Euphytica

, 213:88 | Cite as

Diallel analysis of acid soil tolerant and susceptible maize inbred lines for grain yield under acid and non-acid soil conditions

  • Charles Mutimaamba
  • John MacRobert
  • Jill E. Cairns
  • Cosmos E. Magorokosho
  • Thokozile Ndhlela
  • Collis Mukungurutse
  • Adré Minnaar-Ontong
  • Maryke T. Labuschagne
Article
  • 215 Downloads

Abstract

Maize is not inherently tolerant to soil acidity but due to the ever increasing demand for the crop in the developing world, production of maize on acid soils continues to expand. Breeding for maize acid soil tolerance is the best strategy to improve yield under these conditions. Therefore, the current study was done to determine the general combining ability (GCA) of eight acid-soil tolerant and susceptible inbred lines and the specific combining ability (SCA) of cross combinations of these lines for grain yield under acid and non-acid soils. The eight lines were crossed using a diallel mating design to produce 28 single cross hybrids for evaluation under acid and non-acid soils at four sites for two seasons. Line C2 was the best general combiner under both soil environments while A2/C1 and A1/C2 had the highest desirable SCA effects under optimal conditions. Loss in grain yield and sensitivity to low pH stress was higher among genotypes in light textured soils than heavy soils. Non-additive gene action was more important than additive gene action in conditioning grain yield under both environments. Results revealed that it was feasible to improve grain yield under low pH and optimum soils from the set of genotypes used in the current study.

Keywords

Acid soil Combining ability Maize Yield 

Introduction

Maize is not inherently tolerant to soil acidity (Pandey and Gardner 1992) but due to the ever increasing demand for the crop as a source of food and livelihood in the developing world, production of maize on acid soils continues to expand. Acid soils are generally associated with aluminium (Al) toxicity and phosphorous (P) deficiency, which are usually cited as the leading biophysical constraints to high maize productivity in acid prone environments. P and nitrogen (N) deficiency are the second most important crop production constraint after drought on smallholder farms (Nziguheba et al. 2002; Kamara et al. 2003; Whitbread et al. 2004). Acidic soils generally have low pH, which causes heavy metal intoxication and deficiency of P and micronutrients (Jones et al. 2006; Poschenrieder et al. 2008; Balemi and Schenk 2009) and this leads to reduced crop yields through disruption of normal plant growth. As a result of the detrimental effects of Al toxicity and P deficiency, maize grown in acid soils is associated with fewer and smaller roots, stunted growth, increased barrenness, late flower initiation, and loss of biomass and grain yield due to loss of the plant’s capacity to take up water and nutrients from the soil (Duque-Vargas et al. 1994; Welcker et al. 2005; Navas et al. 2008).

Genetic variation and the associated quantitative inheritance for soil-acidity tolerance in maize has been extensively reported in various studies using a diversity of germplasm, various traits and different physiological, biochemical, genetic and more recently, DNA marker-based analyses (Pérez Velásquéz et al. 2008; Chen et al. 2009; George et al. 2011; Yang et al. 2013). This presents opportunities for restoring and enhancing crop productivity under such soils through plant breeding. Therefore breeding and subsequent deployment of varieties with a combination of tolerance to Al toxicity and efficiency of uptake and utilisation of P, coupled with use of soil amendments such as liming and use of P-bearing fertilisers, is a viable and sustainable strategy for enhancing crop productivity on acid-soils (Granados et al. 1993). Diallel cross studies conducted on maize grain yield in acid soils by various researchers (Salazar et al. 1997; Pérez Velásquez et al. 2008) revealed that both additive and non-additive genetic effects were important in accounting for the total genetic variance for grain yield, but additive effects were, however, more important than dominance and epistatic effects in the genetic control of the inheritance of grain yield in acid-soil conditions. Similar studies carried out by Pandey et al. (2007) on the inheritance of grain yield in acid soils revealed that both additive and dominance effects were predominant over epistatic effects in the genetic control of this trait. Pérez Velásquéz et al. (2008), on the other hand, indicated that dominance effects played a major role compared to additive effects which were, in turn, more preponderant than epistatic effects in both acidic and non-acidic soil environments.

Despite the importance of soil acidity in maize production in southern Africa, there has been very limited investment in breeding. The International Maize and Wheat Improvement Centre (CIMMYT) has an extensive abiotic stress maize phenotyping network of over 140 hectares across eastern and southern Africa (Masuka et al. 2016). Currently breeding is not conducted under low soil acidity (Magorokosho and Tarekegne 2015). Given the prevalence of problems associated with soil acidity within smallholder farmer maize systems there is a need to incorporate soil acidity into the goals of breeding programs within southern Africa. The objective of the current study was to determine the general combining ability (GCA) of acid-soil tolerant and susceptible maize inbred lines and the specific combining ability (SCA) of cross combinations of lines for grain yield under acid and non-acid soils for the purpose of identifying suitable combinations of inbred lines for use in hybrid constitution and inbred line development.

Materials and methods

Germplasm

Lines A1, A2, A3, and A4 were bred for tolerance to soil acidity and Al toxicity by CIMMYT Colombia while lines C1, C2, C3, and C4 had lack of acid soil tolerance but were developed for tolerance to drought, low N, common maize foliar diseases and ear rots by CIMMYT Zimbabwe. These inbred lines are early-medium maturing and have demonstrated appreciable tolerance to drought, heat, combined heat and drought, and good adaptation to non-stress conditions when screened under these environments in trials conducted in Mexico, Kenya, Zimbabwe, Thailand and India (Cairns et al. 2013). Lines A1, A4, C1, and C2 belong to heterotic group B, while A2, A3, C3, and C4 are in heterotic group A. The eight inbred lines were mated using the diallel mating design of Griffing (Method 4, model 1) to produce 28 single cross hybrids for field evaluation under acid and non-acid soils. Seeds from the F1 and reciprocal F1 crosses were composited, assuming absence of maternal effects on soil acidity tolerance in maize, to provide sufficient seed for multi-location evaluation.

The parents and hybrids were evaluated during the seasons 2011/2012 and 2013/2014 at four locations in Zimbabwe; Bindura, Domboshava, Makoholi, and Marondera (Table 1). Rainfall varied from 339 mm at Makoholi to 1100 mm at Marondera in the 2011/12 season, and 693 mm at Domboshava to 1040 mm at Marondera in 2013/14 (Table 1). Soil pH differences between limed and unlimed blocks varied from 0.6 to 1.2 points, where pH level ranged from 3.7 (extremely acidic) at Marondera to 5.6 (slightly acidic) at Bindura (Table 2). Available P ranged from 11 ppm at Domboshava to 65 ppm at Marondera. Soil nutrient levels were generally higher in heavy textured soils at Bindura than light sandy soils at Domboshava, Marondera and Makoholi.
Table 1

Description of testing sites and rainfall amounts received over two seasons (October–May)

Location

Latitude

Longitude

Altitude (masl)a

Total rainfall (mm)

2011/2012

2013/2014

Bindura

17°12′S

31°10′E

1136

781.0

1073.5

Domboshava

17°36′S

31°08′E

1538

846.6

693.0

Makoholi

19°50′S

30°46′E

1189

339.2

838.0

Marondera

18°10′S

31°29′E

1617

1098.9

1039.5

a masl metres above sea level

Table 2

Soil chemical and physical characteristics of the top 30 cm of the soil profile

Parameter

Site

2011/2012

2013/2014

Marondera

Domboshava

Bindura

Makoholi

LpH

Op

LpH

Op

LpH

Op

LpH

Op

Soil pH (CaCl2)

3.8

3.7

4.6

4.6

4.5

5.6

4.5

5.1

Available P2O5 (ppm)

65

48

22

22

42

36

33

19

Calcium (me 100 g−1)

0.8

0.7

0.7

0.7

12

13.2

0.7

0.6

Potassium (me 100 g−1)

0.1

0.1

0.1

0.1

0.5

0.5

0.1

0.1

Magnesium (me 100 g−1)

0.4

0.3

0.3

0.3

3.6

3.6

0.3

0.3

Soil colour

PB

PB

PB

PB

DB

DB

PB

PB

Soil texture

MgS

MgLS

MgS

MgS

MgSCL

MgSCL

MgS

MgS

LpH low pH, Op optimal pH, PB pale brown, DB dark brown, MgS medium grained sand, MgLS medium grained loamy sand, MgSCL medium grained sandy clay loam, ppm parts per million, me milliequivalents

Soil pH: <4.0 extremely acidic, 4.0–4.5 very strongly acidic, 4.5–5.0 strongly acidic, 5.0–5.5 medium acidic, 5.5–6.0 slightly acidic

Experimental design and data collection

The experimental field was sub-divided into two equal blocks at each site to facilitate separate evaluation of genotypes under two treatments which comprised of naturally acidic (low pH) or unlimed soil conditions and optimal or limed conditions. The 28 single cross hybrids were planted under each treatment at all sites. Thirty hybrids comprising of the 28 experimental single cross and two commercial check hybrids were planted in a 5 × 6 alpha-lattice design with three replications and a plot size of 4 m long, one row, 75 cm inter-row and 25 cm in-row spacing. Two seeds were hand-planted per station and thinned to 53,000 plants per hectare at three weeks after crop emergence. Various traits were measured or derived prior to and at flowering as well as at harvest but grain yield was the only trait analysed and reported in this study.

Statistical analyses

Estimates of GCA effects of the parents, SCA effects of the F1 crosses, and their mean squares were done using an 8 × 8 diallel analysis (Hallauer and Miranda 1988) without the checks, following Method 4, Model 1 (Griffing 1956; Dabholkar 1999). A fixed effects model was assumed for the parents and F1 crosses while environments were random in the current study. The underlying statistical linear model for the combined diallel analysis across environments was assumed in the combining ability analysis (Dabholkar 1999). Significance of GCA and SCA effects were tested using a t test. Standard errors of GCA and SCA effects were estimated as the square root of the GCA and SCA variances (Griffing 1956).

Results

Grain yield performance of diallel crosses under low pH and optimal soil environments

Analysis of variance for grain yield of the hybrids indicated that genotype effects were significant at all locations for both soil environments (Table 3). Loss in grain yield of 55.4% at Marondera, 20.4% at Bindura, 17.5% at Makoholi and 11.5% at Domboshava under low pH stress was observed. However, no significant change in grain yield due to low pH stress was observed across locations.
Table 3

Mean grain yield (t ha−1) under low pH and optimal soil conditions over two seasons

Hybrid

Across location

Environments

Low pH

Optimum

Domboshava

Marondera

Makoholi

Bindura

Low pH

Optimum

Low pH

Optimum

Low pH

Optimum

Low pH

Optimum

A1/A2

3.43

3.59

6.26

7.83

1.16

1.35

2.70

3.70

1.40

1.38

A1/A3

3.41

3.32

4.39

5.67

0.82

1.10

2.49

2.20

3.33

4.43

A1/A4

3.53

3.17

4.71

6.07

0.81

1.30

2.62

1.67

3.19

4.05

A1/C1

2.89

2.80

4.30

4.96

0.16

1.34

1.51

2.14

2.91

2.72

A1/C2

3.78

4.38

6.29

7.63

0.47

2.97

2.30

2.10

2.72

5.10

A1/C3

3.10

2.88

3.86

3.38

0.54

1.64

2.52

2.43

3.06

4.14

A1/C4

4.02

3.80

4.79

6.89

0.47

2.92

3.08

2.27

4.10

2.98

A2/A3

2.75

2.88

4.02

5.39

0.93

1.16

1.99

2.42

2.40

2.56

A2/A4

3.03

3.17

5.57

7.12

0.53

1.88

1.73

2.46

1.98

1.18

A2/C1

3.46

3.01

5.59

5.86

0.70

0.71

2.20

2.83

2.68

2.57

A2/C2

3.60

3.72

6.04

7.45

0.60

1.79

2.35

3.44

2.47

1.95

A2/C3

3.21

2.74

3.98

5.36

0.94

1.27

3.41

2.79

2.15

1.64

A2/C4

4.35

3.58

7.29

6.87

0.69

1.23

2.77

3.05

2.94

3.13

A3/A4

2.90

3.17

5.55

5.95

0.98

1.38

0.99

1.86

2.28

3.71

A3/C1

3.13

3.61

4.83

6.45

1.18

1.87

2.04

2.45

2.47

3.42

A3/C2

3.33

3.86

5.95

7.74

0.79

1.38

2.31

2.66

1.77

3.49

A3/C3

2.63

3.50

3.02

4.39

0.51

2.08

2.39

2.70

2.27

4.82

A3/C4

3.47

3.40

5.61

4.34

0.89

0.97

2.25

2.22

2.59

5.61

A4/C1

2.38

2.68

4.21

3.87

0.74

2.39

1.07

2.50

1.81

2.24

A4/C2

3.56

3.54

6.53

6.26

0.62

2.45

1.93

1.91

2.29

3.46

A4/C3

2.80

2.65

5.25

4.95

0.60

1.75

1.44

1.87

1.66

2.14

A4/C4

3.39

3.35

4.87

5.90

1.10

1.52

1.96

2.03

3.21

3.84

C1/C2

2.07

2.42

4.94

6.67

0.77

0.71

0.73

2.23

0.58

0.00

C1/C3

2.57

2.63

3.74

4.89

0.12

1.40

1.43

1.82

2.40

2.23

C1/C4

3.72

3.02

6.17

4.88

0.84

1.97

2.11

3.12

2.85

2.40

C2/C3

3.25

3.39

4.89

6.13

0.61

1.29

2.32

2.30

2.63

4.02

C2/C4

4.45

3.27

7.67

5.52

0.51

1.64

1.71

2.42

3.95

3.62

C3/C4

3.03

3.79

4.49

5.31

0.76

1.02

2.31

2.62

2.16

5.84

Mean

3.26

3.26

5.17

5.85

0.71

1.59

2.09

2.54

2.51

3.15

LSD

1.27

1.39

2.49

2.96

0.96

1.48

0.95

1.16

1.42

1.94

Genotype

**

NS

***

***

***

***

***

***

***

***

G x E

*

***

        

Heritability

0.34

0.01

0.27

0.26

0.00

0.09

0.73

0.41

0.62

0.76

LpH low pH, Op optimal pH, G × E Genotype × environment interaction

* P ≤ 0.05, *** P ≤ 0.001

Under low pH at Domboshava, C2/C4 and A2/C2 were the highest (>7 t ha−1) yielding hybrids, followed by C1/C4, A1/A2, A1/C2 and A4/C2 (>6 t ha−1). Under optimal conditions, A2/A4, A2/C2, A1/C2, A3/C2 and A1/A2 were the top (>7 t ha−1) yielding hybrids. The highest yielding crosses at Marondera under low pH were A3/C1, A1/A2 and A4/C4 (>1 t ha−1). The crosses A1/C2 and A1/C4 were the top (≈3 t ha−1) performing hybrids under optimal conditions at the same site. At Makoholi under low pH, A2/C3 and A1/C4 were the top (>3 t ha−1) yielding crosses. The highest (>3 t ha−1) yielding crosses under optimal conditions at this site were A1/A2, A2/C2 and C1/C4. The highest (≈4 t ha−1) yielding hybrids under low pH at Bindura were A1/C4 and C2/C4. C3/C4, A3/C4 and A1/C2 were the top (≈5 t ha−1) yielding hybrids under optimal conditions at this site.

The hybrids comprised of six tolerant × tolerant (T × T) crosses (Group I), 16 tolerant × susceptible (T × S) crosses (Group II) and six susceptible × susceptible (S × S) crosses (Group III) whose mean grain yields are presented for the respective soil environments in Table 4. No significant differences were observed between each group of crosses for each of the two environments except for Group III (S × S) under low pH. In Group III under low pH, the highest yielding crosses were C2/C4 and C1/C4 while C1/C2 was the lowest yielding cross.
Table 4

Mean yield (t ha−1) of grouped diallel crosses under low pH and optimal soil conditions across eight environments

Cross

Low pH

Optimal

Group I: Tolerant × Tolerant crosses

 A1/A2

3.43

3.59

 A1/A3

3.41

3.32

 A1/A4

3.53

3.17

 A2/A3

2.75

2.88

 A2/A4

3.03

3.17

 A3/A4

2.90

3.17

 Mean

3.17

3.22

Group II: Tolerant × Susceptible crosses

 A1/C1

2.89

2.80

 A1/C2

3.78

4.38

 A1/C3

3.10

2.88

 A1/C4

4.02

3.80

 A2/C1

3.46

3.01

 A2/C2

3.60

3.72

 A2/C3

3.21

2.74

 A2/C4

4.35

3.58

 A3/C1

3.13

3.61

 A3/C2

3.33

3.86

 A3/C3

2.63

3.50

 A3/C4

3.47

3.40

 A4/C1

2.38

2.68

 A4/C2

3.56

3.54

 A4/C3

2.80

2.65

 A4/C4

3.39

3.35

 Mean

3.32

3.34

Group III: Susceptible × Susceptible crosses

 C1/C2

2.07

2.42

 C1/C3

2.57

2.63

 C1/C4

3.72

3.02

 C2/C3

3.25

3.39

 C2/C4

4.45

3.27

 C3/C4

3.03

3.79

 Mean

3.18

3.09

 Overall mean

3.26

3.26

 LSD

1.39

1.42

Analysis of variance of grain yield of diallel crosses under low pH and optimal soil environments

Mean squares were highly significant (P ≤ 0.01) for genotype × environment interaction (G × E), general combining ability (GCA) × environment interaction, genotype, environment and replication across environments (Table 5). About 34.2% of total variation in grain yield was accounted for by environment, 16.8% by G × E, 10.7% by GCA × environment interactions and 0.5% by genotype across environments.
Table 5

Analysis of variance for line GCA effects and SCA effects for grain yield under acid and optimal conditions across eight environments

Source of variation

DF

SS

MS

Environment (Env)

7

682.58

97.51***

Replication (Site)

16

53.94

3.37***

Genotype (Gen)

27

95.70

3.54**

GCA

7

71.11

10.16

SCA

20

38.36

1.92

Env × Gen

189

336.18

1.78***

Env × GCA

49

214.40

4.38***

Env × SCA

140

178.32

1.27

Residual

312

325.00

1.04

Total

767

1995.60

 

GCA general combining ability, SCA specific combining ability, DF degrees of freedom, SS sums of squares, MS mean squares, *** P ≤ 0.001; ** P ≤ 0.01

General combining ability effects of parental inbred lines for grain yield under stress and non-stress soil environments

GCA effects were non-significant for all parental inbred lines across environments and at Marondera for both environments, while significant GCA effects were observed for the rest of the individual locations for both environments (Table 6). Prediction ratio (GCA:SCA) was 0.56 across locations and ranged from 0 to 0.98 at individual locations for both environments, except Makoholi under low pH where it was 1.58. Significant (P < 0.05) positive GCA effects were observed under both environments for lines C2 at Domboshava, A2 at Makoholi and C4 at Bindura, while significant (P < 0.05) negative GCA effects were observed under both environments for A4 at Makoholi and C3 at Domboshava. Under low pH, the highest significant positive GCA effects were observed for line C2 at Domboshava, followed by C4 at Bindura, while lines A1 and A2 at Makoholi gave the lowest positive GCA effects. The highest significant negative GCA effects under low pH were obtained from line C3 at Marondera, while A4 at Makoholi produced the lowest GCA effects under this environment. Under optimal conditions, the highest significant positive GCA effects were detected for lines C2 at Domboshava and A3 at Bindura, while A2 produced the lowest GCA effects at Makoholi. The highest significant negative GCA effects in this environment were observed for lines A2 and C1 at Bindura, while the lowest negative GCA effects were detected for A4 at Makoholi. Line A1 had significant positive GCA effects under low pH at both Makoholi and Bindura, while significant positive GCA effects were observed for A2 under optimal conditions at Domboshava and in both environments at Makoholi.
Table 6

General combining ability effects of inbred lines for grain yield under low pH and optimal soil conditions across eight environments

Line

Across

Domboshava

Marondera

Makoholi

Bindura

LpH

Op

LpH

Op

LpH

Op

LpH

Op

A1

0.20

−0.26

0.25

−0.09

0.25

0.42**

0.07

0.53*

0.46

A2

0.07

0.42

0.82*

0.10

−0.29

0.41**

0.63**

−0.26

−1.28**

A3

0.04

−0.47

−0.17

0.19

−0.20

−0.03

0.02

−0.07

1.00**

A4

−0.16

0.08

−0.13

0.07

0.26

−0.49**

−0.57**

−0.19

−0.24

C1

−0.42

−0.40

−0.56

−0.08

−0.12

−0.60**

−0.23

−0.31

−1.15**

C2

0.21

1.02*

1.08**

−0.10

0.18

−0.17

−0.05

−0.19

−0.14

C3

−0.28

−1.16*

−1.09**

−0.15

−0.11

0.19

−0.04

−0.21

0.46

C4

0.34

0.78

−0.21

0.05

0.03

0.25

0.17

0.71**

0.89**

VE

1.04

2.83

2.00

0.33

0.69

0.30

0.41

0.64

1.16

SE

0.28

0.45

0.38

0.15

0.22

0.15

0.17

0.22

0.29

GCA:SCA

0.56

0.00

0.00

0.00

0.05

1.58

0.88

0.59

0.98

V E error variance, SE standard error, GCA:SCA prediction ratio, LpH low pH, Op optimal pH

* P ≤ 0.05, ** P ≤ 0.01

Specific combining ability effects of diallel crosses for grain yield under low pH and optimal soil environments

SCA effects were non-significant across environments, Domboshava and Marondera for both environments, while significant SCA effects were detected for both environments at Makoholi and Bindura (Table 7). The highest significant (P < 0.05) positive SCA effects were observed for the crosses A2/C1 and A1/C2 under optimal conditions at Bindura while A2/C3 was the only cross with significant (P < 0.05) positive SCA effects under low pH in this study. The highest significant (P < 0.05) negative SCA effects were detected for the crosses C1/C2 and A1/C4 under optimal conditions at Bindura, while the lowest negative SCA effects were produced by A2/A3 under optimal conditions at Makoholi.
Table 7

Specific combining ability effects of crosses for grain yield under low pH and optimal soil conditions across eight environments

Line

Across

Domboshava

Marondera

Makoholi

Bindura

LpH

Op

LpH

Op

LpH

Op

LpH

Op

A1/A2

0.07

0.93

0.91

0.45

−0.20

−0.23

0.97*

−1.37**

−0.95

A1/A3

−0.04

−0.04

−0.26

0.01

−0.54

0.01

0.47

0.37

−0.17

A1/A4

0.09

−0.28

0.11

0.12

−0.80

0.59

−0.21

0.34

0.68

A1/C1

−0.28

−0.20

−0.58

−0.39

−0.38

−0.42

−0.71

0.19

0.27

A1/C2

0.39

0.36

0.45

−0.05

0.95

−0.05

−0.15

−0.13

1.63*

A1/C3

−0.20

0.12

−1.63

0.06

−0.09

−0.19

0.01

0.24

0.07

A1/C4

−0.02

−0.90

0.99

−0.20

1.05

0.30

−0.37

0.36

−1.53*

A2/A3

−0.49

−1.10

−1.11

−0.07

0.06

−0.49

−1.02**

0.22

−0.31

A2/A4

−0.05

−0.11

0.58

−0.35

0.32

−0.29

−0.01

−0.08

−0.46

A2/C1

0.31

0.40

−0.25

−0.03

−0.47

0.29

−0.01

0.74

1.84**

A2/C2

−0.03

−0.58

−0.30

−0.11

0.30

0.01

−0.24

0.41

0.21

A2/C3

0.02

−0.45

−0.23

0.28

0.08

0.71*

0.37

0.10

−0.69

A2/C4

0.16

0.91

0.40

−0.17

−0.10

0.01

−0.05

−0.02

0.36

A3/A4

−0.02

0.77

0.40

0.01

−0.27

−0.58

−0.26

0.04

−0.20

A3/C1

0.58

0.53

1.33

0.36

0.60

0.58

0.44

0.35

0.42

A3/C2

0.03

0.24

0.98

−0.01

−0.20

0.42

−0.19

−0.47

−0.52

A3/C3

0.05

−0.52

−0.20

−0.24

0.80

0.13

0.06

0.04

0.21

A3/C4

−0.11

0.13

−1.14

−0.06

−0.45

−0.06

0.49

−0.55

0.56

A4/C1

−0.11

−0.64

−1.28

0.04

0.67

0.06

0.09

−0.20

0.48

A4/C2

0.27

0.26

−0.53

−0.06

0.42

0.49

0.68

0.16

0.69

A4/C3

−0.05

1.16

0.33

−0.03

0.01

−0.36

0.05

−0.45

−1.23

A4/C4

−0.12

−1.16

0.39

0.28

−0.35

0.10

−0.34

0.19

0.04

C1/C2

−0.67

−0.85

0.30

0.24

−0.94

−0.60

0.26

−1.43**

−2.29**

C1/C3

0.01

0.13

0.69

−0.37

0.05

−0.26

−0.49

0.41

−0.23

C1/C4

0.15

0.62

−0.20

0.16

0.48

0.36

0.42

−0.06

−0.49

C2/C3

0.12

−0.13

0.29

0.15

−0.37

0.20

−0.10

0.52

0.55

C2/C4

−0.11

0.70

−1.20

−0.15

−0.16

−0.47

−0.26

0.93

−0.28

C3/C4

0.05

−0.30

0.75

0.15

−0.48

−0.23

0.11

−0.85

1.33*

VE

1.04

2.83

2.00

0.33

0.69

0.30

0.41

0.64

1.16

SE

0.61

1.00

0.85

0.34

0.50

0.33

0.38

0.48

0.64

GCA:SCA

0.56

0.00

0.00

0.00

0.05

1.58

0.88

0.59

0.98

V E error variance, SE standard error, GCA:SCA prediction ratio, LpH low pH, Op optimal pH

* P ≤ 0.05, ** P ≤ 0.01

Discussion

Analysis of variance showed significant differences in yield response of the diallel crosses under the different soil environments. Loss in grain yield among crosses due to low pH stress was higher (up to 55%) in light textured soils than heavy soils, which implies that genotypes were more sensitive to low pH stress in light soils than heavy soils. The large yield loss at Marondera was unexpected as the pH was similar under low pH and optimal conditions. The low pH under the optimal conditions may have been due to the fact that more ammonium nitrate was applied under optimum conditions at Marondera than under low N conditions. Ammonium-based fertilizers have an acidifying effect when added to the soil as soil bacteria convert the ammonium to nitrate (nitrification process) releasing hydrogen ions (H+) and thus the free hydrogen ions released increase acidity in the soil. The decrease in yield under the low pH conditions may have been a result of the lower ammonium nitrate application under these conditions.

Heritability was low (<0.40) across low pH sites and at Domboshava for both environments, while at Makoholi and Bindura it was moderate (0.40–0.60) to high (>0.70) for both environments. However, it was extremely low (<0.10) across optimal soil conditions and at Marondera for both environments. The crosses C2/C4, A2/C4 and A1/C4 produced the highest grain yield (>4 t ha−1) across low pH environments, while C1/C2 and A4/C1 were the lowest yielding crosses. A1/C2 was the top yielding hybrid across optimal conditions with 4.3 t ha−1, followed by A3/C2, A1/C4 and C1/C3, while the lowest yielding hybrid was C1/C2 with less than 2.5 t ha−1.Heritability of grain yield was low to moderate across low pH environments and extremely low across optimal environments, which indicates that genotype accounted for more variation in grain yield under stress than optimal soil conditions. Therefore, mean squares for genotype, and GCA and SCA were estimated with higher precision under stress than optimal soil environments.

Results of the current study indicated that hybrids comprising of tolerant × susceptible (T × S) inbred lines showed better adaptation to soil stress environments compared to crosses which consisted of tolerant × tolerant (T × T) and susceptible × susceptible (S × S) inbred lines. These results are contrary to findings of Menkir et al. (2010) and Badu-Apraku et al. (2011) who reported higher grain yield among T × T crosses. However, the results of this study appear consistent with findings of Betrán et al. (2003), Derera et al. (2007) and Makumbi et al. (2011) who reported higher yields from crosses comprising of stress tolerant and non-tolerant inbred lines. The large variation in the mean yields of each group observed in this study is in agreement with findings of Menkir et al. (2010) who reported large differences in grain yield under stress and non-stress environments.

The effect of GCA on grain yield, though non-significant across soil environments, was significant for both environments and at three of the locations. Therefore, this indicates that there was sufficient yield variability among the parental inbred lines involved in hybrid combinations to facilitate the improvement of grain yield through selection under the two types of environments. The effect of SCA on grain yield was non-significant in both types of soil environments across locations and at two of the sites. This indicates that there were differences in the yield performance of the diallel crosses at locations with significant SCA effects and therefore an opportunity to identify hybrids with superior yield performance under the different types of soil environments. These results are consistent with findings reported by other researchers (Makumbi et al. 2011; Badu-Apraku et al. 2011, 2013) who found significant GCA and SCA mean squares for grain yield under stress and non-stress environments. This indicated that both additive and non-additive gene action were important in the inheritance of grain yield under both types of soil environments.

Prediction ratio indicates that SCA variance was higher than GCA variance across locations at three of the individual locations for both environments, which corroborates the findings by Badu-Apraku et al. (2013) who reported a preponderance of SCA variance over GCA variances for grain yield. However, these findings are contrary to results found by Makumbi et al. (2011), and Badu-Apraku et al. (2011) who reported that GCA mean squares were larger than SCA mean squares under stress and non-stress environments. The preponderance of SCA mean squares over GCA mean squares in this study implies that non-additive gene action was more important than additive gene action for grain yield and that SCA was the major component accounting for the differences among the 28 diallel crosses. Therefore, this study suggests that the strategies of an efficient maize breeding program, whose primary objective is breeding for low pH tolerance, should be based on exploitation of dominance effects. Results obtained in this study also agree with findings of Narro et al. (2000) who concluded in their diallel analysis in acid soils that both additive and non-additive gene action were important for grain yield.

Significant GCA effects for grain yield were detected among inbred lines under both types of soil environments which is consistent with findings reported by Makumbi et al. (2011) and Badu-Apraku et al. (2011, 2013). Lines C2 and C4 produced the highest desirable GCA effects under low pH and were therefore the best general combiners under this environment, which implies that they are ideal sources of tolerance in breeding maize for low pH tolerance. Line C2 also produced the highest desirable GCA effects under optimal conditions, which suggests that it is suitable as source material in inbred line development which is targeted at development of hybrids with high yield performance under low pH and optimal conditions. Lines A1 and A2 were good general combiners since they both had desirable GCA effects at two locations, with the former under low pH at Makoholi and Bindura and the latter under optimal conditions at Domboshava and both environments at Makoholi. This suggests that these lines are ideal for use in the introgression of desirable genes for adaptability to the soil environments where they produced desirable GCA effects. A2/C3 was the only cross combination with desirable SCA effects under low pH in the current study, while A2/C1 and A1/C2 had the highest significant SCA effects in the desired direction under optimal conditions.

SCA effects of A2/C3 under low pH, SCA of A1/A2 under optimal conditions and GCA effects of the common parent (A2) under both environments at Makoholi were in the desired direction, while SCA of C3/C4 and GCA of one of its parents (C4) were desirable under optimal conditions at Bindura, which indicates that each of these good combiners might have contributed to the desirable performance of their respective progenies. SCA effects of A1/A2 under low pH and SCA of A1/C4 under optimal conditions at Bindura were in the undesirable direction, while one of their respective parents was a good general combiner under the separate environments. In addition, A2/A3 was characterised by desirable SCA effects and good GCA effects of one of its parents (A2) under optimal conditions at Makoholi. These findings imply that good general combiners do not always necessarily confer good cross performance in their progenies. C1/C2 was associated with undesirable SCA effects under both environments and poor GCA effects of one of its parents (C1) under optimal conditions at Bindura, which indicates that this poor general combiner might have contributed to the poor performance of its progeny under optimal conditions. A2/C1 was characterised by desirable SCA effects under optimal conditions at Bindura despite both of its parents being poor general combiners at this location and environment. This suggests that poor × poor general combiners do not always necessarily produce the poorest performing crosses. Results of the current study are consistent with the findings of Chipeta et al. (2015) who concluded that progeny performance is not largely dependent on parental genotype performance per se.

To conclude, loss in grain yield and sensitivity to low pH stress was higher among genotypes in light textured soils than heavy soils. Crosses comprising of tolerant × susceptible inbred lines showed better adaptation to low pH stress than crosses which consisted of tolerant × tolerant and susceptible × susceptible inbred lines. Non-additive gene action was more important than additive gene action in conditioning grain yield under low pH stress and optimal soil conditions. Results showed that it was feasible to improve grain yield under low pH and optimum soils from the set of genotypes used in the current study.

Notes

Acknowledgements

This work was funded by the Bill and Melinda Gates Foundation project Drought Tolerant Maize for Africa (DTMA) and the CGIAR Research Program MAIZE for which the authors are sincerely grateful. Due acknowledgement is gratefully extended to staff of DR&SS, CIMMYT Southern Africa Regional Office, Domboshava Training Centre and Danmaglass Farm.

References

  1. Badu-Apraku B, Oyekunle M, Akinwale RO, Fontem Lum A (2011) Combining ability of early-maturing white maize inbreds under stress and nonstress environments. Agron J 103:544–557CrossRefGoogle Scholar
  2. Badu-Apraku B, Oyekunle M, Akinwale RO, Aderounmu M (2013) Combining ability and genetic diversity of extra-early white maize inbreds under stress and nonstress environments. Crop Sci 53:9–26CrossRefGoogle Scholar
  3. Balemi T, Schenk MK (2009) Genotypic variation of potato for phosphorus efficiency and quantification of phosphorus uptake with respect to root characteristics. J Plant Nutr Soil Sci 172:669–677CrossRefGoogle Scholar
  4. Betrán FJ, Ribaut JM, Beck D, Gonzalez Deleón D (2003) Genetic diversity, specific combining ability, and heterosis in tropical maize under stress and nonstress environments. Crop Sci 43:797–806CrossRefGoogle Scholar
  5. Cairns JE, Crossa J, Zaidi PH, Grudloyma P, Sanchez C, Araus JL, Thaitad S, Makumbi D, Magorokosho C, Bänziger M, Menkir A, Hearne S, Gary NA (2013) Identification of drought, heat, and combined drought and heat tolerant donors in maize. Crop Sci 53:1335–1346CrossRefGoogle Scholar
  6. Chen J, Xu L, Cai Y, Xu J (2009) Identification of QTLs for phosphorus utilization efficiency in maize (Zea mays L.) across P levels. Euph 167:245–252CrossRefGoogle Scholar
  7. Chipeta MM, Bokosi JM, Saka VW (2015) Genetic analysis in a six parent diallel cross of cassava (Manihot esculenta Crantz). Afr J Agric Res 10:1076–1082CrossRefGoogle Scholar
  8. Dabholkar AR (1999) Elements of biometrical genetics. Concept Publishing Company, New DelhiGoogle Scholar
  9. Derera J, Tongoona P, Vivek BS, Laing MD (2007) Gene action controlling grain yield and secondary traits in southern African maize hybrids under drought and non-drought environments. Euph 162:411–422CrossRefGoogle Scholar
  10. Duque-Vargas J, Pandey S, Granados G, Ceballos H, Knapp ER (1994) Inheritance of tolerance to soil acidity in tropical maize. Crop Sci 34:50–54CrossRefGoogle Scholar
  11. George MLC, Salazar F, Warburton M, Narro L, Vallejo FA (2011) Genetic distance and hybrid value in tropical maize under P stress and non-stress conditions in acid soils. Euph 178:99–109CrossRefGoogle Scholar
  12. Granados G, Pandey S, Ceballos H (1993) Response to selection for tolerance to acid soils in a tropical maize population. Crop Sci 33:936–940CrossRefGoogle Scholar
  13. Griffing B (1956) Concept of general and specific combining ability in relation to diallel crossing systems. Aust J Biol Sci 9:463–493CrossRefGoogle Scholar
  14. Hallauer AR, Miranda JB (1988) Quantitative genetics in maize breeding. Iowa State Univ. Press, AmesGoogle Scholar
  15. Jones DL, Blancaflor EB, Kochian LV, Gilroy S (2006) Spatial coordination of aluminium uptake, production of reactive oxygen species, callose production and wall rigidification in maize roots. Plant, Cell Environ 29:1309–1318CrossRefGoogle Scholar
  16. Kamara AY, Menkir A, Badu-Apraku B, Ibikunle O (2003) The influence of drought stress on growth, yield and yield components of selected maize genotypes. J Agric Sci 141:43–50CrossRefGoogle Scholar
  17. Magorokosho C, Tarekegne A (2015) Preliminary results of the 2015 regional trials for eastern and southern Africa. CIMMYT, HarareGoogle Scholar
  18. Makumbi D, Betrán JF, Bänziger M, Ribaut JM (2011) Combining ability, heterosis and genetic diversity in tropical maize (Zea mays L.) under stress and non-stress conditions. Euph 180:143–162CrossRefGoogle Scholar
  19. Masuka B, Atlin GN, Olsen M, Magorokosho C, Labuschagne M, Crossa J, Bañziger M, Pixley KV, Vivek BS, von Biljon A, Marobert J, Alvarado G, Prasanna BM, Makumbi, Tarekegne A, Das B, Zaman-Allah M, Cairns JE (2017) Gains in maize genetic improvement in eastern and southern Africa i) CIMMYT hybrid breeding pipeline. Crop Science 75:168–179Google Scholar
  20. Menkir A, Adetimirin VO, Yallou CG, Gedil M (2010) Relationship of genetic diversity of inbred lines with different reactions to Striga hermonthica (Del.) Benth and the performance of their crosses. Crop Sci 50:602–611CrossRefGoogle Scholar
  21. Narro L, Pérez JC, Pandey S, Crossa J, Salazar F, Arias M, Franco J (2000) Diallel and triallel analysis in an acid soil tolerant maize population. Maydica 45:301–308Google Scholar
  22. Navas AA, Hallauer AR, Pandey S (2008) Molecular studies for determination of quantitative trait loci for acid soil tolerance in maize. Maize Genet Coop Newsl 82:3–4Google Scholar
  23. Nziguheba G, Merckx R, Palm CA (2002) Soil phosphorus dynamics and maize response to different rates of phosphorus fertilizer applied to an Acrisol in western Kenya. Plant Soil 243:1–10CrossRefGoogle Scholar
  24. Pandey S, Gardner CO (1992) Recurrent selection for population, variety and hybrid improvement in tropical maize. Adv Agron 48:1–87CrossRefGoogle Scholar
  25. Pandey S, Narro L, Friesen DK, Waddington SR (2007) Breeding maize for tolerance to soil acidity. Plant Breed Rev 28:59–100Google Scholar
  26. Pérez Velásquez JC, de Souza CL, Narro LA, Pandey S, de León CL (2008) Genetic effects for maize traits in acid and non-acid soils. Genet Mol Biol 97:89–97CrossRefGoogle Scholar
  27. Poschenrieder C, Gunse B, Corralles I, Barcelo J (2008) A glance into aluminum toxicity and resistance in plants. Sci Total Environ 400:237–260CrossRefGoogle Scholar
  28. Salazar F, Pandey S, Narro L, Pérez JC, Ceballos H, Parentoni S, Bahia Filho AFC (1997) Diallel analysis of acid soil tolerant and intolerant tropical maize populations. Crop Sci 37:1457–1462CrossRefGoogle Scholar
  29. Welcker C, The C, Andreau B, De Leon C, Parentoni SN, Bernal J, Félicité J, Zonkeng C, Salazar F, Narro L, Charcosset A, Horst WJ (2005) Heterosis and combining ability for maize adaptation to tropical soils: implications for future breeding strategies. Crop Sci 45:2405–2413CrossRefGoogle Scholar
  30. Whitbread AM, Jiri O, Maasdorp B (2004) The effect of managing improved fallows of Mucuna pruriens on maize production and soil carbon and nitrogen dynamics in sub-humid Zimbabwe. Nutr Cycl Agro-Ecosyst 69:59–71CrossRefGoogle Scholar
  31. Yang ZB, Rao IM, Horst WJ (2013) Interaction of aluminium and drought stress on root growth and crop yield on acid soils. Plant Soil 372:3–25CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2017

Authors and Affiliations

  • Charles Mutimaamba
    • 1
    • 3
  • John MacRobert
    • 2
  • Jill E. Cairns
    • 2
  • Cosmos E. Magorokosho
    • 2
  • Thokozile Ndhlela
    • 3
  • Collis Mukungurutse
    • 4
  • Adré Minnaar-Ontong
    • 1
  • Maryke T. Labuschagne
    • 1
  1. 1.Department of Plant Sciences (Plant Breeding)University of the Free StateBloemfonteinSouth Africa
  2. 2.CIMMYTHarareZimbabwe
  3. 3.Crop Breeding InstituteHarareZimbabwe
  4. 4.Chemistry and Soils Research InstituteHarareZimbabwe

Personalised recommendations