African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 9, Num. 2, 2001, pp. 377-384

African Crop Science Journal, Vol. 9. No. 2, pp. 377-384

STUDIES ON GENETIC VARIATIONS IN A SORGHUM VARIETY IRRADIATED WITH COBALT 60 (CO⁶⁰)

D.A. ABA, C.C. NWASIKE, M.YEYE and A.A. ZARIA

Plant Science Department, Institute for Agricultural Research, Ahmadu Bello University, Samaru, Zaria, Nigeria

(Recieved 10 May, 1999; accepted 8 June, 2000)

Code Number: CS01019

INTRODUCTION

Sorghum (Sorghum bicolor Moench) is a major staple food in many parts of the world. It has very extensive use as a food crop in Africa, Asia, and is a major feed grain in North America and South Africa (Pickett and Oswalt, 1971). In areas where sorghum is used as a food crop the food recipes in a given area have imposed severe restriction on the types of grain that may be used in varietal development. In India, the preference is primarily for the white, creamy white to yellow and pearly seeds types. In Africa, the variability of choice is tremendous with the various types being the primary factor in food preferences. As a feed grain in the United States, it is utilised as whole grain (Pickett and Oswalt, 1971).

In Nigeria about 40 percent of the calories from all food stuffs come from cereals and of this, sorghum contributes 50 percent, with almost all the production coming from the Northern zone of the country (Oke, 1976). Sorghum is a very versatile and hardy crop that is adapted to varying environments, from the wet guinea savanna through to the vast dry sudan savanna and semi- arid sahel zones in Nigeria. The sorghum growing areas in Nigeria lie between latitudes 6° 30’N and 13° 55’N of the country.

Sorghum (Sorghum bicolar (L.) Moench), locally called guinea corn is a very important crop in the savanna and derived savanna zones of Nigeria, grown on about 6 million ha which cover about 9.7% of the world sorghum growing areas. This relates to about 46% of the total land area devoted to cereals production with a current estimated production of about 8 million metric tonnnes (NAERLS, 1996; FAO/ICRISAT, 1996) in Nigeria. Of this amount, 3.12 million tonnes are consumed as human food and the remaining 0.88 million tonnes is used for various industrial purposes, the most important being brewing (Ikediobi, 1989, unpubli.).

In developing various types of sorghum, the breeder is always faced with the low genetic variability existing in the germplasm. To overcome this constraint, a breeder may employ a breeding method like hybridisation and selection between and within populations and/or use mutation breeding to provide the desired variability. These procedures help to broaden the genetic base of the breeding materials (Musoke et al., 1999). Parra-Negrette et al. (1984) have evaluated induced mutation and hybridisation methods for producing genetic variability in 15 quantitative characters of sorghum. Their results showed large variability for plant maturity, plant height, panicle length and exertion.

Much of the sorghum produced in Nigeria comes from the four ecological zones of the country: Sahel, Sudan, Northern Guinea and Southern Guinea savannas, where the rainfall is erratic and unpredictable and periods of droughts frequent. This situation calls for the development of early maturing, shorter and higher yielding sorghums that are resistant to disease and insect pests. The objectives of this study, therefore, were to (i) study the genetic variation in some agronomic characters of sorghum including the yield components, (ii) study the relationship of these characters to one another, and (iii) study the extent to which these characters are inherited.

MATERIAL AND METHODS

Genetic variability was created in sorghum "Short Kaura" with gamma-ray from cobalt source (Co⁶⁰). This variety is a medium plant of about 1.5 m with broad leaf, thick stem, compact head (panicle) and large yellow grains, and matures in about 125 to 150 days. Using genetic effective cell number (GECN) of 2, and a mutation rate of 10-3, a total of 2300 seeds were irradiated with 250 Gy as outlined by Nwasike (1985, unpubli.). Compensating for 70 per cent viability of seeds, 65 per cent survival, and 80 per cent fertility after treatment, the total number of seeds used in M₁ were calculated as;

For plant density of 500 per M₂ in M₁ plus 10 per cent for planting control, the seeds required was calculated as;

Since 7 seeds/m^-2 was the seeding rate and 20 seeds per row in M₂, the plot size was determined as 3159/7 seeds/m² = 451.28 seeds m^-2.

Treated M0 seeds were planted in the field during the off season 1985. Selected M₁ seeds were grown as bulk and, subsequently, bulked seeds were grown as M₂ population. Individual panicles or plants selected in M₃ were planted ear-to-row for family selections. Ninety family lines were selected in M₃, after which 34 family selections were made in M₄.

Thirty-four M₄ family selections were threshed as individuals parnicles and each divided into three equal portions to allow for the same materials to be evaluated for three years. Each of the 3 portions served as a treatment, and were evaluated in a randomised complete block design each year for three years (1989-1991) at three sites of the Institute for Agricultural Research farms, Ahmadu Bello University, Zaria, Nigeria. At each site there were four replications with four row plots (6m long) and the plants were spaced 25 cm within the row. Prior to ridging 32 kg N ha^-1 of single superphosphate fertiliser (P₂O₅) was broadcasted as basal dressing. Split application of nitrogen fertiliser (Urea) was done, the first 32 kg N ha^-1 was basal, while 32 kg N ha^-1 was applied 3 weeks after thinning (top dressing).

Correlation coefficients between characters x₁ and x₂ were calculated using the formula of Singh and Chaudhary (1977) as;

where:
r (x₁x₂) = the correlation between characters x₁ and x₂
Cov (x₁ x₂) = the covariance between characters x₁ and x₂
V (x₁) = the variance of character x₁ and
V (x₂) = the variance of character x₂

Heritability. Estimates of broad sense heritabilities were made from family means using variance components as given by the formula:

σ²_g = genotypic variance of a character
σ²_ph = phenotypic variance of a character
h² = heritability of a character (broad sence)

RESULTS

The mean square values for the 22 characters for the individual years as well as for the three years combined indicated that in 1989, all the characters had highly significant (P=0.01) mean squares except protein percentage and width of spikelet. A similar trend was observed for the mean squares of the characters in 1990 and 1991 where we had highly significant (P=0.01) mean squares for all the characters except for protein percentage and width of spikelet which were not significant in the two years. There were also highly significant (P=0.01) differences between years for all characters except width of spikelet. Likewise, there were highly significant genotype x year interaction mean squares for all characters except width of spikelet which showed a non-significant interaction.

Broad sense heritability estimates were computed for all characters since they had positive σ²_g, basing on the components of variance estimates. Heritability from the combined analysis (Table 1) showed that percentage protein which had almost the lowest variance components for σ²_g, σ²_ph, and σ²e had the highest heritability value of 77.9%, while length of first leaf blade had the lowest heritability value of 5.5% (Table 3).

In the combined correlation matrix (Table 2) grain weight per panicle was positively correlated to length of panicle, spikelet number per panicle, grain weight per spikelet, days to 50% flowering, length of spikelet, first and second leaf blades. Days to 50% flowering was negatively correlated with all non-reproductive structures except length of 1st, 2nd and 3rd leaf blades, and length of spikelet. Grain weight per spikelet was positively correlated to length of spikelet, width of spikelet, length of 2nd, 3rd and 4th leaf blades, lengths of 2nd and 3rd internode. Significant (P=0.05) correlations were found between lengths of 1st and 2nd leaf sheaths, between 1st and 2nd internodes, between 2nd and 3rd internodes and between 3rd and 4th and 5th internode. Highly significant (P=0.01) correlations were also found between lengths of 2nd and 3rd leaf sheaths and between 3rd and 4th leaf sheaths.

The results of principal component analysis for the different characters are shown in Table 3. Seven characters accounted for 68% of the total variance. These characters are length of second leaf sheath, length of third leaf sheath, length of fourth leaf sheath, length of first internode, length of second internode, length of third internode, length of fourth internode and length of fifth internode. This indicates that these characters were the ones mostly affected by the irradiation, thus creating most of the variability in them. Since these characters are positively correlated to most yield components, they could be selected for as an indirect way of improving yield in our plant types.

DISCUSSION

A genetic plant type may be defined as a genetically conditioned association of morphologic traits (Morishima et al., 1967), represented by a certain combination of leaf, stem (culm) and panicle characters. Growth type or shape is easily recognised, but difficult to measure.

The result of the analysis of variance for the three years combined showed significant (P=0.01) mean squares for years for all the characters except for grain weight per spikelet and width of spikelet, an indication that the yearly changes were high and could influence the performance of the genotypes (Taye Kufa et al., 2001, this volume). The variances due to genotype x year interactions were large, similar to the findings of Morishima and Oka (1968), and indicate that the plants response to the growing seasons was genetically controlled. According to Morishima et al. (1967), selection for seasonal adaptability and for high yielding capacity may be made simultaneously within a season. In cowpea, Ndiaga Cisse (2001, this volume) suggests that more than one season is needed to identify genotypes adapted to semi-arid regions.

The genotypic variance was high for all the characters except for percentage protein and width of spikelet. All other characters had lower genotypic variances (σ²g) than their respective phenotypic variances and error variances, except for percentage protein. This probaly indicates that the gene controlling these characters are almost stable at M₄ generation, which tends to agree with the findings of Scossiroli et al. (1960), Swaminathan (1961), Scossiroli and Polligrini-Scossiroli (1962), Gaul (1964) and Borojevic (1963, 1965) who reported that variability of quantitative characters in sorghum treated with mutagen increased up to four times in the M₂ and M₃, gradually decreased in later generations and stabilised around M₅. Menkir et al. (1997) using RAPD assessment method in some sorghum landrances found high genetic diversity within the races bicolor and guinea, but low in the Kafor. All the characters showed lower genotype x year interaction than their respective genotypic and phenotypic variances. This suggests that the years were adequate for measuring the genetic diversity of the characters (Quendeba et al., 1995). It also indicated that the years did not influence the performances of the traits which make it difficult to select for these traits in these particular years (Opeke and Fakorede, 1986).

Heritability estimates computed from the variance components showed that percentage protein, which had almost the lowest variance components had the highest heritability of 77.4 percent, in the combined years. This indicates that their genotypic variances were higher or larger than their respective phenotypic variances. This could mean that their genetic variation is large due to major genes controlling the characters and that selection would be effective in the desired direction for each of these traits in the population (Opeke and Fakorode, 1986).

A correlation matrix represents the inter-relationship of concerned variates distributed in a multidimensional space (Morishima and Oka, 1968). Length of panicle was positively correlated to all characters in the combined correlation matrix except grain weight/spikelet and days to 50% flowering. Selecting for the non reproductive structures which account for a large proportion of the variations in the principal component analysis will likely improve yield through increase in length of panicle and lengths of spikelet.

Quendeba et al. (1995) reported that grain yield and spike yield of millet landraces were significantly correlated with spike length, spike girth, stem diameter, spike number/plant, plant height and 1000 grain weight. They also reported a negative correlation between spike length with peduncle exertion, but positive correlation to spike yield and grain yield.

Principal component analysis indicated that seven factors accounted for 68% of the total variation (Table 3). The first factor accounted for 22.6% of the total variation and was also associated with length of 1st leaf sheath, length of 2nd leaf sheath, length of 3rd leaf sheath, and 1st - 5th internodes lengths. Factor 2 accounted for 11.85% of the total variation which is explained by 1st leaf blade, 2nd leaf blade, 3rd leaf blade, days to 50% flowering, length of spikelet and grain weight per spikelet. Thus, selecting for plants with long leaf blades, which are tall and early maturing may lead to increase in grain weight/spikelet (yield), as an indirect approach of improving yield in these plant types.

REFERENCES

1. Borojevic, K. 1963. Genetic advances in the height of induced micro and macro mutations in the breeding of polyploid plants. In: Proceedings of Symposium Application of Nuclear Energy in Agriculture, Rome, 1961: 241 pp.
2. Borojevic, K. 1965. The effect of irradiation on the number of kernels per spike in wheat, the use of induced mutations in plant breeding. Rep. FAO/IAEA. Tech. Meeting Rome 1964), Pergaman Press, Oxford (1965) 504.
3. FAO/ICRISAT, 1996. A world sorghum and millet economics: Facts, trends and outlook. Food and Agricultural Organisation of the United Nations, Rome, Italy and International Crops Research Institute for the Semi-Arid Tropics, Patancheru 502324, A.P. India. 68pp.
4. Gaul, H. 1964. Mutation in plant breeding. Radiation Botany 4:155.
5. Menkir, A. Peter, G. and Gebisa, E. 1997. RAPD based assessment of genetic diversity in cultivated races of sorghum. Crop Science 37:564-569.
6. Morishima, H., Oka, H.I. and Chung, T. 1967. Analysis of seasonal change in genetic plant type. Japan Journal of Breeding.
7. Morishima, H. and Oka, H. J. 1968. Analysis of genetic variation in plant types of rice. III. Variations in general size and allometric pattern among mutant lines. Japan Journal of Breeding 18:181-189.
8. Musoke, C., Rubaihayo, P.R. and Magambo, M. 1999. Gamma rays and ethylmethane sulphonate in vitro induced fusarium wilt resistant mutants in bananas. African Crop Science Journal 7:313-320.
9. NAERLS, 1996. Prospects and problems of the 1996 cropping season. A report of a study conducted by the National Agricultural Extension and Research Liaison Services (NAERLS) and Agricultural Planning, Monitoring and Evaluation Unit (APMEU) 2 - 4 October, 1996. NAERLS, Ahmadu Bello University, Samaru, Zaria, Nigeria. 62 pp.
10. Ndiaga Cisse, 2001. Genotype x row spacing and environment interaction of cowpea in semi-arid zones. African Crop Science Journal 9:359-367.
11. Opeke, B.O. and Fakorede, M.A.B. 1986. Genetic variability, heritability estimates, correlations and predicted response to S1 selection for seedling emergence and yield in three maize populations. Nigerian Journal of Agronomy 1:1 - 4.
12. Oke, O.L. 1976. The potential of millet and sorghum as food in Nigeria. In: Proceedings of a Symposium on Sorghum and Millet for human food. Vienna 11-12 November, 1976. IAEA, Vienna.
13. Parra-Negrette, L.E., Mendoza, O. and Orfizcereceres, J. 1984. Comparison of mutant lines. Parental collections and elite lines. North American Sorghum Newsletter 29:14-15.
14. Pickett, R.C. and Oswalt, D.L. 1971. Sorghum nutritional quality improvement: In: Sorghum in Seventies. Oxford & IBH Publishing Co., New Delhi, Bombay, Calcuta. pp. 445-463.
15. Quendeba, B., Ejeta, G., Hanna, W.W. and Anand, K.K. 1995. Diversity among African Pearl millet landrace populations. Crop Science 35:919 - 924.
16. Scossiroli, R.E., Palenzona, O. and Rusmani, B. 1960. Radiation experiments on Triticum durum and T. vulgare: Effects of ionizing radiations on seeds. In: Proceedings of a Conference held at Karlscrune, IAEA, Vienna. pp. 373-386.
17. Scossiroli, R.E. and Polligrini-Scossiroli, S. 1962. Use of irradiation applied to seed to induce new genetic variability for quantitative traits in durum wheat. In: Symposium Genetic and Wheat Breeding. Agricultural Research Institute, Hung Academy of Science. pp. 231-236.
18. Singh, R.K. and Chaudhary, B.D. 1977. Biometrical Methods in Quantitative Genetics Analysis. Kalyani Publishers, New Delhi - Ludiana. 317 pp.
19. Swaminathan, M.S. 1961. Evaluation of the use of induced micro-and-macro mutations in the breeding of polyploid crop plants. In: Symposium Application of Nuclear Energy in Agriculture, Rome, 1961. IAEA, Vienna, Austria. pp. 241-277.
20. Taye Kufa, Tesfaye Shimber, Alemsged Yilma, Anteneh Netsere and Endale Taye, 2001. The impact of close spacing on yield of arabica coffee under contrasting agro-ecologies of Ethiopia. African Crop Science Journal 9:401-409.