search
for
 About Bioline  All Journals  Testimonials  Membership  News


African Crop Science Journal
African Crop Science Society
ISSN: 1021-9730 EISSN: 2072-6589
Vol. 8, Num. 3, 2000, pp. 203-211
African Crop Science Journal, Vol. 8. No. 3, pp. 203-211

African Crop Science Journal, Vol. 8. No. 3, pp. 203-211

Inheritance of resistance to pod shattering in soybean

P.Tukamuhabwa, P.R. Rubaihayo, K. Dashiell1 and E.Adipala
Department of Crop Science, Makerere University, P.O. Box 7062, Kampala, Uganda1International Institute of tropical Agriculture, PMB 5320, Oyo road, Ibadan, Nigeria

(Received 12 November, 1999; accepted 14 June, 2000)

Code Number: CS00022

Introduction

Pod shattering is the opening of pods along both the dorsal and ventral sutures of the soybean pod ( Glycine max). Shortage of labour and harvesting equipment can postpone harvesting when the farmer is otherwise prepared to harvest, leading to seed losses of 50-100% often associated with pod shattering during dry weather conditions in susceptible varieties (IITA, 1986). Ferraz de Toledo et al. (1994) reported that shattering in soybean was the most important trait among the important characteristics of the soybean plant. Work at the International Institute of Tropical Agriculture (IITA) showed that resistance to shattering was of great economic benefit to farmers in the hot tropics and areas where machines are used for harvesting in terms of reduced yield losses and planning for time and labour (IITA, 1986). Misra et al. (1980) observed that information on genetics of shattering in soybeans was scanty and inconsistent, and little effort had been devoted to improving soybean for resistance to shattering in the tropics where soybean has been observed to shatter easily (IITA,1992). Hoag (1972) estimated losses due to shattering to be as high as 60%, accounting for loss of over 170 m US dollars annually in United States. In field trials at Arlington, USA, Philbrook and Oplinger (1989) observed that shattering contributed 37% of total field loss in soybean.

Caviness, (1969) found susceptibility to shattering to be partially dominant but analysis of pod shattering in F1 populations by Tiwari and Bhatnagar (1992) suggested that susceptibility was dominant or partially dominant to resistance. They further observed additive gene action to predominate in the expression of the trait, although dominance also played a major role. According to Caviness (1969) segregation for soybean shattering in F2 was overlapping suggesting polygenic inheritance. Misra et al. (1980) experienced difficulty in estimation of shattering trait of soybean genotypes, which they attributed to absence of discrete observations at F2 generation, suggesting presence of several genes. Suzuki et al. (1981), however, suggested that if the difference between genotype means was small compared to the environmental variance, even one or two genes could result in continuous variation. Carpenter and Fehr (1986) observed discrete classes of shattering scores in segregating populations of soybean involving a susceptible wild relative Glycine soja and two resistant cultivars. They observed shattering frequency to decrease with each generation of backcrossing to the cultivars and concluded that the shattering trait could be completely eliminated by three to four backcrosses, indicating that only a few genes were involved. Caviness (1963) suggested that four major genes were involved in controlling susceptibility to shattering. On the other hand Tsuchiya (1986) suggested that depending on methodology, minimum number of genes controlling shattering in soybean varied between one and two, while Akpan (1988) reported 6 - 12 genes to be involved. In a recent study by Bailey et al. (1997) it was observed that one major quantitative loci, and a few minor QTLs controlled pod shattering in soybean.

Average broad sense heritability (h2b) estimate of shattering in soybean was put at 0.9 by Caviness (1969), 93% by Tsuchiya (1987) and 98.8% by Tiwari and Bhatnagar (1991) indicating that the character is highly heritable. Heritability in the broad sense (h2b) in self pollinating crops is less informative than heritability in the narrow sense (h2n) which is a direct measure of additive variance. The narrow sense heritability can be achieved by parent offspring regression if parental values are means of both parents (Kearsey and Pooni, 1997) or from the expected mean squares of the progenies of interest (Brim et al., 1961). The presence of epistasis and dominance tend to inflate values of h2n (Fernadez and Miller, 1985) and thus, heritability determined by the regression coefficient in case of random mating offers a more secure approach to h2n than the partition of variance.

The study reported in this paper was carried out to assess the inheritance mechanism of pod shattering in soybean and determine narrow sense heritability and number of genes controlling soybean pod shattering.

MATERIALS AND METHODS

Eight parental lines were characterised for pod shattering in the field at Namulonge (Uganda) in December, 1996 and their level of resistance to pod shattering confirmed using the oven method (Jiang et al., 1991). Genotypes AGS 292 and TGm 737P were susceptible to pod shattering, Gc81090-48, Duiker, TGx 1448-2E and Roan resistant, and genotypes Kabanyolo1 and Samsoy 1 were intermediate.

A total of 15 soybean crosses were made from the eight lines in two sets. The first set comprised of nine crosses; TGm 737P x Duiker, TGx 1448-2E x TGm 737P, TGx 1448-2E x Samsoy 1, TGx 1448-2E x Duiker, Kabanyolo 1 x TGm 737P, Duiker x Kabanyolo 1, Kabanyolo 1 x Duiker, TGx 1448-2E x Kabanyolo 1 and Kabanyolo 1 x TGx 1448-2E. These crosses were made at Namulonge in 1996, second season (October). Their F1 and F2 progenies were planted in 1997 during the first (March) and second (September) seasons, respectively. Seed bulks of each cross were made for all crosses and planted as F2 populations and were grown in a completely randomised design and analysed according to Kearsey and Pooni (1996) to determine heritability, by regressing F2 populations means on value of their mid parents.

A second set of crosses consisted of: TGx 1448-2E x AGS 292, AGS 292 x Roan, Gc81090-48 x AGS 292, TGx 1448-2E x TGm 737P, TGm 737P x Duiker and Roan x TGm 737P which were between susceptible and resistant parental lines. Their F1 and F2 progenies were planted in 1998 during the first (March) and second (September) seasons, respectively. F1 plants from the crosses varied from 5 to 12 plants and yielded an F2 plant population of 59 to 310 plants per cross and were used to study the population distribution at F1 and F2 to determine the number of genes controlling pod shattering.

Determination of maternal influence. Maternal influence on the inheritance of soybean pod shattering was determined using crosses of TGm 737P x Duiker, TGm 737P x Roan and AGS 292 x TGx 1448 - 2E and their reciprocals. Shattering was evaluated in the F2 and F3 generations. Reciprocal crosses were only available for these three crosses, due to difficulties involved in obtaining enough F1 hybrid seed in autogamous crops like soybean. The means of the crosses and their respective reciprocal crosses were compared using Z test ( Fowler and Cohen,1990) to determine the difference between the mean of each cross and its respective reciprocal.

Evaluation procedure for pod shattering. Shattering scores were carried out on the progenies at F1, F2 and F3 generations. The pods in paper bags were allowed to equilibrate for ten days at room temperature after which they were subjected to oven drying at 80 °c for 12 hours. The percentage of pod shattering induced was determined on the 0 – 10 scale used by Bailey et al. (1997) where 0 < 1% = 0, 1-10% =1, 11-20% = 2, 21-30%=3, 31-40% = 4, 41-50% = 5, 51-60%=6, 61-70%=7, 71-80 = 8, 81-90%=9, 91-100%=10. Basing on the scale by Bailey et al. (1997), a 1 - 3 scale was developed in which 0 - 10% shattered = 1, 11 - 70% shattered = 2 and 71 - 100% shattered = 3. Basing on results from the 1 – 3 scale, phenotypic classes were assigned as follows: progenies with the score of 1 were regarded as resistant, progenies with score of 2 as intermediates and 3 as susceptible. A Chi square test to determine the departure of the observed frequencies from the expected frequencies was carried out based on critical values of 4.605, 5.991, 9.210 for two degrees of freedom on the 0.10, 0.05 and 0.01 probability levels, respectively.

Determination of number of genes. Estimation of the number of genes controlling resistance for pod shattering in soybean was carried out according to Bjarko and Line (1987) using the formulae:

(1) F2 generation: n = (GR)2 [1.5 - 2 h(1 - h)] / 8(VF2 - (VPS + VPR + 2VF1)4)
h = (F1M - PR)/(PS - PR)

(2) F3 generation: n = (GR)2 / 5.33 [VF3 - (VPR + VPS)/2]

GR = Genotype range, estimated as phenotypic range of segregating populations or difference betweenthe two parents (PS - PR)
PR = Mean of resistant parent
PS = Mean of susceptible parent
F1M = Mean of F1 progenies
VPR , VPS = Variance of resistant and susceptible respective parents
VF1 , VF2 , VF3 = Variance of F1, F2 and F3 respective generations
n = The estimated number of segregating genes

Number of loci were also estimated using the original Castle - Wright’s method (Zeng et al., 1990) as follows:

(3) Kcw = D2 / 8VG = D2 / 8(VF2 - VE)
where;
Kcw = Number of loci estimated by the original Castle - Wright formula
D = Difference in parental mean (P2 - P1)
VF2 = F2 Phenotypic variance
VE = Environmental variance

RESULTS AND DISCUSSION

The shattering scores of F1 hybrids of susceptible and resistant parents and their parental lines are presented in Table 1. All F1 plants indicated a phenotype closer to the susceptible parent than the resistant parent, suggesting partial dominance for susceptibility. Similar results were reported by Caviness (1969). Genetic ratios for levels of pod shattering resistance at F2 are presented in Table 2. A Chi square test on the ratios obtained in all six crosses fitted the 1 resistant: 3 Intermediates: 12 susceptible model, suggesting dominant epistasis to be present and that shattering resistance is controlled by two genes. The frequency distribution of F2 populations for the 0 - 10 scale is shown in Figure 1. Contrary to observations made by Ting (1946) and Caviness (1969) the frequency distribution of F2 was not overlapping in a normal distribution manner but presented a wide and skewed distribution. The complex distribution observed on F2 progenies can be attributed to epistasis which cause deviations from the expected distribution ( Fehr, 1987; Griffith et al.. 1997).

Table 1. Shattering scores of seven parental lines and their F1 progenies
Genotype/cross Number of plants Mean shatter % Shattering scores1
AGS 292 20 100 3
Duiker 20 10 1
Gc81090-48 20 5 1
Roan 20 0 1
TGm 737P 20 100 3
Kabanyolo 1 20 49 2
Samsoy 1 20 45 2
TGx1444-2E 20 5 1
Roan x TGm 737P 5 100 3
TGx 1448-2E x AGS 292 12 92 3
AGS 292 x Roan 6 100 3
Gc81090-48 x AGS 292 10 90 3
TGx 1448-2E x TGm 737P 5 100 3
TGm 737P x Duiker 5 100 3
TGm 737P x Roan 6 83 3
1Based on 1 - 3 scale, where 1 = 0-10%; 2 = 11-70%; and 3 = 71-100% shattered


Table 2. Phenotypic ratios of resistant (R) : intermediate (I) : susceptible (S) F2 populations when fitted on a 1:3:12 genetic model
Cross Phenotype Observed Expected X2
TGx 1448-2E x AGS 292 R 5 4.88 0.00
I 14 14.63 0.03
S 59 58.50 0.00
Total 78 78.00 0.03***
AGS 292 x Roan R 7 7.81 0.08
I 23 23.44 0.00
S 95 93.75 0.01
Total 125 125.00 0.02***
Gc81090-48 x AGS 292 R 5 3.69 0.47
I 12 11.06 0.08
S 42 44.25 0.11
Total 59 59.00 0.66***
TGx 1448-2E x TGm 737P R 31 19.38 6.98
I 51 58.13 0.08
S 223 232.50 0.39
Total 310 310.00 7.45*
TGm 737P x Duiker R 12 13.38 0.14
I 31 40.13 2.08
S 171 160.50 0.69
Total 214 214.00 2.90***
TGm 737P x Roan R 4 5.00 0.20
I 16 15.00 0.07
S 60 60.00 0.00
Total 80 80.00 0.27***
1Based on 1-3 scale, where 1 = 0-10%; 2 = 11-70%; and 3 = 71-100% shattered
2Significant goodness of fit: *** P<0.10; *P<0.01

A suggested model for the genetic control mechanism of pod shattering in soybean is presented in Table 3. This model suggests that genes at same locus interact resulting in susceptibility alleles being dominant over alleles for resistance plus non allelic interaction (epistasis). An assumption is made on basis of a two gene model, whose genotypes are represented by A+A+B+B+and A-A-B-B- for homozygous dominant and homozygous recessive parental lines, respectively. The observed ratio of 1:3:12 in which the presence of at least one dominant allele (A+) at one locus masks the expression of both alleles present, results in the dominant phenotype, and presence of (B+) in absence of (A+) results in an intermediate phenotype. The final distribution is a reflection of classical dominant epistasis which causes modification of 9:3:3:1 Mendelian ratio (Kearsey and Pooni, 1996) and fits the proposal that epistasis influences soybean pod shattering (Bailey et al. 1997)

Table 3. Relationship between genotype and phenotype in an F2 population showing classical dominant epistasis
Genotype Frequency Phenotype Ratio
A–A–B–B– 1/16 R 1
A–A–B+B+ 1/16 I 3
A–A–B+B– 2/16 I  
A+A+B+B+ 1/16 S 12
A+A+B+B– 2/16 S  
A+A+B–B– 1/16 S  
A+A–B+B+ 2/16 S  
A+A–B+B– 4/16 S  
A+A–B–B– 2/16 S  

In self pollinating plants like soybean, epistasis is more important than dominance and can generate different phenotypes some of which represent real genetic advance over their parents (Allard, 1960). However, Kearsey (1993) reported that non allelic interactions cause problems when predicting response to selection, by preventing identification of the most promising crosses at an early stage in the breeding programme. Kearsey and Pooni (1996) noted that the effect of epistasis is more profound on means as opposed to variances and epistatic effects of generation means are more amenable to analysis than those of variances. Hallauer and Miranda (1988) recommended identification of loci with favourable epistatic combinations for improvement of economically important traits in crops. When epistasis is important, recurrent selection and reciprocal recurrent selection have been recommended as efficient techniques for selecting favourable epistatic gene combinations (Allard, 1960; Moreno-Gonzalez and Cubero, 1993).

The parent - offspring regression analysis indicated a heritability estimate of 0.79 ( Table 4) suggesting that inheritance of resistance to pod shattering was highly heritable. Heritability estimates obtained from regressing offsprings on their mid-parent values are a direct estimate of heritability in the narrow sense (Kearsey and Pooni, 1996), and explains the reason why the estimate is lower than estimates reported by Caviness (1969), and Tiwari and Bhatnagar (1991) which were broad sense estimates. However, the high estimate of heritability in this study may have been inflated by the presence of epistasis (Ferdnadez and Miller, 1985).

Table 4. Analysis of variance for mid-parent regression on F2 progenies for pod shattering
Source DF SS MS F
Regression 1 3.3696 3.3696 64.62**
Error 8 0.4172 0.5215  
Total 9 3.7868    
CV (%) = 6.5; R2 = 0.89; B = 0.79 = h2n
**: Significant difference at P = 0. 01

Using F2 and F3 progenies, estimates of genes ranged from 1.63 to 2.33 following Bjarko and Lines’ (1987) method ( Table 5). The results suggested a mean of 2.01 genes at F2 and 2.06 at F3 suggesting 2 genes being involved in controlling pod shattering. However, the number of loci varied between 1 and 2 using Castle and Wright’s method which Zeng et al. (1990) observed to underestimate the number of loci. These results are in agreement with the observed phenotypic ratio of 1:3:12 at F2. Similar results were reported by Tsuchiya (1986) while contrary estimates were given by Caviness (1963), who indicated four major genes and Akpan (1988) who reported 6 to 12 genes to be responsible for pod shattering in soybean. Other diverging results were reported by Bailey et al. (1997) who showed shattering to be conditioned by one major gene and a few modifier genes.

Table 5. Estimated number of genes controlling soybean shattering in six crosses of soybean at F2 and F3 generations
Cross Number of progenies (n) Filial generation
F2w F2j F3j
TGx 1448-2E x AGS 292 78 1.45 1.97 2.01
AGS 292 x Roan 125 1.74 2.33 2.28
Gc81090-48 x AGS 292 59 1.34 1.63 1.93
TGx 1448-2E x TGm 737P 310 2.11 2.19 2.20
TGm 737P x Duiker 112 1.45 1.90 1.60
TGm 737P x Roan 80 1.48 2.02 2.32
Mean 127.33 1.60 2.01 2.06
westimates by the original Castle - Wright’s method (Zeng et al., 1990); jestimate according to Bjarko and Line (1987)

The results of the maternal influence are shown in Table 6. There were no significant differences between the crosses and their reciprocals indicating absence of maternal influence on pod shattering in soybean, and suggesting that choice of maternal parent is not important in a hybridisation programme focussing on improvement of soybean for resistance to pod shattering.

Table 6. Maternal influence on soybean shattering in F2 and F3 populations of three crosses and their reciprocals
Cross Filial generation
F2 F3
Number of observations Mean shatter score Standard error Number of observations Mean shatter score Standard error
TGm 737P x Duiker 100 4.20 ±0.15 100 3.47 ±0.71
Duiker x TGm 737P 100 4.50   100 4.48  
TGm 737P x Roan 80 4.09 ±0.25 70 4.27 ±0.16
Roan x TGm 737P 80 4.58   70 4.49  
TGx 1448-2E x AGS 292 80 4.44 ±0.14 140 4.19 ±0.01
AGS 292 x TGx 1448-2E 80 4.16   140 4.21  

In conclusion, pod shattering in soybean is under control of two genes and is partially dominant over resistance. Inheritance of pod shattering is influenced by non allelic interactions resulting in classical dominant epistasis and the pod shattering trait in soybean is highly heritable with narrow sense heritability of 0.79 but is not infuenced by maternal effects,

ACKNOWLEDGEMENT

The authors thank the National Agricultural Research Organisation of Uganda and the International Institute of Tropical Agriculture for funding the study.

REFERENCES

  1. Akpan, E.S. 1988. Inheritance of pod shattering and its relationship with some Agronomic characters in soybeans. MSc. Thesis, Ahmadu Bello University, Zaria, Nigeria.
  2. Allard, R.W. 1960. Principles of Plant Breeding. John Wiley and Sons Inc., New York. 485 pp.
  3. Bailey,M.A., Mian, M.A.R., Carter, T.E., Ashley, D.A. and Boerma, H.R. 1997. Pod dehiscence of soybean: Identification of quantitative trait loci. The Journal of Heredity 88:152- 154.
  4. Bjarco, M.E. and Line, R.F. 1987. Heritability and number of genes controlling leaf rust resistance in four cultivars of wheat. Phytopathology 78:457- 461.
  5. Caviness, C.E. 1963. A physiological and genetic study of shattering in soybean. PhD Dissertation, University of Missaouri.
  6. Caviness, C.E. 1969. Heritability of pod dehiscence and its association with some agronomic characters in soybeans. Crop Science 9:207-209.
  7. Fehr, W.R., 1987. Principles of cultivar development. Vol. 1. Theory and Techniques. Macmillan, New York. 536 pp.
  8. Fernandez, G.C. and Miller, J.C. 1985. Estimation of heritability by parent - offspring regression. Theoretical and Applied Genetics 70:650 - 654.
  9. Ferraz de Toledo, J. and Alves de Almeida. L. 1994. Genetics and breeding. In: Tropical soybean: Improvement and Production. FAO, Rome. 254 pp.
  10. Fowler, J. and Cohen, L. 1990. Practical Statistics for Field Biology. John Wiley and Sons, Chichester, UK. 227 pp.
  11. Griffiths.A.J.F., Miller, J.H., Suzuki,T.D., Lewontin R.C. and Gelbart, W.M. 1997. An Introduction
  12. to Genetic Analysis. W.H. Freeman and Company, New York. 916 pp.
  13. Hauller, A.R. and Miranda, J.B. 1981.Quantitative Genetics in Maize Breeding, Iowa State University Press,USA. 468 pp.
  14. Hoag, D.L. 1972. Properties related to soybean shatter. University of Illinois, Urbana. USA.
  15. International Institute of Tropical Agriculture (IITA), 1986. A laboratory method for evaluating resistance to pod shattering in soybeans. Annual Report. 58 - 59. IITA, Ibadan, Nigeria.
  16. International Institute of Tropical Agriculture (IITA), 1992. Soybean Breeding. Grain Legume Improvement Program, Archival Report. IITA, Ibadan, Nigeria. pp.1 - 8.
  17. Kearsey, J.M. and Pooni, S.H. 1996. The genetical analysis of quantitative traits. Chapman and Hall, London, UK. 381 pp.
  18. Misra, R., Singh, Sharma, S. and Mehta, S. 1980. Note on induced variation in shattering habit of soybean. Indian Journal of Agricultural Science 51:678-80.
  19. Moreno-Gonzalez and Dudley. J.W. 1981. Epistasis in related and un related maize hybrids
  20. determined by three methods. Crop Science 21:645- 651.
  21. Philbrook, B.D. and Oplinger, E.S.1989. Soybean field losses as influenced by harvest delays.
  22. Agronomy Journal 81:251-258.
  23. Suzuki, D.T., Grifths, A.J. and Lewontin, R.C. 1981. An Introduction to Genetic Analysis.
  24. N.H. Freeman and Co., San Francisco, U.S.A. 910 pp.
  25. Ting, C.L. 1946. Genetical studies on the wild and cultivated soybean. Journal of American Society of Agronomy 38:381-393.
  26. Tiwari, S. and Bhatnagar, P. 1991. Pod shattering as related to other agronomic attributes in soybean. National Research Centre for soybean (ICAR), Indore 425 001, India.
  27. Tiwari, S. and Bhatnagar, P. 1992. Genetics of pod shattering in soybean. National Research
  28. Centre for soybean (ICAR), Indore 425 001, India.
  29. Tsuchiya, T. 1986. Studies on shattering resistance in soybean Glycine max breeding. Report of Hokkaido Prefectural Agricultural Experi-ment Station 58:1-53.
  30. Tsuchiya, T. 1987. Physiological and genetic analysis of pod shattering in soybean. Japan
  31. Agricultural Research Quarterly 21:166-175.
  32. Zeng, Z.B., Houle, D. and Cockerham C.C. 1990. How informative is Wright’s estimator of the number of genes affecting a quantitative character? Genetics 126:235- 247.

The following images related to this document are available:

Line drawing images

[cs00022a.gif]
Home Faq Resources Email Bioline
© Bioline International, 1989 - 2024, Site last up-dated on 01-Sep-2022.
Site created and maintained by the Reference Center on Environmental Information, CRIA, Brazil
System hosted by the Google Cloud Platform, GCP, Brazil