Puccinia sorghi, the common rust of maize in India, is one of the major disease problems, especially in rabi (= winter) crops in the state of Bihar. The two widely-grown hybrids, Hi-Starch and Ganga Safed 2, have shown a high degree of susceptibility. The known monogenic sources of resistance carrying gene Rp-d have been rated to be susceptible at Kalimpong, a location in the North-eastern Himalayas (Payak et al., 1974, Indian J. Gen. 34:31-35). It was, therefore, considered desirable to identify and utilize more broad-based or generalized type of resistance. In the All India Coordinated Maize Pathology Project, the inbred lines CM (= Coordinated Maize) 103, CM 104, CM 105, CM 106, CM 113, CM 500, GE 440 and Eto 25 have been rated to be resistant on the basis of their exposure to pathotypes prevalent at Bajaura (North Western Himalayas), Hyderabad (Andhra Pradesh) and Dholi (Bihar).
A study was initiated in 1972-73 to determine the genetic architecture of rust resistance employing the approaches of combining ability and di-allel analyses and estimation of gene effects. Heritability (narrow sense) as well as heterosis were also worked out. The trial was conducted at Dholi and Hyderabad in randomized block design under artificial disease epidemic. The parental materials consisted of six resistant (Eto 25, CM 103, CM 104, CM 105, CM 500 and GE 440) and two susceptible (CM 201 and PI 217407) inbred lines. Data obtained on the eight parental lines, 28 possible F1 single crosses (reciprocals mixed), 28 F2 generations and two sets (28 each) of back crosses were subjected to the above-mentioned biometrical analyses. Here we present in a summarized form the results and interpretations that have been arrived at.
Efforts to place the inheritance on a classical Mendelian basis were not successful. Significant deviation and heterogeneity components at Hyderabad and Dholi with the exception of 13:3 ratio of F2 and 3:1 ratio in test cross at the latter location indicated that on the whole the data did not fit in the expected segregation ratios. Thus the continuous nature of variation and failure to place the inheritance on a definite but small number of genes suggested that resistance to P. sorghi in the materials studied is controlled by polygenes. The plants were scored for disease according to 1-5 scale (1 = no disease and 5 = 76% or more leaf area covered). The disease scores in lines ranged from 1.5 to 3.0. However, the rust incidence was adequate enough at both the locations to draw valid conclusions. Original as well as transformed data were utilized for analyses.
The highly significant variation due to general combining ability (g.c.a.) at both locations suggested that it was playing a greater role in the expression of resistance. Lines CM 105, GE 440, CM 105, CM 103 and Eto 25 at Hyderabad and GE 440, CM 104, CM 105 and Eto 25 at Dholi gave negative values for g.c.a. thereby indicating that they were the best general combiners in the order of listing. Pooled analysis revealed that GE 440, CM 105 and CM 104 were the best general combiners. PI 217407 proved to be the poorest general combiner in location-wise as well as in the pooled analysis as it had the highest positive value for g.c.a.
In a similar type of study conducted at Hawaii in 1977 by Kim and Brewbaker (Crop Sci. 17:456-461) CM 105 was determined to be a good general combiner for resistance to this rust. However, CM 104 was rated by them as susceptible which in our study showed resistant reaction throughout at both locations. It probably suggests that the pathotype(s) of Puccinia sorghi prevalent in Hawaii may be different from those present in India.
The mean rating values and g.c.a. effects for the lines included in both these studies showed good agreement suggesting that the per se performance of the lines was a good indication of their ability to transmit resistance. The specific combining ability (s.c.a.) effects contributed by combinations among resistant lines were non-significant. Susceptible x susceptible cross CM 201 x PI 217407 was not as susceptible as expected; this may be due to recessive alleles at different loci. The data revealed that while stability of performance was lacking for both the combining abilities, g.c.a. was more variable over locations. This observation highlights the importance of testing at more than one location.
The variance component analysis showed that both additive and non-additive gene systems are operative for resistance. Variances due to additive and non-additive gene effects were of the same magnitude at Hyderabad but at Dholi that due to the former was greater. Pooled analysis also showed that variance due to additive effects was greater than that arising out of non-additive ones or dominance. The inbred line, CM 105, contained a higher number of dominant genes for resistance. Graphic analysis showed the presence of over-dominance. Heritability estimates (narrow sense) were up to 40%, which is comparable with the value of 47% in the Hawaiian study.
R. C. Sharma and M. M. Payak
Return to the MNL 53 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page