallest hybrids across locations were T4 x L6 and T4 x L3 (154 cm) (Table 3).

 

Table 3.  Performance of 28 crosses for the measured traits in maize at the two locations combined,  season 2008.

Traits/

Crosses

         DT                    PH                      EL                    ED                    KW                   GY

 

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

 

T1 x L1

48.5

   22

  14.6

13

14.2

    6

3.8       

 4

22.0    

  9

3.1       

  5

 

T1 x L2

48.5   

20

148.3   

14

14.2     

 7

3.5      

22

23.2    

  1

3.4         

  2

 

T1 x L3

50.0   

13

149.8   

 7

13.7    

18

3.7     

  9

21.7    

14

2.9       

12

 

T1 x L4

50.1   

12

145.0   

18

13.3    

22

3.7      

14

22.1    

  7

2.9       

11

 

T1 x L5

49.0    

19

145.6   

16

12.9    

25

3.5      

23

22.2    

  6

3.0       

10

 

T1 x L6

50.1   

11

152.3   

4

14.3    

  5

3.7      

11

21.8    

18

2.7       

21

 

T1 x L7

46.8   

27

138.9   

25

15.2      

  2

3.4      

26

20.3    

24

2.9       

16

 

T2 x L1

52.3   

  1

155.9   

 1

14.1    

  8

4.0      

  1

20.8    

22

3.3       

  3

 

T2 x L2

49.5   

17

149.2   

10

13.2    

21

3.7    

15

19.9    

27

2.4       

26

 

T2 x L3

51.2   

  4

145.2   

17

12.2    

27

3.7    

16

22.8    

  3

3.1       

  7

 

T2 x L4

50.2   

  9

141.0   

22

13.2    

24

3.7    

17

22.1    

  8

2.4       

15

 

T2 x L5

49.5     

18

140.8   

24

14.0    

10

3.7    

13

21.3    

17

2.0       

28

 

T2 x L6

50.0     

14

143.4   

19

14.6    

  4

3.3     

27

20.1    

25

3.1       

  8

 

T2 x L7

48.2     

21

149.1   

11

13.9    

14

3.4    

25

19.7    

28

3.5       

  1

 

T3 x L1

50.3     

  7

150.3   

 6

13.9    

12

3.6    

20

21.6    

16

2.8       

18

 

T3 x L2

49.7     

16

149.8   

 8

13.7    

16

3.7    

  7

21.7    

13

2.9       

13

 

T3 x L3

48.0     

23

139.2   

24

13.3    

20

3.8    

  2

22.4    

  5

2.7       

22

 

T3 x L4

50.2     

10

142.9   

21

11.9    

28

3.7    

12

20.6    

23

3.0       

  9

 

T3 x L5

51.2     

  3

151.4   

  5

16.1    

  1

3.6    

21

22.5    

  4

2.9       

17

 

T3 x L6

50.8     

  5

138.8   

26

13.9    

13

3.3    

28

20.9    

21

2.6       

24

 

T3 x L7

52.2     

  2

135.1   

28

14.1    

  9

3.5    

24

21.0    

20

2.2       

27

 

T4 x L1

50.3     

  8

146.1   

15

12.8    

26

3.7    

18

21.7    

12

2.5       

25

 

T4 x L2

50.0     

15

149.8   

 9

13.7    

17

3.7    

  8

21.7    

15

2.9       

14

 

T4 x L3

50.3     

  6

154.2   

 3

14.0    

11

3.8    

  5

23.2    

  2

3.1       

  6

 

T4 x L4

47.5     

25

148.9   

12

13.6    

19

3.7    

  6

21.8    

10

3.3       

  4

 

T4 x L5

45.7     

28

135.1   

27

13.8    

15

3.7    

10

21.8    

11

2.8       

19

 

T4 x L6

48.0     

24

154.2   

 2

13.2    

23

3.8    

  3

20.1    

26

2.7          

20

 

T4 x L7

47.2     

26

143.1   

20

15.2    

  3

3.6    

19

21.2    

19

2.6       

23

 

Mean

49

 

145.9

 

13.8

 

3.7

 

21.3

 

2.8

 

 

CV%

  6.7

 

10

 

13

 

9.8

 

14.5

 

 27.8     

 

 

S.E±

  0.64

 

    3.8

 

  0.46

 

0.08

 

  0.56

 

0.14

 

 
















 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm),  KW= kernels weight and GY= grain yield (t/ha).

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 

    The results indicate that crosses were later than their parents. Also, the taller crosses were late maturing than short ones. Generally, the crosses were taller than their parents which suggested some degree of hybrid vigor.

   Ear length trait is an important selection index for grain yield in maize. The ear length means of parents, as expected, were found to be shorter than those of the crosses at the two sites, with the general mean of 13.5 cm. The parents mean ranged between 12.7 cm for L7 to 15 cm for L2 (Table 2). The crosses mean varied from 11.9 cm for (T3 x L4) to 16.1 cm for (T3 x L5). However, long ear length were recorded for crosses T1 x L7 (15.2 cm), and T2 x L6 (14.6cm) (Table 3).Vedia and Claure (1995) found that ear length was the most important yield component and when used as a selection index genetic gain in recurrent selection reached 9.94% for yield and 5.75% for the ear traits. Therefore, any increase in ear length would be expected to increase number of kernels/row and hence increase grain yield.

    Ear diameter is a good indicator of the number of kernel rows/ear. The mean of ear diameter across sites for parents ranged between 3.4 cm for L6 and L7 to 4.1 cm for L4 (Table 2). Among the crosses, the large ear diameter ranged from 3.3 cm for T3 x L6 to 4.0 cm for T2 x L1. The crosses which had a big ear diameter were T3 x L3 and T4 x L6 (3.8cm) (Table 3). This result was in agreement with the findings of Tracy (1990) who found that the maize hybrids with high yield had more ears/plant, longer ears and a better ear shape and row configuration.

The mean of one hundred kernels weight for parents was 21.4 g, and it ranged between 19.9 g for L2 to 22.8 g for T3 (Table 2). Among the crosses, the mean was 21.3 g. The best crosses which obtained the highest kernel weight were T1 x L2 and T4 x L3 (23.2) followed by T2 x L3 (22.8 g) (Table 3).

Yield is a polygenic character is influenced by the fluctuating enviro-nment. Moreover, it is a complex trait depending on many components (Sharaan and Ghallab, 1997). In this study, there was a considerable amount of variability among the genotypes for this trait. The studied parents in the two locations showed a general mean of 2.4 t/ha. The parents means ranged between 2.12 t/ha for L4 to 2.93 t/ha for T4 (Table 2), while, the crosses means ranged between 2.0 t/ha for (T2 x L5) to 3.55 t/ha for (T2 x L7) (Table 3).  Most of the crosses (19 hybrids) had significantly higher mean grain yield than the overall mean. It is of interest to mention that the top ranking and the best yielder hybrids were T1 x L2 (3.4 t/ha), T2 x L1 (3.3 t/ha), T4 x L4 (3.3 t/ha), T1 x L1 (3.30 t/ha) and T4 x L3 (3.1 t/ha). These results agreed with those of Khalafalla and Abdalla (1997), who pointed to the fact that hybrids (crosses) produce higher grain yields than the open pollinated varieties due to the good performance of hybrids under Sudan conditions.

 Combining ability

    The breeding method to be adopted for improvement of a crop depends primarily on the nature of gene action involved in the expression of quantitative traits of economic importance. Combining ability leads to identification of parents with general combining ability effects and in locating cross combining showing high specific combining ability effects. In this study the ratio of GCA to SCA mean variance for most traits was less than one, suggesting that the inheritance of these traits was due to non additive gene action, with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and largely

controlled by additive gene action in the base material (Table 4).

Table 4. Mean squares of six agronomic traits for maize parents and 28 lines x tester crosses tested at two locations, Medani and Mutaq 2008.

Source of variation

DF

DT

PH

EL

ED

KW

GY

Location

  1

 3322.70**

13721**   

4.48**

50.90**

287.8**

 26.9**

Line

  6

     04.22

119.19

4.03

  0.02

    5.69

   0.27

Tester

  3

     18.81

   46.63

0.85

  0.07

    3.06

   0.28

Line x tester

18

     05.85**

   93.70*

2.38*

  0.05*

    1.62*

   0.44*

Line x tester x

location

18

     11.91**

217.90**

2.80

  0.11

     6.68

   0.63

Pooled error

76

     05.24

108.60

1.76

  0.04

     3.08

   0.19

GCA

 

       0.2

    -5.0

0.2

  0.00

    -0.7

   0.08

SCA

 

       0.6

    13.7

0.5

  0.02

     0.7

   0.03

GCA/SCA

 

       0.4

    -0.4

0.4

 -0.15

   -1.0

3.07

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 









 

   This result indicates that dominance and epistatic interaction effects seemed to be predomint for this trait and therefore heterosis breeding may be gratifying. The good combiner parents, those having negative GCA effects in Medani, for 50% days to tasseling were L5 followed by T4, T1 and L7, indicating earliness for flowering time, while, the latest, having positive GCA effect was T3 (Table 5).The earliest crosses having negative SCA effects were T3 x L6, T2 x L7 and T2 x L4, while, the latest crosses were T2 x L5, T4 x L5 and T4 x L4 (Table.6).

    The earliest parent in Mutaq was L7 (Table 5) and the earliest crosses were T2 x L4, T4 x L4 and T3 x L4 (Table 6). Common parents across locations that contributed to earliness were T4 and L5. The latest were L6 followed by T3 and T2 (Table 5). Parent L4 had good contribution for earliness to their hybrids progeny across locations.

Thus, the inbred lines which exhibited good general combining ability for at least one character can be used for development of early maturity and high grain yield. The contribution of the total variance for general and specific combining ability for this trait differs from location to another, but SCA was high in both locations (50.4% and 71.7%) compared with GCA which indicates that this trait is  controlled by additive gene action (Figs 1 and 2).

Trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. Only three top cross hybrid parents in Medani have   negative GCA effects for plant height, i.e., L7, L3 and T3; they were best combiners for short plant type. Tallness which is an undesirable trait is shown by parents L1, L2 and T1 (Table 5). Crosses having negative SCA effects and consequently short plant type were T4 x L2, T1 x L4 and T2 x L4,  while, tall hybrids with positive SCA effects were T3 x L1, T2 x L5 and T4 x L5 (Table 6).




 

   The best combiners for the short plant type with negative GCA in Mutaq were L7, L6 and L2 while, the taller parents with high positive GCA effects were L5 and L1 (Table 5). Among the crosses the shortest hybrids were T2 x L4, followed by T3 x L5 and the tallest hybrids were T2 x L5 and T3 x L4 (Table 6). This showed that, there is a relationship between late flowering and tall plant type. This is quite obvious among the hybrids such as T3 x L1 and T3 x L4.  Contribution for this trait is higher in crosses (80% and 53%) compared to parents (20% and 40%) at the two locations (Figs 1 and 2). The earliness and shortness are desirable traits especially under rainfed conditions for better water use efficiency and the escape of drought and avoidance of late season infestation with stem borer.

    Ear length is a good index for higher grain yield, therefore any increase in ear length would be expected to increase number of kernels/row and hence directly improve grain yield. In Medani site, the long ear length parents having a positively significant GCA effects such as L5, L7 and T1, while parents showing the short ear length were L4 and L2 (Table 5). The best crosses for this trait having a positive SCA effects and hence the longest ear length were T2 x L5 and T4 x L7. On the other hand the best combiners in Mutaq were L7 and T3 (Table 5), while the best crosses were T1 x L1 and T4 x L4 (Table 6). In the two locations, the best contribution was (73% and 65.9%) obtained by SCA compared with (27% and44.1) for GCA (Figs1 and 2). These results emphasized that ear length has a direct effect for improving grain yield. This is in agreement with the finding of Vedia and Claure (1995) who found that ear aspect was the most important yield component.

     Based on GCA estimates, the best combiners for ear diameter and length in Medani are L1 and L5, while best crosses were T1 x L2, T3 x L5 and T3 x L7. The good combiners in Mutaq site are L2, L3 and L4, while the best crosses are T3 x L4, T4 x L1 and T1 x L5 (Tables 5 and 6). A higher contribution among this trait is obtained by SCA (55.9% and 65.9%) in both locations compared with GCA (Figs 1and 2).

    Favorable GCA values were given by T1 and L3 as the good combiners for kernel weight in Medani and the best crosses were T4 x L7 and T2 x L7.  Among the studied parent material in Mutaq, only three parents have positive GCA effects (L3, L4 and T3).

 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

 Figure 1. Parent contribution of the maize GCA and SCA to the total variance

                of yield and its componets at Medani, season 2008.

 

 

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

Figure 2. Parent contribution of the maize GCA and SCA to the total variance of yield and its componets at Mutaq, season 2008.

 

The best crosses were shown by T1 x L4 and T3 x L4 (Tables 5 and 6). The higher average contribution was given by the SCA (50.8% and 61) compared with the GCA at two locations (Figs 1and 2). This indicted that the inheritance of this trait was controlled by non additive gene effects.

    At Medani site, all the results depicted in Table 5 showed that the parents differ considerably with respect to estimate of GCA effects for grain yield. The parents having positive GCA effects were T1 followed by L4 and L6. Parents having negative GCA effects were L2 and L6. The best crosses having positive SCA effects were T3 x L3 followed by T4 x L5 and T1 x L2. Negative SCA effects were shown by T3 x L4, T2 x L2 and T1 x L4 (Table 6). The higher combiner in Mutaq, were L2, L1 and L4. The best crosses were T3 x L5, T1 x L5 and T4 x L5, while negative SCA effects were shown by T1 x L3, T2 x L5 and T4 x L3 (Tables 5 and6). The great contribution was given by SCA (62.4% and 62%) compared with GCA at the two locations (Figs1 and 2).

     General combing ability variance for grain yield is greater than the mean square for specific combining ability indicating the importance of additive gene action in controlling grain yield. This finding is in agreement with that of Barakat and Osman (2008) who found GCA effects are larger than SCA effects for grain yield indicating that the additive genetic variance is a major source of variations responsible for inheritance of grain yield.

 

CONCLUSION

   The ratio of general combining ability variance for grain yield was greater than specific combining ability indicating the importance of additive gene action in controlling this trait hence the good combiner parent for grain yield across locations was L4 so it could be used in recurrent selection. Also enormous variability was detected in the studied population which makes cyclic selection more effective. The best cross was T4 x L5 indicating that dominance and epesitic interaction seemed to be predomint, hence, higher heterosis gratified and recommended cross T4 x L5 for future testing in multi-locations trials for commercial utilization in order to be released as a hybrid.

REFERENCES

Alhussein, M.B. 2007. Growth Performance and Grain Yield Stability of   some Open Pollinated arieties of Maize (Zea mays L.). M.Sc. Thesis, University of Gezira, Wad Medani Sudan.

Barakat, A.A and M.M. Osman. 2008. Gene action and combining ability estimates for some romising maize inbred lines by top cross system. Journal of Agricultural Sciences. Mansoura niversity Journal 33:280-709

Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel  rossing system. Australian Journal of Biological Sciences 9: 463-493.

 Khalafalla, M.M. and H.A. Abdalla. 1997. Performance of some maize genotypes (Zea mays L.) nd  their  F1  hybrid  for  yield  and  its components. University of Khartoum Journal of Agricultural Sciences 5(2): 56-68.

Meseka, S.K. 2000.DiallelAnalysis for Combining Ability of Grain Yield and Yield Components n Maize (Zea  mays L.). M.Sc. Thesis, Faculty of Agricultural Sciences, University of ezira, Wad Medani, Sudan.

McCann, J. 2005. Maize and Grace: Africa’s Encounter with a New Crop, 1500-2000. Harvard niversity Press, New York

Nour, A.M., I. N. Elzain and M.A. Dafalla. 1997. Crop Development and   Improvement. Annual Report f the Maize Research Program. Agricultural Research Corporation, Wad Medani,Sudan.

Sharaan, A.N. and K.H. Ghallab. 1997. Character association at different location in sesame. Sesame and Safflower Newsletter 12: 66-75.

Tracy, W.F. 1990.  Potential of field corn germplasm for improvement of sweet corn. Crop Science 30:1041-1045.

Vedia, M.L. and E.T. Claure. 1995. Selection index for yield in the maize population. Crop Science 7: 505-510.

 

 

 

 

 

 

ABSTRACT

 

   The development of hybrids is the main objective of maize breeding. However, success depends largely on the identification of the best parents to ensure maximum combining ability. This study was conducted to estimate genetic variability and combining ability for grain yield and yield components of seven local inbred lines and four introduced open pollinated varieties of maize (Zea mays L.) across two irrigated locations, Medani and Matuq, Gezira, Sudan in 2008. The experiment was arranged in a randomized complete block design with three replicates. The traits measured were days to 50% tassel, plant height, ear length, ear diameter, hundred kernels weight and grain yield. Significant differences were observed among the parents and crosses for most of studied traits in both seasons. The crosses showed high genetic variability and tall plants than their parents which suggested some degree of hybrid vigor. The tallest hybrids across locations were T3 x L5 and T4 x L3. This indicates that the crosses were late maturing than their parents. The highest yielding hybrids had long ears and better shape, e.g., T2 x L1 and T1 x L7.The top five ranking crosses for grain yield across locations were T2 x L7 (3.45 t/ha), T1 x L2 (3.44 t/ha), T2 x LI (3.32 t/ha), T4 x L4 (3.30 t/ha) and T1 x L1 (3.13 t/ha).   The inheritance of most traits was controlled by non-additive gene action except ear height and grain yield. The best combiners for grain in Medani were T4, L4 and L5, while in Mutaq were L2, L4 and L6. The ratio of GCA to SCA variance for the most traits was less than one, suggesting that the inheritance was due to non additive gene effect with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and was largely controlled by additive gene action in the base material. From these results it is recommended that parents T4, L1 and L6 to be used in recurrent selection, while, crosses T3 x L5, T1 x L5 and T4 x L6 to be tested in multi-locations trials for commercial utilization.

 

INTRODUCTION

       Maize generally is one of the most diverse crop both genetically and phenotypically. Due to its wide adaptability and productivity, maize spread rapidly around the world after the Europeans brought the crop from the Americas in the 15th and 16th centuries (McCann, 2005). The Portuguese introduced the crop to Africa at the beginning of the 16th century and since then the crop has replaced sorghum and millet as the main staple food in most of the continent where the climatic conditions are favorable (McCann, 2005). Today, there  is an increasing interest in maize production in Sudan due to its suitability to cultivation in the agricultural irrigated schemes, especially in the Gezira.It can occupy an important position in the economy of the country due to the possibility of blending it with wheat for making bread (Nour et al., 1997; Meseka, 2000).

    The grain yield of existing maize varieties and local landraces in Sudan is low. Also, maize   hybrids have been reported to show high potential for grain yield than the open pollinated varieties and landraces (Alhussein, 2007). Advantages of hybrids over open pollinated cultivars are higher yield, uniformity, high quality and resistance to diseases and pests. In spite of having yield potential, the production of maize in Sudan is very low. One of the reasons for this is the cultivation of exotic hybrids, which are not well adapted to our agro-climatic conditions. One of the strategies of the Agricultural Research Corporation (ARC) of the Sudan for maize breeding program is to develop new hybrids as an attempt to incorporate both advantages for higher yield and adaptability to environmental conditions. Thus, getting the benefit from the use of hybrids is the main purpose in maize breeding program of ARC.    Therefore, the objective of this study is to estimate the magnitude of combining ability in 28 topcross hybrids of maize for grain yield and its components across two irrigated locations and to identify high yielding topcross hybrids for future testing and commercial utilization.

 

 

MATERIALS AND METHODS

    The plant material used consisted of 7 local inbred lines used as lines (L), and 4 introduced open pollinated varieties used as testers (T) crossed in line x tester arrangement (Table 1). Hand pollination was used to develop the breeding material. Pollen grain was collected into a paper bag from the tassel of male parent (tester) and then dusted on the silk of the female parent (line). The ear was covered with a bag and information regarding the cross was written on the bag. A total of 28 cross combinations was obtained through hand pollination. In July 2008, the 11 parental material and 28 cross hybrids were grown and evaluated at two irrigated locations, Medani, Gezira Research Station (GRS) and Matuq, Matuq Research Station (MRS), Gezira State, Sudan. The trials were arranged a randomized complete block design with three replicates. The plot size was maintained as 2 rows x 3 m long with inter and intra row spacings of 80 and 25 cm, respectively.  Seeds were sown at the rate of 3- 4 seeds per hill.  Plants were thinned to one plant per hill after three weeks from sowing. Nitrogen was applied at 86 kg/ha in a split dose after thinning and before flowering. The crop was irrigated at intervals of 10-14 days, and plots were kept free of weeds by hand weeding.  Data were analyzed using the Statistical Analysis System (SAS) computer package. The analysis was done for each season for characters days to 50% tasseling, plant height, ear length, ear diameter, kernels weight and grain yield and then combined. Mean performance was separated using Duncan's Multiple Range Test (DMRT). Data from each location was analyzed separately and across locations to determine the general and specific combining ability of each line was measured according to Griffing,s Method 2 (1956).

 

Table 1. Pedigree of the lines and testers used in the study.

Parents

Pedigree

Source 

L1

RING-B-S1-2    

Inbred line developed by ARC

L2

PR-89 B-5655-S1-1

Inbred line introduced from CIMMYT, Mexico

L3

RING-B- S1-3   

Inbred line developed by ARC

L4

RING- B-S1-1

Inbred line developed by ARC

L5

RING-A-S1-1

Inbred line developed by ARC

L6

RING-A-S1-2

Inbred line developed by ARC

L7

PR-89 B-5655-S1-3

Inbred line introduced from CIMMYT, Mexico

T1

SOBSIY-HG AB                        

OPV introduced from CIMMYT, Kenya

T2

ACROSS- 500 HGY-B             

OPV introduced from CIMMYT, Kenya

T3

CORRALE10 -02 SIYQ           

OPV introduced from CIMMYT,  Kenya

T4

BAILO- 02SIYQ                        

OPV introduced from CIMMYT,  Kenya

RESULTS AND DISCUSSION

 

   The performance of the material tested for most traits is high across the two locations. However, significant differences among the parents and their hybrids for most traits were shown indicating the diversity of the material.

Mean separation and ranking

    Mean days to 50% tasseling indicates that the pollen shedding ability of maize genotypes is an indicator of the earliness of genotypes. Mean days to tasseling across locations for parents scored 52 days as the general mean. Mean of parents ranged between 49 and 55 days for L6 and T3, respectively (Table 2). The mean of crosses ranged between 46 days for (T4 x L5) to 52 days for (T2 x L1) (Table 3). Identification of early tasseling genotypes is very important in developing hybrids and choosing hybrids to suit different agro-ecological zones as well as grower’s requirements. Earliness was a desirable trait especially under rainfed conditions. It is important for better use of water resources and avoidance of late season infestation with stem borers. Hence, the earliest crosses were T1 x L7 (47 days), T4 x L7 (47 days), T4 x L4 (48 days) and T4 x L6 (48 days) (Table 3).

 

Table 2. Mean performance of eleven parents for the measured traits in maize at the two locations, season 2008.

Traits /

Parents

       DT   

      PH   

        EL    

       ED    

        KW  

      GY  

Mean   Rank

Mean  Rank

 Mean Rank 

Mean  Rank  

 

Mean Rank

 

Mean   Rank

L1

49.1      10

131.4     10

14.2         4

3.7          3

20.7         6

   2.8         2

L2

50.0        9

148.5       4

15.0         1

3.6          7

19.9       11

   2.6         5

L3

51.7        6

145.2       6

13.2         9

3.6          6

20.7         8

   2.4         8

L4

50.0        8

152.0       3

14.3         3

4.1          1

20.3       10

   2.1        11

L5

51.7        5

145.6       5

13.7         5

3.6          4

22.6         2

   2.7         3

L6

49.1      11

139.1       9

13.4         8

3.4        11

22.1         3

   2.2         9

L7

50.1        7

131.1     11

12.7       11

3.4        10

20.7         7

   2.4         7

T1

52.7        4

139.3       8

13.6         7

3.9          2

21.3         5

   2.2       10

T2

54.2        2

155.9       2

14.8         2

3.6          5

21.7         4

   2.4         6

T3

55.2        1

157.7       1

13.7         6

3.5          8

22.8         1

   2.6         4

T4

52.8        3

143.2       7

12.9       10

3.5          9

20.5         9

   2.9         1

Mean

52.3

144.4

13.5

3.5

21.4

   2.4

CV%

  6.7

  10.0

13.0

9.8

14.5

 27.8

S.E±

  0.98

    2.33

0.38

0.08

  0.81

   0.15

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha).

 

    Tallness is not a good character in grain maize production, since tall maize plants tend to be susceptible to stem and root lodging.  Highly significant differences for tallness were detected among the studied parents with the general mean being of 144.4 cm. The trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. In the studied parents mean plant height ranged between 131.1 cm for L7 to 158 cm for T3 which was the tallest and latest parent across locations (Table 2). The crosses mean varied from 135.1 cm for (T3 x L7) to 155.9 cm for (T2 x L1).The tallest hybrids across locations were T4 x L6 and T4 x L3 (154 cm) (Table 3).

 

Table 3.  Performance of 28 crosses for the measured traits in maize at the two locations combined,  season 2008.

Traits/

Crosses

         DT                    PH                      EL                    ED                    KW                   GY

 

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

 

T1 x L1

48.5

   22

  14.6

13

14.2

    6

3.8       

 4

22.0    

  9

3.1       

  5

 

T1 x L2

48.5   

20

148.3   

14

14.2     

 7

3.5      

22

23.2    

  1

3.4         

  2

 

T1 x L3

50.0   

13

149.8   

 7

13.7    

18

3.7     

  9

21.7    

14

2.9       

12

 

T1 x L4

50.1   

12

145.0   

18

13.3    

22

3.7      

14

22.1    

  7

2.9       

11

 

T1 x L5

49.0    

19

145.6   

16

12.9    

25

3.5      

23

22.2    

  6

3.0       

10

 

T1 x L6

50.1   

11

152.3   

4

14.3    

  5

3.7      

11

21.8    

18

2.7       

21

 

T1 x L7

46.8   

27

138.9   

25

15.2      

  2

3.4      

26

20.3    

24

2.9       

16

 

T2 x L1

52.3   

  1

155.9   

 1

14.1    

  8

4.0      

  1

20.8    

22

3.3       

  3

 

T2 x L2

49.5   

17

149.2   

10

13.2    

21

3.7    

15

19.9    

27

2.4       

26

 

T2 x L3

51.2   

  4

145.2   

17

12.2    

27

3.7    

16

22.8    

  3

3.1       

  7

 

T2 x L4

50.2   

  9

141.0   

22

13.2    

24

3.7    

17

22.1    

  8

2.4       

15

 

T2 x L5

49.5     

18

140.8   

24

14.0    

10

3.7    

13

21.3    

17

2.0       

28

 

T2 x L6

50.0     

14

143.4   

19

14.6    

  4

3.3     

27

20.1    

25

3.1       

  8

 

T2 x L7

48.2     

21

149.1   

11

13.9    

14

3.4    

25

19.7    

28

3.5       

  1

 

T3 x L1

50.3     

  7

150.3   

 6

13.9    

12

3.6    

20

21.6    

16

2.8       

18

 

T3 x L2

49.7     

16

149.8   

 8

13.7    

16

3.7    

  7

21.7    

13

2.9       

13

 

T3 x L3

48.0     

23

139.2   

24

13.3    

20

3.8    

  2

22.4    

  5

2.7       

22

 

T3 x L4

50.2     

10

142.9   

21

11.9    

28

3.7    

12

20.6    

23

3.0       

  9

 

T3 x L5

51.2     

  3

151.4   

  5

16.1    

  1

3.6    

21

22.5    

  4

2.9       

17

 

T3 x L6

50.8     

  5

138.8   

26

13.9    

13

3.3    

28

20.9    

21

2.6       

24

 

T3 x L7

52.2     

  2

135.1   

28

14.1    

  9

3.5    

24

21.0    

20

2.2       

27

 

T4 x L1

50.3     

  8

146.1   

15

12.8    

26

3.7    

18

21.7    

12

2.5       

25

 

T4 x L2

50.0     

15

149.8   

 9

13.7    

17

3.7    

  8

21.7    

15

2.9       

14

 

T4 x L3

50.3     

  6

154.2   

 3

14.0    

11

3.8    

  5

23.2    

  2

3.1       

  6

 

T4 x L4

47.5     

25

148.9   

12

13.6    

19

3.7    

  6

21.8    

10

3.3       

  4

 

T4 x L5

45.7     

28

135.1   

27

13.8    

15

3.7    

10

21.8    

11

2.8       

19

 

T4 x L6

48.0     

24

154.2   

 2

13.2    

23

3.8    

  3

20.1    

26

2.7          

20

 

T4 x L7

47.2     

26

143.1   

20

15.2    

  3

3.6    

19

21.2    

19

2.6       

23

 

Mean

49

 

145.9

 

13.8

 

3.7

 

21.3

 

2.8

 

 

CV%

  6.7

 

10

 

13

 

9.8

 

14.5

 

 27.8     

 

 

S.E±

  0.64

 

    3.8

 

  0.46

 

0.08

 

  0.56

 

0.14

 

 
















 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm),  KW= kernels weight and GY= grain yield (t/ha).

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 

    The results indicate that crosses were later than their parents. Also, the taller crosses were late maturing than short ones. Generally, the crosses were taller than their parents which suggested some degree of hybrid vigor.

   Ear length trait is an important selection index for grain yield in maize. The ear length means of parents, as expected, were found to be shorter than those of the crosses at the two sites, with the general mean of 13.5 cm. The parents mean ranged between 12.7 cm for L7 to 15 cm for L2 (Table 2). The crosses mean varied from 11.9 cm for (T3 x L4) to 16.1 cm for (T3 x L5). However, long ear length were recorded for crosses T1 x L7 (15.2 cm), and T2 x L6 (14.6cm) (Table 3).Vedia and Claure (1995) found that ear length was the most important yield component and when used as a selection index genetic gain in recurrent selection reached 9.94% for yield and 5.75% for the ear traits. Therefore, any increase in ear length would be expected to increase number of kernels/row and hence increase grain yield.

    Ear diameter is a good indicator of the number of kernel rows/ear. The mean of ear diameter across sites for parents ranged between 3.4 cm for L6 and L7 to 4.1 cm for L4 (Table 2). Among the crosses, the large ear diameter ranged from 3.3 cm for T3 x L6 to 4.0 cm for T2 x L1. The crosses which had a big ear diameter were T3 x L3 and T4 x L6 (3.8cm) (Table 3). This result was in agreement with the findings of Tracy (1990) who found that the maize hybrids with high yield had more ears/plant, longer ears and a better ear shape and row configuration.

The mean of one hundred kernels weight for parents was 21.4 g, and it ranged between 19.9 g for L2 to 22.8 g for T3 (Table 2). Among the crosses, the mean was 21.3 g. The best crosses which obtained the highest kernel weight were T1 x L2 and T4 x L3 (23.2) followed by T2 x L3 (22.8 g) (Table 3).

Yield is a polygenic character is influenced by the fluctuating enviro-nment. Moreover, it is a complex trait depending on many components (Sharaan and Ghallab, 1997). In this study, there was a considerable amount of variability among the genotypes for this trait. The studied parents in the two locations showed a general mean of 2.4 t/ha. The parents means ranged between 2.12 t/ha for L4 to 2.93 t/ha for T4 (Table 2), while, the crosses means ranged between 2.0 t/ha for (T2 x L5) to 3.55 t/ha for (T2 x L7) (Table 3).  Most of the crosses (19 hybrids) had significantly higher mean grain yield than the overall mean. It is of interest to mention that the top ranking and the best yielder hybrids were T1 x L2 (3.4 t/ha), T2 x L1 (3.3 t/ha), T4 x L4 (3.3 t/ha), T1 x L1 (3.30 t/ha) and T4 x L3 (3.1 t/ha). These results agreed with those of Khalafalla and Abdalla (1997), who pointed to the fact that hybrids (crosses) produce higher grain yields than the open pollinated varieties due to the good performance of hybrids under Sudan conditions.

 Combining ability

    The breeding method to be adopted for improvement of a crop depends primarily on the nature of gene action involved in the expression of quantitative traits of economic importance. Combining ability leads to identification of parents with general combining ability effects and in locating cross combining showing high specific combining ability effects. In this study the ratio of GCA to SCA mean variance for most traits was less than one, suggesting that the inheritance of these traits was due to non additive gene action, with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and largely

controlled by additive gene action in the base material (Table 4).

Table 4. Mean squares of six agronomic traits for maize parents and 28 lines x tester crosses tested at two locations, Medani and Mutaq 2008.

Source of variation

DF

DT

PH

EL

ED

KW

GY

Location

  1

 3322.70**

13721**   

4.48**

50.90**

287.8**

 26.9**

Line

  6

     04.22

119.19

4.03

  0.02

    5.69

   0.27

Tester

  3

     18.81

   46.63

0.85

  0.07

    3.06

   0.28

Line x tester

18

     05.85**

   93.70*

2.38*

  0.05*

    1.62*

   0.44*

Line x tester x

location

18

     11.91**

217.90**

2.80

  0.11

     6.68

   0.63

Pooled error

76

     05.24

108.60

1.76

  0.04

     3.08

   0.19

GCA

 

       0.2

    -5.0

0.2

  0.00

    -0.7

   0.08

SCA

 

       0.6

    13.7

0.5

  0.02

     0.7

   0.03

GCA/SCA

 

       0.4

    -0.4

0.4

 -0.15

   -1.0

3.07

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 









 

   This result indicates that dominance and epistatic interaction effects seemed to be predomint for this trait and therefore heterosis breeding may be gratifying. The good combiner parents, those having negative GCA effects in Medani, for 50% days to tasseling were L5 followed by T4, T1 and L7, indicating earliness for flowering time, while, the latest, having positive GCA effect was T3 (Table 5).The earliest crosses having negative SCA effects were T3 x L6, T2 x L7 and T2 x L4, while, the latest crosses were T2 x L5, T4 x L5 and T4 x L4 (Table.6).

    The earliest parent in Mutaq was L7 (Table 5) and the earliest crosses were T2 x L4, T4 x L4 and T3 x L4 (Table 6). Common parents across locations that contributed to earliness were T4 and L5. The latest were L6 followed by T3 and T2 (Table 5). Parent L4 had good contribution for earliness to their hybrids progeny across locations.

Thus, the inbred lines which exhibited good general combining ability for at least one character can be used for development of early maturity and high grain yield. The contribution of the total variance for general and specific combining ability for this trait differs from location to another, but SCA was high in both locations (50.4% and 71.7%) compared with GCA which indicates that this trait is  controlled by additive gene action (Figs 1 and 2).

Trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. Only three top cross hybrid parents in Medani have   negative GCA effects for plant height, i.e., L7, L3 and T3; they were best combiners for short plant type. Tallness which is an undesirable trait is shown by parents L1, L2 and T1 (Table 5). Crosses having negative SCA effects and consequently short plant type were T4 x L2, T1 x L4 and T2 x L4,  while, tall hybrids with positive SCA effects were T3 x L1, T2 x L5 and T4 x L5 (Table 6).




 

   The best combiners for the short plant type with negative GCA in Mutaq were L7, L6 and L2 while, the taller parents with high positive GCA effects were L5 and L1 (Table 5). Among the crosses the shortest hybrids were T2 x L4, followed by T3 x L5 and the tallest hybrids were T2 x L5 and T3 x L4 (Table 6). This showed that, there is a relationship between late flowering and tall plant type. This is quite obvious among the hybrids such as T3 x L1 and T3 x L4.  Contribution for this trait is higher in crosses (80% and 53%) compared to parents (20% and 40%) at the two locations (Figs 1 and 2). The earliness and shortness are desirable traits especially under rainfed conditions for better water use efficiency and the escape of drought and avoidance of late season infestation with stem borer.

    Ear length is a good index for higher grain yield, therefore any increase in ear length would be expected to increase number of kernels/row and hence directly improve grain yield. In Medani site, the long ear length parents having a positively significant GCA effects such as L5, L7 and T1, while parents showing the short ear length were L4 and L2 (Table 5). The best crosses for this trait having a positive SCA effects and hence the longest ear length were T2 x L5 and T4 x L7. On the other hand the best combiners in Mutaq were L7 and T3 (Table 5), while the best crosses were T1 x L1 and T4 x L4 (Table 6). In the two locations, the best contribution was (73% and 65.9%) obtained by SCA compared with (27% and44.1) for GCA (Figs1 and 2). These results emphasized that ear length has a direct effect for improving grain yield. This is in agreement with the finding of Vedia and Claure (1995) who found that ear aspect was the most important yield component.

     Based on GCA estimates, the best combiners for ear diameter and length in Medani are L1 and L5, while best crosses were T1 x L2, T3 x L5 and T3 x L7. The good combiners in Mutaq site are L2, L3 and L4, while the best crosses are T3 x L4, T4 x L1 and T1 x L5 (Tables 5 and 6). A higher contribution among this trait is obtained by SCA (55.9% and 65.9%) in both locations compared with GCA (Figs 1and 2).

    Favorable GCA values were given by T1 and L3 as the good combiners for kernel weight in Medani and the best crosses were T4 x L7 and T2 x L7.  Among the studied parent material in Mutaq, only three parents have positive GCA effects (L3, L4 and T3).

 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

 Figure 1. Parent contribution of the maize GCA and SCA to the total variance

                of yield and its componets at Medani, season 2008.

 

 

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

Figure 2. Parent contribution of the maize GCA and SCA to the total variance of yield and its componets at Mutaq, season 2008.

 

The best crosses were shown by T1 x L4 and T3 x L4 (Tables 5 and 6). The higher average contribution was given by the SCA (50.8% and 61) compared with the GCA at two locations (Figs 1and 2). This indicted that the inheritance of this trait was controlled by non additive gene effects.

    At Medani site, all the results depicted in Table 5 showed that the parents differ considerably with respect to estimate of GCA effects for grain yield. The parents having positive GCA effects were T1 followed by L4 and L6. Parents having negative GCA effects were L2 and L6. The best crosses having positive SCA effects were T3 x L3 followed by T4 x L5 and T1 x L2. Negative SCA effects were shown by T3 x L4, T2 x L2 and T1 x L4 (Table 6). The higher combiner in Mutaq, were L2, L1 and L4. The best crosses were T3 x L5, T1 x L5 and T4 x L5, while negative SCA effects were shown by T1 x L3, T2 x L5 and T4 x L3 (Tables 5 and6). The great contribution was given by SCA (62.4% and 62%) compared with GCA at the two locations (Figs1 and 2).

     General combing ability variance for grain yield is greater than the mean square for specific combining ability indicating the importance of additive gene action in controlling grain yield. This finding is in agreement with that of Barakat and Osman (2008) who found GCA effects are larger than SCA effects for grain yield indicating that the additive genetic variance is a major source of variations responsible for inheritance of grain yield.

 

CONCLUSION

   The ratio of general combining ability variance for grain yield was greater than specific combining ability indicating the importance of additive gene action in controlling this trait hence the good combiner parent for grain yield across locations was L4 so it could be used in recurrent selection. Also enormous variability was detected in the studied population which makes cyclic selection more effective. The best cross was T4 x L5 indicating that dominance and epesitic interaction seemed to be predomint, hence, higher heterosis gratified and recommended cross T4 x L5 for future testing in multi-locations trials for commercial utilization in order to be released as a hybrid.

REFERENCES

Alhussein, M.B. 2007. Growth Performance and Grain Yield Stability of   some Open Pollinated arieties of Maize (Zea mays L.). M.Sc. Thesis, University of Gezira, Wad Medani Sudan.

Barakat, A.A and M.M. Osman. 2008. Gene action and combining ability estimates for some romising maize inbred lines by top cross system. Journal of Agricultural Sciences. Mansoura niversity Journal 33:280-709

Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel  rossing system. Australian Journal of Biological Sciences 9: 463-493.

 Khalafalla, M.M. and H.A. Abdalla. 1997. Performance of some maize genotypes (Zea mays L.) nd  their  F1  hybrid  for  yield  and  its components. University of Khartoum Journal of Agricultural Sciences 5(2): 56-68.

Meseka, S.K. 2000.DiallelAnalysis for Combining Ability of Grain Yield and Yield Components n Maize (Zea  mays L.). M.Sc. Thesis, Faculty of Agricultural Sciences, University of ezira, Wad Medani, Sudan.

McCann, J. 2005. Maize and Grace: Africa’s Encounter with a New Crop, 1500-2000. Harvard niversity Press, New York

Nour, A.M., I. N. Elzain and M.A. Dafalla. 1997. Crop Development and   Improvement. Annual Report f the Maize Research Program. Agricultural Research Corporation, Wad Medani,Sudan.

Sharaan, A.N. and K.H. Ghallab. 1997. Character association at different location in sesame. Sesame and Safflower Newsletter 12: 66-75.

Tracy, W.F. 1990.  Potential of field corn germplasm for improvement of sweet corn. Crop Science 30:1041-1045.

Vedia, M.L. and E.T. Claure. 1995. Selection index for yield in the maize population. Crop Science 7: 505-510.

 

 

 

ABSTRACT

 

   The development of hybrids is the main objective of maize breeding. However, success depends largely on the identification of the best parents to ensure maximum combining ability. This study was conducted to estimate genetic variability and combining ability for grain yield and yield components of seven local inbred lines and four introduced open pollinated varieties of maize (Zea mays L.) across two irrigated locations, Medani and Matuq, Gezira, Sudan in 2008. The experiment was arranged in a randomized complete block design with three replicates. The traits measured were days to 50% tassel, plant height, ear length, ear diameter, hundred kernels weight and grain yield. Significant differences were observed among the parents and crosses for most of studied traits in both seasons. The crosses showed high genetic variability and tall plants than their parents which suggested some degree of hybrid vigor. The tallest hybrids across locations were T3 x L5 and T4 x L3. This indicates that the crosses were late maturing than their parents. The highest yielding hybrids had long ears and better shape, e.g., T2 x L1 and T1 x L7.The top five ranking crosses for grain yield across locations were T2 x L7 (3.45 t/ha), T1 x L2 (3.44 t/ha), T2 x LI (3.32 t/ha), T4 x L4 (3.30 t/ha) and T1 x L1 (3.13 t/ha).   The inheritance of most traits was controlled by non-additive gene action except ear height and grain yield. The best combiners for grain in Medani were T4, L4 and L5, while in Mutaq were L2, L4 and L6. The ratio of GCA to SCA variance for the most traits was less than one, suggesting that the inheritance was due to non additive gene effect with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and was largely controlled by additive gene action in the base material. From these results it is recommended that parents T4, L1 and L6 to be used in recurrent selection, while, crosses T3 x L5, T1 x L5 and T4 x L6 to be tested in multi-locations trials for commercial utilization.

 

INTRODUCTION

       Maize generally is one of the most diverse crop both genetically and phenotypically. Due to its wide adaptability and productivity, maize spread rapidly around the world after the Europeans brought the crop from the Americas in the 15th and 16th centuries (McCann, 2005). The Portuguese introduced the crop to Africa at the beginning of the 16th century and since then the crop has replaced sorghum and millet as the main staple food in most of the continent where the climatic conditions are favorable (McCann, 2005). Today, there  is an increasing interest in maize production in Sudan due to its suitability to cultivation in the agricultural irrigated schemes, especially in the Gezira.It can occupy an important position in the economy of the country due to the possibility of blending it with wheat for making bread (Nour et al., 1997; Meseka, 2000).

    The grain yield of existing maize varieties and local landraces in Sudan is low. Also, maize   hybrids have been reported to show high potential for grain yield than the open pollinated varieties and landraces (Alhussein, 2007). Advantages of hybrids over open pollinated cultivars are higher yield, uniformity, high quality and resistance to diseases and pests. In spite of having yield potential, the production of maize in Sudan is very low. One of the reasons for this is the cultivation of exotic hybrids, which are not well adapted to our agro-climatic conditions. One of the strategies of the Agricultural Research Corporation (ARC) of the Sudan for maize breeding program is to develop new hybrids as an attempt to incorporate both advantages for higher yield and adaptability to environmental conditions. Thus, getting the benefit from the use of hybrids is the main purpose in maize breeding program of ARC.    Therefore, the objective of this study is to estimate the magnitude of combining ability in 28 topcross hybrids of maize for grain yield and its components across two irrigated locations and to identify high yielding topcross hybrids for future testing and commercial utilization.

 

 

MATERIALS AND METHODS

    The plant material used consisted of 7 local inbred lines used as lines (L), and 4 introduced open pollinated varieties used as testers (T) crossed in line x tester arrangement (Table 1). Hand pollination was used to develop the breeding material. Pollen grain was collected into a paper bag from the tassel of male parent (tester) and then dusted on the silk of the female parent (line). The ear was covered with a bag and information regarding the cross was written on the bag. A total of 28 cross combinations was obtained through hand pollination. In July 2008, the 11 parental material and 28 cross hybrids were grown and evaluated at two irrigated locations, Medani, Gezira Research Station (GRS) and Matuq, Matuq Research Station (MRS), Gezira State, Sudan. The trials were arranged a randomized complete block design with three replicates. The plot size was maintained as 2 rows x 3 m long with inter and intra row spacings of 80 and 25 cm, respectively.  Seeds were sown at the rate of 3- 4 seeds per hill.  Plants were thinned to one plant per hill after three weeks from sowing. Nitrogen was applied at 86 kg/ha in a split dose after thinning and before flowering. The crop was irrigated at intervals of 10-14 days, and plots were kept free of weeds by hand weeding.  Data were analyzed using the Statistical Analysis System (SAS) computer package. The analysis was done for each season for characters days to 50% tasseling, plant height, ear length, ear diameter, kernels weight and grain yield and then combined. Mean performance was separated using Duncan's Multiple Range Test (DMRT). Data from each location was analyzed separately and across locations to determine the general and specific combining ability of each line was measured according to Griffing,s Method 2 (1956).

 

Table 1. Pedigree of the lines and testers used in the study.

Parents

Pedigree

Source 

L1

RING-B-S1-2    

Inbred line developed by ARC

L2

PR-89 B-5655-S1-1

Inbred line introduced from CIMMYT, Mexico

L3

RING-B- S1-3   

Inbred line developed by ARC

L4

RING- B-S1-1

Inbred line developed by ARC

L5

RING-A-S1-1

Inbred line developed by ARC

L6

RING-A-S1-2

Inbred line developed by ARC

L7

PR-89 B-5655-S1-3

Inbred line introduced from CIMMYT, Mexico

T1

SOBSIY-HG AB                        

OPV introduced from CIMMYT, Kenya

T2

ACROSS- 500 HGY-B             

OPV introduced from CIMMYT, Kenya

T3

CORRALE10 -02 SIYQ           

OPV introduced from CIMMYT,  Kenya

T4

BAILO- 02SIYQ                        

OPV introduced from CIMMYT,  Kenya

RESULTS AND DISCUSSION

 

   The performance of the material tested for most traits is high across the two locations. However, significant differences among the parents and their hybrids for most traits were shown indicating the diversity of the material.

Mean separation and ranking

    Mean days to 50% tasseling indicates that the pollen shedding ability of maize genotypes is an indicator of the earliness of genotypes. Mean days to tasseling across locations for parents scored 52 days as the general mean. Mean of parents ranged between 49 and 55 days for L6 and T3, respectively (Table 2). The mean of crosses ranged between 46 days for (T4 x L5) to 52 days for (T2 x L1) (Table 3). Identification of early tasseling genotypes is very important in developing hybrids and choosing hybrids to suit different agro-ecological zones as well as grower’s requirements. Earliness was a desirable trait especially under rainfed conditions. It is important for better use of water resources and avoidance of late season infestation with stem borers. Hence, the earliest crosses were T1 x L7 (47 days), T4 x L7 (47 days), T4 x L4 (48 days) and T4 x L6 (48 days) (Table 3).

 

Table 2. Mean performance of eleven parents for the measured traits in maize at the two locations, season 2008.

Traits /

Parents

       DT   

      PH   

        EL    

       ED    

        KW  

      GY  

Mean   Rank

Mean  Rank

 Mean Rank 

Mean  Rank  

 

Mean Rank

 

Mean   Rank

L1

49.1      10

131.4     10

14.2         4

3.7          3

20.7         6

   2.8         2

L2

50.0        9

148.5       4

15.0         1

3.6          7

19.9       11

   2.6         5

L3

51.7        6

145.2       6

13.2         9

3.6          6

20.7         8

   2.4         8

L4

50.0        8

152.0       3

14.3         3

4.1          1

20.3       10

   2.1        11

L5

51.7        5

145.6       5

13.7         5

3.6          4

22.6         2

   2.7         3

L6

49.1      11

139.1       9

13.4         8

3.4        11

22.1         3

   2.2         9

L7

50.1        7

131.1     11

12.7       11

3.4        10

20.7         7

   2.4         7

T1

52.7        4

139.3       8

13.6         7

3.9          2

21.3         5

   2.2       10

T2

54.2        2

155.9       2

14.8         2

3.6          5

21.7         4

   2.4         6

T3

55.2        1

157.7       1

13.7         6

3.5          8

22.8         1

   2.6         4

T4

52.8        3

143.2       7

12.9       10

3.5          9

20.5         9

   2.9         1

Mean

52.3

144.4

13.5

3.5

21.4

   2.4

CV%

  6.7

  10.0

13.0

9.8

14.5

 27.8

S.E±

  0.98

    2.33

0.38

0.08

  0.81

   0.15

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha).

 

    Tallness is not a good character in grain maize production, since tall maize plants tend to be susceptible to stem and root lodging.  Highly significant differences for tallness were detected among the studied parents with the general mean being of 144.4 cm. The trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. In the studied parents mean plant height ranged between 131.1 cm for L7 to 158 cm for T3 which was the tallest and latest parent across locations (Table 2). The crosses mean varied from 135.1 cm for (T3 x L7) to 155.9 cm for (T2 x L1).The tallest hybrids across locations were T4 x L6 and T4 x L3 (154 cm) (Table 3).

 

Table 3.  Performance of 28 crosses for the measured traits in maize at the two locations combined,  season 2008.

Traits/

Crosses

         DT                    PH                      EL                    ED                    KW                   GY

 

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

 

T1 x L1

48.5

   22

  14.6

13

14.2

    6

3.8       

 4

22.0    

  9

3.1       

  5

 

T1 x L2

48.5   

20

148.3   

14

14.2     

 7

3.5      

22

23.2    

  1

3.4         

  2

 

T1 x L3

50.0   

13

149.8   

 7

13.7    

18

3.7     

  9

21.7    

14

2.9       

12

 

T1 x L4

50.1   

12

145.0   

18

13.3    

22

3.7      

14

22.1    

  7

2.9       

11

 

T1 x L5

49.0    

19

145.6   

16

12.9    

25

3.5      

23

22.2    

  6

3.0       

10

 

T1 x L6

50.1   

11

152.3   

4

14.3    

  5

3.7      

11

21.8    

18

2.7       

21

 

T1 x L7

46.8   

27

138.9   

25

15.2      

  2

3.4      

26

20.3    

24

2.9       

16

 

T2 x L1

52.3   

  1

155.9   

 1

14.1    

  8

4.0      

  1

20.8    

22

3.3       

  3

 

T2 x L2

49.5   

17

149.2   

10

13.2    

21

3.7    

15

19.9    

27

2.4       

26

 

T2 x L3

51.2   

  4

145.2   

17

12.2    

27

3.7    

16

22.8    

  3

3.1       

  7

 

T2 x L4

50.2   

  9

141.0   

22

13.2    

24

3.7    

17

22.1    

  8

2.4       

15

 

T2 x L5

49.5     

18

140.8   

24

14.0    

10

3.7    

13

21.3    

17

2.0       

28

 

T2 x L6

50.0     

14

143.4   

19

14.6    

  4

3.3     

27

20.1    

25

3.1       

  8

 

T2 x L7

48.2     

21

149.1   

11

13.9    

14

3.4    

25

19.7    

28

3.5       

  1

 

T3 x L1

50.3     

  7

150.3   

 6

13.9    

12

3.6    

20

21.6    

16

2.8       

18

 

T3 x L2

49.7     

16

149.8   

 8

13.7    

16

3.7    

  7

21.7    

13

2.9       

13

 

T3 x L3

48.0     

23

139.2   

24

13.3    

20

3.8    

  2

22.4    

  5

2.7       

22

 

T3 x L4

50.2     

10

142.9   

21

11.9    

28

3.7    

12

20.6    

23

3.0       

  9

 

T3 x L5

51.2     

  3

151.4   

  5

16.1    

  1

3.6    

21

22.5    

  4

2.9       

17

 

T3 x L6

50.8     

  5

138.8   

26

13.9    

13

3.3    

28

20.9    

21

2.6       

24

 

T3 x L7

52.2     

  2

135.1   

28

14.1    

  9

3.5    

24

21.0    

20

2.2       

27

 

T4 x L1

50.3     

  8

146.1   

15

12.8    

26

3.7    

18

21.7    

12

2.5       

25

 

T4 x L2

50.0     

15

149.8   

 9

13.7    

17

3.7    

  8

21.7    

15

2.9       

14

 

T4 x L3

50.3     

  6

154.2   

 3

14.0    

11

3.8    

  5

23.2    

  2

3.1       

  6

 

T4 x L4

47.5     

25

148.9   

12

13.6    

19

3.7    

  6

21.8    

10

3.3       

  4

 

T4 x L5

45.7     

28

135.1   

27

13.8    

15

3.7    

10

21.8    

11

2.8       

19

 

T4 x L6

48.0     

24

154.2   

 2

13.2    

23

3.8    

  3

20.1    

26

2.7          

20

 

T4 x L7

47.2     

26

143.1   

20

15.2    

  3

3.6    

19

21.2    

19

2.6       

23

 

Mean

49

 

145.9

 

13.8

 

3.7

 

21.3

 

2.8

 

 

CV%

  6.7

 

10

 

13

 

9.8

 

14.5

 

 27.8     

 

 

S.E±

  0.64

 

    3.8

 

  0.46

 

0.08

 

  0.56

 

0.14

 

 
















 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm), ED= ear diameter (cm),  KW= kernels weight and GY= grain yield (t/ha).

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 

    The results indicate that crosses were later than their parents. Also, the taller crosses were late maturing than short ones. Generally, the crosses were taller than their parents which suggested some degree of hybrid vigor.

   Ear length trait is an important selection index for grain yield in maize. The ear length means of parents, as expected, were found to be shorter than those of the crosses at the two sites, with the general mean of 13.5 cm. The parents mean ranged between 12.7 cm for L7 to 15 cm for L2 (Table 2). The crosses mean varied from 11.9 cm for (T3 x L4) to 16.1 cm for (T3 x L5). However, long ear length were recorded for crosses T1 x L7 (15.2 cm), and T2 x L6 (14.6cm) (Table 3).Vedia and Claure (1995) found that ear length was the most important yield component and when used as a selection index genetic gain in recurrent selection reached 9.94% for yield and 5.75% for the ear traits. Therefore, any increase in ear length would be expected to increase number of kernels/row and hence increase grain yield.

    Ear diameter is a good indicator of the number of kernel rows/ear. The mean of ear diameter across sites for parents ranged between 3.4 cm for L6 and L7 to 4.1 cm for L4 (Table 2). Among the crosses, the large ear diameter ranged from 3.3 cm for T3 x L6 to 4.0 cm for T2 x L1. The crosses which had a big ear diameter were T3 x L3 and T4 x L6 (3.8cm) (Table 3). This result was in agreement with the findings of Tracy (1990) who found that the maize hybrids with high yield had more ears/plant, longer ears and a better ear shape and row configuration.

The mean of one hundred kernels weight for parents was 21.4 g, and it ranged between 19.9 g for L2 to 22.8 g for T3 (Table 2). Among the crosses, the mean was 21.3 g. The best crosses which obtained the highest kernel weight were T1 x L2 and T4 x L3 (23.2) followed by T2 x L3 (22.8 g) (Table 3).

Yield is a polygenic character is influenced by the fluctuating enviro-nment. Moreover, it is a complex trait depending on many components (Sharaan and Ghallab, 1997). In this study, there was a considerable amount of variability among the genotypes for this trait. The studied parents in the two locations showed a general mean of 2.4 t/ha. The parents means ranged between 2.12 t/ha for L4 to 2.93 t/ha for T4 (Table 2), while, the crosses means ranged between 2.0 t/ha for (T2 x L5) to 3.55 t/ha for (T2 x L7) (Table 3).  Most of the crosses (19 hybrids) had significantly higher mean grain yield than the overall mean. It is of interest to mention that the top ranking and the best yielder hybrids were T1 x L2 (3.4 t/ha), T2 x L1 (3.3 t/ha), T4 x L4 (3.3 t/ha), T1 x L1 (3.30 t/ha) and T4 x L3 (3.1 t/ha). These results agreed with those of Khalafalla and Abdalla (1997), who pointed to the fact that hybrids (crosses) produce higher grain yields than the open pollinated varieties due to the good performance of hybrids under Sudan conditions.

 Combining ability

    The breeding method to be adopted for improvement of a crop depends primarily on the nature of gene action involved in the expression of quantitative traits of economic importance. Combining ability leads to identification of parents with general combining ability effects and in locating cross combining showing high specific combining ability effects. In this study the ratio of GCA to SCA mean variance for most traits was less than one, suggesting that the inheritance of these traits was due to non additive gene action, with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and largely

controlled by additive gene action in the base material (Table 4).

Table 4. Mean squares of six agronomic traits for maize parents and 28 lines x tester crosses tested at two locations, Medani and Mutaq 2008.

Source of variation

DF

DT

PH

EL

ED

KW

GY

Location

  1

 3322.70**

13721**   

4.48**

50.90**

287.8**

 26.9**

Line

  6

     04.22

119.19

4.03

  0.02

    5.69

   0.27

Tester

  3

     18.81

   46.63

0.85

  0.07

    3.06

   0.28

Line x tester

18

     05.85**

   93.70*

2.38*

  0.05*

    1.62*

   0.44*

Line x tester x

location

18

     11.91**

217.90**

2.80

  0.11

     6.68

   0.63

Pooled error

76

     05.24

108.60

1.76

  0.04

     3.08

   0.19

GCA

 

       0.2

    -5.0

0.2

  0.00

    -0.7

   0.08

SCA

 

       0.6

    13.7

0.5

  0.02

     0.7

   0.03

GCA/SCA

 

       0.4

    -0.4

0.4

 -0.15

   -1.0

3.07

DT= days to 50% tasseling, PH= plant height, EL= ear length, ED= ear diameter, KW= kernels weight, GY= grain yield.

*, ** Significant at, 0.05 and 0.01 levels of probability, respectively.

 









 

   This result indicates that dominance and epistatic interaction effects seemed to be predomint for this trait and therefore heterosis breeding may be gratifying. The good combiner parents, those having negative GCA effects in Medani, for 50% days to tasseling were L5 followed by T4, T1 and L7, indicating earliness for flowering time, while, the latest, having positive GCA effect was T3 (Table 5).The earliest crosses having negative SCA effects were T3 x L6, T2 x L7 and T2 x L4, while, the latest crosses were T2 x L5, T4 x L5 and T4 x L4 (Table.6).

    The earliest parent in Mutaq was L7 (Table 5) and the earliest crosses were T2 x L4, T4 x L4 and T3 x L4 (Table 6). Common parents across locations that contributed to earliness were T4 and L5. The latest were L6 followed by T3 and T2 (Table 5). Parent L4 had good contribution for earliness to their hybrids progeny across locations.

Thus, the inbred lines which exhibited good general combining ability for at least one character can be used for development of early maturity and high grain yield. The contribution of the total variance for general and specific combining ability for this trait differs from location to another, but SCA was high in both locations (50.4% and 71.7%) compared with GCA which indicates that this trait is  controlled by additive gene action (Figs 1 and 2).

Trends in breeding work are to develop cultivars that are dwarf or of moderate height to avoid lodging of the crop which adversely affects yield. Only three top cross hybrid parents in Medani have   negative GCA effects for plant height, i.e., L7, L3 and T3; they were best combiners for short plant type. Tallness which is an undesirable trait is shown by parents L1, L2 and T1 (Table 5). Crosses having negative SCA effects and consequently short plant type were T4 x L2, T1 x L4 and T2 x L4,  while, tall hybrids with positive SCA effects were T3 x L1, T2 x L5 and T4 x L5 (Table 6).




 

   The best combiners for the short plant type with negative GCA in Mutaq were L7, L6 and L2 while, the taller parents with high positive GCA effects were L5 and L1 (Table 5). Among the crosses the shortest hybrids were T2 x L4, followed by T3 x L5 and the tallest hybrids were T2 x L5 and T3 x L4 (Table 6). This showed that, there is a relationship between late flowering and tall plant type. This is quite obvious among the hybrids such as T3 x L1 and T3 x L4.  Contribution for this trait is higher in crosses (80% and 53%) compared to parents (20% and 40%) at the two locations (Figs 1 and 2). The earliness and shortness are desirable traits especially under rainfed conditions for better water use efficiency and the escape of drought and avoidance of late season infestation with stem borer.

    Ear length is a good index for higher grain yield, therefore any increase in ear length would be expected to increase number of kernels/row and hence directly improve grain yield. In Medani site, the long ear length parents having a positively significant GCA effects such as L5, L7 and T1, while parents showing the short ear length were L4 and L2 (Table 5). The best crosses for this trait having a positive SCA effects and hence the longest ear length were T2 x L5 and T4 x L7. On the other hand the best combiners in Mutaq were L7 and T3 (Table 5), while the best crosses were T1 x L1 and T4 x L4 (Table 6). In the two locations, the best contribution was (73% and 65.9%) obtained by SCA compared with (27% and44.1) for GCA (Figs1 and 2). These results emphasized that ear length has a direct effect for improving grain yield. This is in agreement with the finding of Vedia and Claure (1995) who found that ear aspect was the most important yield component.

     Based on GCA estimates, the best combiners for ear diameter and length in Medani are L1 and L5, while best crosses were T1 x L2, T3 x L5 and T3 x L7. The good combiners in Mutaq site are L2, L3 and L4, while the best crosses are T3 x L4, T4 x L1 and T1 x L5 (Tables 5 and 6). A higher contribution among this trait is obtained by SCA (55.9% and 65.9%) in both locations compared with GCA (Figs 1and 2).

    Favorable GCA values were given by T1 and L3 as the good combiners for kernel weight in Medani and the best crosses were T4 x L7 and T2 x L7.  Among the studied parent material in Mutaq, only three parents have positive GCA effects (L3, L4 and T3).

 

DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

 Figure 1. Parent contribution of the maize GCA and SCA to the total variance

                of yield and its componets at Medani, season 2008.

 

 

 DT= days to 50% tasseling, PH= plant height (cm), EL= ear length (cm)

 ED= ear diameter (cm), KW= kernels weight (g), GY= grain yield (t/ha)

 

Figure 2. Parent contribution of the maize GCA and SCA to the total variance of yield and its componets at Mutaq, season 2008.

 

The best crosses were shown by T1 x L4 and T3 x L4 (Tables 5 and 6). The higher average contribution was given by the SCA (50.8% and 61) compared with the GCA at two locations (Figs 1and 2). This indicted that the inheritance of this trait was controlled by non additive gene effects.

    At Medani site, all the results depicted in Table 5 showed that the parents differ considerably with respect to estimate of GCA effects for grain yield. The parents having positive GCA effects were T1 followed by L4 and L6. Parents having negative GCA effects were L2 and L6. The best crosses having positive SCA effects were T3 x L3 followed by T4 x L5 and T1 x L2. Negative SCA effects were shown by T3 x L4, T2 x L2 and T1 x L4 (Table 6). The higher combiner in Mutaq, were L2, L1 and L4. The best crosses were T3 x L5, T1 x L5 and T4 x L5, while negative SCA effects were shown by T1 x L3, T2 x L5 and T4 x L3 (Tables 5 and6). The great contribution was given by SCA (62.4% and 62%) compared with GCA at the two locations (Figs1 and 2).

     General combing ability variance for grain yield is greater than the mean square for specific combining ability indicating the importance of additive gene action in controlling grain yield. This finding is in agreement with that of Barakat and Osman (2008) who found GCA effects are larger than SCA effects for grain yield indicating that the additive genetic variance is a major source of variations responsible for inheritance of grain yield.

 

CONCLUSION

   The ratio of general combining ability variance for grain yield was greater than specific combining ability indicating the importance of additive gene action in controlling this trait hence the good combiner parent for grain yield across locations was L4 so it could be used in recurrent selection. Also enormous variability was detected in the studied population which makes cyclic selection more effective. The best cross was T4 x L5 indicating that dominance and epesitic interaction seemed to be predomint, hence, higher heterosis gratified and recommended cross T4 x L5 for future testing in multi-locations trials for commercial utilization in order to be released as a hybrid.

REFERENCES

Alhussein, M.B. 2007. Growth Performance and Grain Yield Stability of   some Open Pollinated arieties of Maize (Zea mays L.). M.Sc. Thesis, University of Gezira, Wad Medani Sudan.

Barakat, A.A and M.M. Osman. 2008. Gene action and combining ability estimates for some romising maize inbred lines by top cross system. Journal of Agricultural Sciences. Mansoura niversity Journal 33:280-709

Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel  rossing system. Australian Journal of Biological Sciences 9: 463-493.

 Khalafalla, M.M. and H.A. Abdalla. 1997. Performance of some maize genotypes (Zea mays L.) nd  their  F1  hybrid  for  yield  and  its components. University of Khartoum Journal of Agricultural Sciences 5(2): 56-68.

Meseka, S.K. 2000.DiallelAnalysis for Combining Ability of Grain Yield and Yield Components n Maize (Zea  mays L.). M.Sc. Thesis, Faculty of Agricultural Sciences, University of ezira, Wad Medani, Sudan.

McCann, J. 2005. Maize and Grace: Africa’s Encounter with a New Crop, 1500-2000. Harvard niversity Press, New York

Nour, A.M., I. N. Elzain and M.A. Dafalla. 1997. Crop Development and   Improvement. Annual Report f the Maize Research Program. Agricultural Research Corporation, Wad Medani,Sudan.

Sharaan, A.N. and K.H. Ghallab. 1997. Character association at different location in sesame. Sesame and Safflower Newsletter 12: 66-75.

Tracy, W.F. 1990.  Potential of field corn germplasm for improvement of sweet corn. Crop Science 30:1041-1045.

Vedia, M.L. and E.T. Claure. 1995. Selection index for yield in the maize population. Crop Science 7: 505-510.

 

 

 

 

 

 

ABSTRACT

 

   The development of hybrids is the main objective of maize breeding. However, success depends largely on the identification of the best parents to ensure maximum combining ability. This study was conducted to estimate genetic variability and combining ability for grain yield and yield components of seven local inbred lines and four introduced open pollinated varieties of maize (Zea mays L.) across two irrigated locations, Medani and Matuq, Gezira, Sudan in 2008. The experiment was arranged in a randomized complete block design with three replicates. The traits measured were days to 50% tassel, plant height, ear length, ear diameter, hundred kernels weight and grain yield. Significant differences were observed among the parents and crosses for most of studied traits in both seasons. The crosses showed high genetic variability and tall plants than their parents which suggested some degree of hybrid vigor. The tallest hybrids across locations were T3 x L5 and T4 x L3. This indicates that the crosses were late maturing than their parents. The highest yielding hybrids had long ears and better shape, e.g., T2 x L1 and T1 x L7.The top five ranking crosses for grain yield across locations were T2 x L7 (3.45 t/ha), T1 x L2 (3.44 t/ha), T2 x LI (3.32 t/ha), T4 x L4 (3.30 t/ha) and T1 x L1 (3.13 t/ha).   The inheritance of most traits was controlled by non-additive gene action except ear height and grain yield. The best combiners for grain in Medani were T4, L4 and L5, while in Mutaq were L2, L4 and L6. The ratio of GCA to SCA variance for the most traits was less than one, suggesting that the inheritance was due to non additive gene effect with the exception of grain yield being more than one, indicating that inheritance of this trait was due to GCA effects, and was largely controlled by additive gene action in the base material. From these results it is recommended that parents T4, L1 and L6 to be used in recurrent selection, while, crosses T3 x L5, T1 x L5 and T4 x L6 to be tested in multi-locations trials for commercial utilization.

 

INTRODUCTION

       Maize generally is one of the most diverse crop both genetically and phenotypically. Due to its wide adaptability and productivity, maize spread rapidly around the world after the Europeans brought the crop from the Americas in the 15th and 16th centuries (McCann, 2005). The Portuguese introduced the crop to Africa at the beginning of the 16th century and since then the crop has replaced sorghum and millet as the main staple food in most of the continent where the climatic conditions are favorable (McCann, 2005). Today, there  is an increasing interest in maize production in Sudan due to its suitability to cultivation in the agricultural irrigated schemes, especially in the Gezira.It can occupy an important position in the economy of the country due to the possibility of blending it with wheat for making bread (Nour et al., 1997; Meseka, 2000).

    The grain yield of existing maize varieties and local landraces in Sudan is low. Also, maize   hybrids have been reported to show high potential for grain yield than the open pollinated varieties and landraces (Alhussein, 2007). Advantages of hybrids over open pollinated cultivars are higher yield, uniformity, high quality and resistance to diseases and pests. In spite of having yield potential, the production of maize in Sudan is very low. One of the reasons for this is the cultivation of exotic hybrids, which are not well adapted to our agro-climatic conditions. One of the strategies of the Agricultural Research Corporation (ARC) of the Sudan for maize breeding program is to develop new hybrids as an attempt to incorporate both advantages for higher yield and adaptability to environmental conditions. Thus, getting the benefit from the use of hybrids is the main purpose in maize breeding program of ARC.    Therefore, the objective of this study is to estimate the magnitude of combining ability in 28 topcross hybrids of maize for grain yield and its components across two irrigated locations and to identify high yielding topcross hybrids for future testing and commercial utilization.

 

 

MATERIALS AND METHODS

    The plant material used consisted of 7 local inbred lines used as lines (L), and 4 introduced open pollinated varieties used as testers (T) crossed in line x tester arrangement (Table 1). Hand pollination was used to develop the breeding material. Pollen grain was collected into a paper bag from the tassel of male parent (tester) and then dusted on the silk of the female parent (line). The ear was covered with a bag and information regarding the cross was written on the bag. A total of 28 cross combinations was obtained through hand pollination. In July 2008, the 11 parental material and 28 cross hybrids were grown and evaluated at two irrigated locations, Medani, Gezira Research Station (GRS) and Matuq, Matuq Research Station (MRS), Gezira State, Sudan. The trials were arranged a randomized complete block design with three replicates. The plot size was maintained as 2 rows x 3 m long with inter and intra row spacings of 80 and 25 cm, respectively.  Seeds were sown at the rate of 3- 4 seeds per hill.  Plants were thinned to one plant per hill after three weeks from sowing. Nitrogen was applied at 86 kg/ha in a split dose after thinning and before flowering. The crop was irrigated at intervals of 10-14 days, and plots were kept free of weeds by hand weeding.  Data were analyzed using the Statistical Analysis System (SAS) computer package. The analysis was done for each season for characters days to 50% tasseling, plant height, ear length, ear diameter, kernels weight and grain yield and then combined. Mean performance was separated using Duncan's Multiple Range Test (DMRT). Data from each location was analyzed separately and across locations to determine the general and specific combining ability of each line was measured according to Griffing,s Method 2 (1956).

 

Table 1. Pedigree of the lines and testers used in the study.

Parents

Pedigree

Source 

L1

RING-B-S1-2    

Inbred line developed by ARC

L2

PR-89 B-5655-S1-1

Inbred line introduced from CIMMYT, Mexico

L3

RING-B- S1-3   

Inbred line developed by ARC

L4

RING- B-S1-1

Inbred line developed by ARC

L5

RING-A-S1-1

Inbred line developed by ARC

L6

RING-A-S1-2

Inbred line developed by ARC

L7

PR-89 B-5655-S1-3

Inbred line introduced from CIMMYT, Mexico

T1

SOBSIY-HG AB                        

OPV introduced from CIMMYT, Kenya

T2

ACROSS- 500 HGY-B             

OPV introduced from CIMMYT, Kenya

T3

CORRALE10 -02 SIYQ           

OPV introduced from CIMMYT,  Kenya

T4

BAILO- 02SIYQ                        

OPV introduced from CIMMYT,  Kenya

RESULTS AND DISCUSSION

 

   The performance of the material tested for most traits is high across the two locations. However, significant differences among the parents and their hybrids for most traits were shown indicating the diversity of the material.

Mean separation and ranking

    Mean days to 50% tasseling indicates that the pollen shedding ability of maize genotypes is an indicator of the earliness of genotypes. Mean days to tasseling across locations for parents scored 52 days as the general mean. Mean of parents ranged between 49 and 55 days for L6 and T3, respectively (Table 2). The mean of crosses ranged between 46 days for (T4 x L5) to 52 days for (T2 x L1) (Table 3). Identification of early tasseling genotypes is very important in developing hybrids and choosing hybrids to suit different agro-ecological zones as well as grower’s requirements. Earliness was a desirable trait especially under rainfed conditions. It is important for better use of water resources and avoidance of late season infestation with stem borers. Hence, the earliest crosses were T1 x L7 (47 days), T4 x L7 (47 days), T4 x L4 (48 days) and T4 x L6 (48 days) (Table 3).

 

Table 2. Mean performance of eleven parents for the measured traits in maize at the two locations, season 2008.

Traits /

Parents

       DT   

      PH   

        EL    

       ED    

        KW  

      GY  

Mean   Rank

Mean  Rank

 Mean Rank 

Mean  Rank  

 

Mean Rank

 

Mean   Rank

L1

49.1      10

131.4     10

14.2         4

3.7          3

20.7         6

   2.8         2

L2

50.0        9

148.5       4

15.0         1

3.6          7

19.9       11

   2.6         5

L3

51.7        6

145.2       6

13.2         9

 

 

 

 

 

 

 

 

 

 

published in Gezira Journa; of Agricultural Science

© 2016 University Of Gezira. All rights reserved | Design by Informatics Administration