Previous PageTable Of ContentsNext Page

Genetic analysis of resistance to the cassava mosaic virus disease

Y. Lokko1, Alfred Dixon1, S.K. Offei1 and E.Y. Danquah2

1 International Institute of Tropical Agriculture, (IITA), Ibadan Nigeria, c/o Lambourn UK Ltd, Carolyn House, 26 Dingwall
Road Croydon CR9 3EE; UK web www.iita.org, emails y.lokko@cgiar.org, A.Dixon@cgiar.org 2 Dept of Crop Science,
University of Ghana, P.O. Box 25, Legon-Accra, Ghana.
Web www.ug.edu.gh, e-mail: offei@ug.edu.gh, edanquah@ug.edu.gh

Abstract

The relative importance of general combining ability (GCA), specific combining ability (SCA) and heterosis for resistance to the cassava mosaic virus, causal organism of the cassava mosaic disease (CMD), an important disease of cassava in Africa, was evaluated in a 3X18 North Carolina design II experiment involving different sources of resistance to the disease. Variations due to GCA of the males and the SCA of the female by male interaction were significant. The ratio of the total GCA components to the total GCA plus SCA component however, was 0.76 which indicates that GCA was more important in controlling CMD resistance among the crosses. The best general combiner was TMSI90257 and three crosses with significant SCA effects were more resistant than the average. The one-degree of freedom orthogonal contrast (parents versus crosses) test for average mid parent heterosis was not significant. However, one cross involving susceptible parents exhibited significant midparent heterosis for resistance to CMD.

Key Words

Additive gene effect, genotype by environment interaction (GXE)

Media Summary

New sources for the genetic control of resistance to the cassava mosaic disease (CMD have been determined.

Introduction

The cassava mosaic disease (CMD) is the most prevalent and serious disease of the cassava crop (Manihot esculenta) in sub-Saharan Africa. Resistance breeding, which started in the 1920’s, deployed resistance from Manihot glaziovii into cassava and the clone 58308, which was developed from this process, has been the main source of resistance in breeding for resistance to the disease (Hahn et al., 1989).

However, additional sources of resistance to CMD are required to ensure that durable resistance is maintained, since extensive use of closely related cultivars could result in vulnerability to pests and diseases. Hahn et al., (1977) noted that, some levels of resistance exist among the landraces. These resistant landraces could serve as new sources of resistance to broaden the genetic base of diversify for resistance to the disease within commercial cassava lines and breeding populations. However, selection of parents to be incorporated in a breeding programme cannot be based on their performance alone. Information regarding the relative magnitude and estimates of combining ability are essential, as this would allow the breeder to determine the appropriate breeding strategy to adopt. The objectives of this study were to determine the mode of gene action, combining ability, and to estimate heterosis for resistance to CMD in the various sources of resistance.

Materials and Methods

Experimental procedures

F1 crosses developed in a 3X18 North Carolina design II experiment involving six improved cassava accessions and 15 African landraces with varying levels of resistance and susceptibility to CMD together with the parents and three check genotypes (Table 1) were evaluated in three environments in Nigeria. The design in each environment was a randomised complete block with two replicates. Individual progeny which ranged from 52 to 934 genotypes per cross, ten parent and check plants were assessed for their reaction to CMD under natural infection by whiteflies at 12 weeks after planting (WAP), using the standard five point scoring scale system for CMD where a score of one indicates no obvious symptom and a score of five indicate severe mosaic symptoms.

Data analysis

Using the GLM procedure in SAS (SAS, 1999), the analyses of variance, was performed on weighted mean disease scores of the genotypes, with the environments considered random and genotypic effects were fixed. Parental accessions with mean disease score less than or equal to two were grouped as resistant while those with mean disease scores greater than two were susceptible.

Components of the genotypes were partitioned into checks and test genotypes (parents and crosses). The parents were further partitioned into female, male, resistant, susceptible and parental contracts. Crosses were also further partitioned into the variation due to the GCA effects of the females and males, SCA effect of their interaction (female x male) and a one degree of freedom orthogonal contrast (parent versus cross) was performed to test the significance of average mid-parent heterosis. The variation due to the genotype by environment interaction (GXE) effects of each of the main effects was also determined. Main effects were tested with their respective GXE effect and the GXE effects were tested with the pooled error.

The relative importance of GCA and SCA in predicting progeny performance was determined from the ratio of the mean square component associated with the fixed GCA effects of males θm and females θf to the sum of the mean square components of θm, θf and the SCA effect. Combining ability estimates were based on the methods described by Beil and Atkins, (1967) and tested for significance according to Cox and Frey, (1984). Per se performance of the parents with their GCA effects was tested with correlation. Average and individual mid and high parent heterosis were estimated and tested for significance as described by Dixon et al., (1990).

Results

The variation due to environments, replicates nested in environments and genotypes were significant except for the check by test genotype, the female, resistant and susceptible parent, the GCA effects of the females and the parent versus cross contract (Table 2). All the GXE effects except the check, female versus male parent and the SCA by environment interactions were significant. Significant variation among the crosses was due to GCA effect of males and the SCA effect. However, the ratio of the mean square component associated with GCA to the sum of mean square components associated with GCA and SCA, was 0.76. GCA was also significant (p<0.05) and positively correlated to parents means.

In this study, a negative GCA effect was desirable for resistance. Negative GCA effects of a parent indicated a larger contribution towards resistance while positive significant value suggests a contribution towards susceptibility. The overall best general combiner, which contributed the most to resistance was, TMS I90257 while worst general combiner for resistance to CMD was TME2 (Table 1). Negative SCA effects were also desirable for resistance. A cross with significant and negative SCA implied that this cross was more resistant than the average GCA of the parent accessions and a cross with significant and positive SCA implied that this cross was more susceptible than the average GCA of the parent accessions. The most resistant cross across environments, TMSI30572 X TMSI90257 had the lowest mean of 1.48 and largest negative and significant SCA effect of -0.44. Significant and negative SCA was also detected for TMSI30001 X TMSI60142 and also for TMS I30555 X TME12 across environments. The cross TMS I30555 X TME2 with a mean of 2.57, had a positive and significant SCA of 0.63, and was the most susceptible cross across environments. Significant and positive SCA effects were estimated in all the three crosses involving the susceptible male TME2 and for TMS I30001 X TME31, TMS I30555 X TME5, TMS I30572 X TME117, and TMS I30572 X TME31 across environments.

Despite the significance of the parent mean squares, which suggest diverse variability among the parents, the test for average heterosis (parent versus cross) was not significant indicating the absence of appreciable average mid-parent heterosis. Furthermore, mid and high parent heretosis values computed were generally positive indicating that the F1 crosses were generally more susceptible than their mid or high parents. However, the cross TMS I30555 X TME41 involving two susceptible parents, had significant (p<0.05) and negative mid parent heterosis of -31.94%, which suggests the presence of positive CMD resistance factors in the susceptible parents I30555 and TME117 resulting in enhanced resistance in their progeny.

Table 1 Description of cassava accessions used as parents and checks, Mean CMD severity scores and GCA effects for CMD symptom severity of parents

Parent

Accession

Pedigree information, local name and origin

Mean‡

GCA

Female 1

TMSI30001

Pedigree information lost

1.20

0.13

Female 2

TMSI30555

58308 X Oyarugba dudu

2.23

-0.04

Female 3

TMSI30572

58308 X Branca de Santa Caterina (OP)

1.55

-0.09

Male 1

TMSI4(2)1425

58308 X Oyarugba fufun

1.96

-0.06

Male 2

TMSI60142

KR685 OP

1.63

-0.20

Male 3

TMSI90257

58308 X Oyarugba dudu

1.48

-0.26*

Male 4

TME1

Antiota (Ondo, Ondo State, Nigeria)

1.48

-0.09

Male 5

TME2

Odungbo (Opeji, Ogun State, Nigeria)

2.68

0.55*

Male 6

TME4

Atu (Iwo, Kwara State, Nigeria)

1.18

-0.02

Male 7

TME5

Bagiwawa (New Busa, Niger State, Nigeria)

1.20

0.06

Male 8

TME6

Lapai-1 (Lapai, Niger State, Nigeria)

1.37

-0.07

Male 9

TME7

Oko-Iyawo (New Lapai, Niger State, Nigeria)

1.12

-0.03

Male 10

TME8

Amala (Ireuekpen, Edo State, Nigeria)

1.36

-0.12

Male 11

TME9

Olekanga (Ogbomosho, Oyo State, Nigeria)

1.15

-0.05

Male 12

TME10

Orente (Ogbomosho, Oyo State, Nigeria)

2.76

0.05

Male 13

TME11

Igueeba (Warri, Delta State, Nigeria)-

1.44

-0.09

Male 14

TME12

Tokunbo (Ibadan, Oyo State, Nigeria)

1.39

-0.09

Male 15

TME14

Abbey Ife (Abbey-Ife, Osun State, Nigeria)

1.37

0.06

Male 16

TME31

Bakince-Iri (Bahago, Sokoto State, Nigeria)

2.77

0.20

Male 17

TME41

Danbusa (Kanji, Niger State, Nigeria)

3.15

0.01

Male 18

TME117

Isunikankiyan (Ibadan, Oyo State, Nigeria)

3.46

0.17

 

91/02324

TME1 OP

   

* Significantly different from zero at the 0.05 probability level
SEMean 0.12, SEGCA 0.06 (LSD GCAmales 0.238; LSD GCAfemales 0.258)

Table 2 Analysis of variance for CMD severity at 12 weeks after planting (WAP) among all genotypes in 3X18 NCD II mating scheme, evaluated in three environments

Source of Variation

Across Environment

       

Df1

MS2

Environment (E)

   

2

59.17**

Replicates within E

   

3

1.40**

Genotypes in population (G)

   

78

1.39**

 

Checks (Ck)

   

3

7.72**

 

Test Genotypes (TG)

 

74

1.12**

 

Ck versus TG

   

1

1.32

   

Parent (P)

 

20

3.17**

     

Female (F)

2

1.64

     

Male (M)

17

3.51**

     

F versus M

1

0.46*

     

Susceptible (S)

5

1.08

     

Resistant (R)

14

0.29

     

R versus S

1

53.98**

   

Crosses (C)

 

53

0.31**

     

F (GCA)

2

1.40

     

M (GCA)

17

0.57*

     

F x M (SCA)

34

0.11*

   

P vs C

 

1

3.10

GXE

     

155

0.28**

 

Ck x E

   

5

0.07

 

TG x E

   

148

0.26**

 

Ck vs TG x E

   

2

2.71**

   

P x E

 

40

0.26**

     

F x E

4

0.26**

     

M x E

34

0.28**

     

F vs M x E

2

0.14

     

S x E

10

0.36**

     

R x E

28

0.36**

     

R vs S x E

2

1.01**

   

C x E

 

106

0.15**

     

F (GCA) x E

4

0.54*

     

M (GCA) x E

34

0.27*

     

F x M (SCA) x E

68

0.07

   

P vs CxE

 

2

6.20**

Error (genotypes)

   

233

0.07

Error (crosses)

   

159

0.06

* Significantly different from zero at the 0.5 probability level
** Significantly different from zero at the 0.01 probability level
1
df – degrees of freedom
2
MS – mean square

Conclusion

The study showed that GCA, or additive gene effect, is more important in determining progeny performance. Due to the importance of GXE on the genetic effects, significant SCA effects in crosses not involving parents with significant GCA, the parents identified as potential sources of resistance for breeding need to be selected after progeny testing to determine the most suitable parents.

Reference

Beil G.M and R.E. Aitkins, 1967. Estimates and specific combining abilities in F1 hybrids for grain yield and its components in grain sorghum, Sorghum vulgare Pers. Crop Science 7:225-228.

Cox, D.J. and K.L., Frey, 1984. Combining ability and selection of parents for interspecific oat matings. Crop Science 24:963-967.

Dixon, A.G.O, P.J. Bramel-Cox and T.L. Harvey, 1990. Diallele analysis of resistance in sorghum to greenbug biotype E: antibiosis and tolerance. Crop Science 30:1055-1059.

Hahn S.K., John, C. Isoba G. and Ikoun, T. (1989). Resistance breeding in root and tuber crops at the International Institute for Tropical Agriculture (IITA), Ibadan Nigeria. Crop Protection 8: 147-168.

Hahn, S.K., Howland A.K. and Terry, E.R. (1977). Cassava breeding at IITA. In: Leaky C.L.A. (ed.) Proceedings of the third Symposium of the International Society for tropical root crops. IITA, Ibadan Nigeria. 2-9 Dec. 1973.IITA Ibadan, Nigeria pp 4-10.

SAS Institute. 1999. SAS companion for Microsoft windows environment, version 6, 1st ed. SAS Institute, Cary, North Carolina, USA.

Previous PageTop Of PageNext Page